Journal of Electromyography and Kinesiology 17 (2007) 568–577

www.elsevier.com/locate/jelekin

Similar response of agonist and antagonist muscles after eccentricexercise revealed by electromyography and mechanomyography

Artur Jaskolski *, Renata Andrzejewska, Jarosław Marusiak,Katarzyna Kisiel-Sajewicz, Anna Jaskolska

Department of Kinesiology, University School of Physical Education, ul. Rzezbiarska 4, 51-629 Wrocław, Poland

Received 14 February 2006; received in revised form 18 May 2006; accepted 31 May 2006

Abstract

The purpose of this study was to investigate the influence of eccentric contractions (ECC) on the biceps (BB) and triceps brachii (TB)muscles during maximal voluntary contraction (MVC) of elbow flexors using electrical (EMG) and mechanomyographical activities(MMG). Each of 18 male students performed 25 submaximal contractions (50% MVC) of the elbow flexors. Root mean square ampli-tude (RMS) and median frequency (MDF) were calculated for the EMG and MMG signals recorded during MVC. All measurementswere taken before, immediately after, 24, 48, 72, and 120 h post-ECC from the BB and TB muscles. MVC was reduced by 34% imme-diately after exercise and did not return to the resting value within 120 h (P 6 0.05). The EMG MDF decreased significantly (P 6 0.05) inboth muscles after ECC. The MMG RMS at 24 h, 48, 72 and 120 h post-ECC was significantly lower compared to that recorded imme-diately after ECC in both muscles (P 6 0.05). The present research showed that (i) there were similar changes in electrical and mechanicalactivities during MVC after submaximal ECC in agonist and antagonist muscles suggesting a common drive controlling the agonist andantagonist motoneuron pool, (ii) the ECC induced different changes in EMG than in MMG immediately after ECC and during 120 h ofrecovery that suggested an increased tremor and contractile impairments, i.e., reduced rate of calcium release from the sarcoplasmic retic-ulum (acute effect), and changes in motor control mechanisms of agonist and antagonist muscles, and increased muscle stiffness (chroniceffect).� 2006 Elsevier Ltd. All rights reserved.

Keywords: Biceps brachii; Triceps brachii; Eccentric exercise; EMG; MMG

1. Introduction

Activities that engage a high component of eccentriccontractions (ECC) typically result in muscle damage(Clarkson et al., 1986). The well-documented symptomsof muscle damage include prolonged strength loss,delayed-onset muscle soreness (DOMS), increased stiffnessand swelling (Newham et al., 1983; Friden et al., 1983;Clarkson et al., 1992). Prolonged impaired force produc-tion after ECC is considered to be one of the most validand unfailing indirect measures of muscle damage inhumans (Warren et al., 1999). Isometric strength is reduced

1050-6411/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jelekin.2006.05.002

* Corresponding author. Tel.: +48 71 3473627; fax: +48 71 3473609.E-mail address: artur.jaskolski@awf.wroc.pl (A. Jaskolski).

immediately after ECC and recovery is progressive andprolonged. Howell et al. (1993) reported an immediate35% force loss of the elbow flexors after submaximal eccen-tric exercise and implied that the recovery time may be aslong as 5–6 weeks.

Failure to maintain the required or expected forcecaused by ECC was accompanied by (a) no changes inthe muscle’s electrical activity of the first dorsal interosseusmuscle (Bajaj et al., 2002) and biceps brachii (BB) muscle(Pearce et al., 1998; Michaut et al., 2001), (b) an evidentincreasing tendency in electrical activity of the elbow flex-ors (Sbriccoli et al., 2001), or (c) a decrease in electricalactivity of the quadriceps muscle (Hortobagyi et al.,1998). The discrepancies among the studies are probablyrelated to a different protocol (including intensity of eccen-

A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577 569

tric exercise and control contractions) applied to the differ-ent muscle groups tested. When EMG is recorded simulta-neously with force during maximal voluntary contraction(MVC), the capability of the contractile elements to answerthe electrical signals can be assessed by a MVC force/EMGRMS ratio that provides information as to whether thereduction in force is due to neural factors or contractile fac-tors (Komi and Viitasalo, 1977). Michaut et al. (2001) andMcHugh et al. (2000) did not find any post-ECC changes inthe neuromuscular efficiency of the BB, while in the studyof Deschenes et al. (2000) done on the quadriceps musclethe efficiency significantly decreased and lasted through10 days following exercise.

During voluntary contraction, the active motor unitsgenerate pressure waves that are recorded by different sen-sors and named mechanomyogram (MMG) (Orizio, 2004).It has been suggested that the MMG may detect thechanges in mechanical properties of active muscle (Orizio,2004) and pain during submaximal exercise (Madeleine andArendt-Nielsen, 2005). To our knowledge there is only onestudy that examined changes in agonist MMG activity dur-ing the recovery period following ECC (Bajaj et al., 2002).These authors reported an MMG amplitude increaseimmediately after, 24 and 48 h after ECC as compared topre-ECC (baseline) in the first dorsal interosseous muscleof the hand (agonist muscle). There is no study reportingMMG activity of antagonist muscle. In addition, simulta-neous recording and analysis of the EMG and MMG sig-nals have been used to indicate electromechanicalefficiency and muscle stiffness (Barry et al., 1990); both ofthese can change after ECC. During fatiguing sustainedisometric contraction at 20%, 40%, and 60% MVC, Mama-ghani et al. (2002) recorded different changes in EMG andMMG of agonists.

It is known that the force and torque generated arounda given joint depend on the agonist and antagonist muscleactivity (Falconer and Winter, 1985). However, few studieshave examined the acute effect of ECC in antagonist musclepairs based on EMG recordings (Michaut et al., 2001).Michaut et al. (2001) did not record any significant changesin EMG amplitude of the BB and triceps brachii (TB) mus-cles at MVC within a week after a bout of maximal ECC ofelbow flexors. The similar behavior of the pair of antago-nist muscles may support the notion that central nervoussystem shares excitatory and inhibitory input to agonistmotor neurons pool and subpopulation of the antagonistmotor neuron pool; the phenomenon known as a commondrive. Moreover, reversal order of motor unit recruitmenthas been documented during maximal, as well as duringsubmaximal ECC (Enoka, 1996; McHugh et al., 2000).Based on this common drive theory and the preferentialactivation of the fast-twitch motor units in maximal andsubmaximal ECC, we expected similar changes in theEMG and MMG activities in a pair of antagonist musclesof the elbow joint immediately after submaximal ECC andduring 5 days of recovery. We also hypothesized that thechanges in the MMG after ECC would be different from

those in the EMG of a pair of antagonist muscles becausethe MMG assesses not only motor unit activation strategy(as does EMG) but also changes in muscle stiffness (EMGdoes not).

The aim of this study was to examine the changes inEMG and MMG recorded during MVC from the bicepsand triceps brachii muscles (BB and TB, respectively) aftermuscle damage induced by submaximal ECC and to followthese changes during a 5-day recovery.

2. Methods and subjects

2.1. Subjects

Eighteen non-weight-trained men volunteered to partic-ipate in this study (mean age 23 ± 1.3 years, mean stature178 ± 3.8 cm, mean body mass 79 ± 10.5 kg). Accordingto the Declaration of Helsinki, each subject gave writtenpermission after being informed of the nature of theexperiment, and completed a health history questionnaire.Participants were apparently healthy and free of anymuscular and skeletal disorders and did not take anti-inflammatory medicine that might interfere with theinvestigation. None was involved in any specific trainingprogram of the upper limb musculature at the time of thestudy. The experimental procedure was accepted by thelocal ethics committee.

2.2. Experimental procedure

The present study consisted of five sessions carried outat the same time of day (±2 h) (Fig. 1). The participantswere familiarized with the isokinetic and isometric dyna-mometer and testing procedure two days before testing ses-sions (day 0). An electrode location on the BB (long head)and TB (lateral head) was also determined as the place withthe biggest EMG RMS. Although the muscle belly is notthe optimal location for EMG electrodes, the EMG/MMG probes were located at the belly to obtain thebest-quality MMG signals. The positions of the electrodesover the two muscles were marked on the skin with awaterproof permanent marker to ensure same electrodepositioning throughout the experiment.

The first testing session (session 1) consisted of measure-ments done immediately before and after 25 submaximalECC (25 ECC). Sessions 2, 3, 4, and 5 consisted of mea-surements conducted during the recovery period: session2 = 24 h (Post 24), session 3 = 48 h (Post 48), session4 = 72 h (Post 72), session 5 = 120 h (Post 120) afterECC (see Fig. 1). Before each session, subjects performeda standard warm-up followed by 6 min rest (10 times ofupper limb bends in the elbow joint in sagittal plane with-out external load, with 1.5 kg, and 2.0 kg, and 10 pushupsagainst a wall). Myoelectrical and mechanical activities ofthe BB and TB muscles and the corresponding MVC forcewere assessed during each of the five sessions, and analyzedfor changes in amplitude and frequency domain. The three

25 x ECC

50 % MVC at 60º/ s

from 60º to 150º

ECC

Warm up

100% MVC at 90º with EMG and

MMG (3 min rest)

4min rest

Post-ECC

Warm up

100% MVC at 90 º with EMG and MMG

(3 min rest)

Post - I A

100% MVC at 90º with EMG and MMG

(3 min rest)

Post 24h, 48h, 72h, 120h

Session 1 Sessions 2, 3, 4, 5

Control

Pre-ECC

Fig. 1. Schematic representation of testing protocol. Control = measurements before ECC (baseline); IA = measurements immediately after ECC; 24, 48,72, 120 h = measurements at the respective hours after ECC; MVC = maximal voluntary contraction.

570 A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577

signals were continuously and simultaneously stored on apersonal computer Pentium II 300 MHz. Three minutesof rest were given between the two MVC tests. Four min-utes of rest was applied between MVC measurement andthe 25 ECC.

The laboratory temperature was controlled and kept at20–23 �C. A participant and two testers were in the testingroom.

2.3. Protocol

2.3.1. MVC assessment

The individual MVC force of the elbow flexors of thenon-dominant hand was measured with the BIODEXdynamometer at a joint angle between arm and forearmof 90� (1.57 rad). Subjects were instructed to exert theMVC as fast and hard as possible when a sound signalwas emitted by a computer, and to release (relax) the forceas fast as they could at the second signal (3 s later). Thepeak force of each 3-s value was determined and the meanof the two measurements was used for analysis because themean value of the trial scores is usually the best indicationof a subject’s typical performance and is more reliable thanthe best score (Heinonen et al., 1994). The force signal wasamplified and recorded on personal computer with sam-pling rate 10 kHz.

For six subjects, antagonist muscle MVC force was alsomeasured (during elbow extension) to check for a potentialexercise-induced changes in their neuromuscular efficiency.

2.3.2. Eccentric exercise

Subjects were seated in the Biodex dynamometer (Bio-dex System 3: Biodex Shirley) with their non-dominantarm flexed to approximately 40� and their hand holdingthe handle of the arm attachment. The rotational axis ofthe dynamometer was aligned to the epitrochlea–epicon-dyle axis of the arm.

Subjects were required to resist elbow extension gener-ated by the dynamometer, which was set to yield at 50%of the subject’s MVC. Movement speed was set at60� · s�1. The subjects were instructed to exert just enoughforce to marginally resist the movement within a range of

60–150� (180� = full extension), but not to the point wherethe machine stopped. When the subject’s arm reached finalposition (150�), the subject was instructed to relax his arm,which was returned to its starting position by the operatorwithin 2 s. The process was repeated 25 times for a totalexercise time of approximately 75 s.

Preliminary testing was done to design a protocol thataimed to induce DOMS and reduce force significantlyusing one bout of submaximal ECC (they are more com-mon in every-day activities than maximal contractions) ina short time.

2.3.3. Surface EMG and MMG

During MVC measurement, EMG and MMG signalswere recorded from probes placed on the BB and TB mus-cles. The EMG was recorded using two custom-made dif-ferential Ag/Ag Cl electrodes, 4 mm in diameter,interelectrode distance: 25 mm. An air-coupled condensermicrophone (type BCM 9765, BeStar Acoustic, Jiangsu,China; diameter 9.7 mm, weight 18 g, 54 dB sensitivitywhere 1 Pa = 7 mV) placed between the active EMG elec-trodes was used to record the MMG signal (Jaskolskaet al., 2003). The distance from the microphone to skin sur-face was 9 mm. Electrodes were placed over the belly of themuscles along their longitudinal axis and attached to theskin with double-sided adhesive tape. The ground electrodewas fixed to the right wrist. The skin was cleaned with alco-hol before electrode placement.

The EMG and MMG signals were amplified to get thebest signal-to-noise ratio. The EMG signal was tentativelyamplified by 500 using a preamplifier installed in the elec-trodes and the MMG signal was amplified by 1–20. Theamplified EMG and MMG signals were then sampled ata rate of 10 kHz, band-pass filtered 20–500 Hz and3–120 Hz, respectively, and converted in numerical formatwith a 14-bit A/D board (Analogue Devices, Norwood,MA). The same one-second period covering the greatestvalue of MVC was chosen to average and analyze theEMG and MMG signals.

Countermeasures used to avoid a cross-talk and its reli-ability were presented in previous papers (Jaskolska et al.,2003, 2006).

A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577 571

2.3.4. Muscle soreness (pain) assessment

A subjective and objective method of pain assessmentwas used. Subjective reporting pain was measured by pal-pation using ‘‘zero-one’’ scale. The subject left the armalongside the trunk in the standing position. The experi-menter pressed the belly of the BB and TB with index fin-ger. This procedure was done always by the sameexperimenter, who was trained during preliminary studyto press the muscle with the same force. The subjectreported ‘‘yes’’ if he felt pain or ‘‘no’’ if he did not feelthe pain, and this was recorded as 1 or 0, respectively.

A force level at which pain occurred was used as anobjective method of pain assessment with the same set upas for the MVC force recording, but with an instructionto develop the force very slowly and to release it as soonas the pain was sensed. Then, this force was assessed off-line and expressed as a percent of the MVC force recordedbefore ECC (baseline).

2.3.5. Data analysis

MVC and the EMG and MMG signals were averagedfrom the two trials. The EMG and MMG signals fromthe agonist (BB) and antagonist (TB) muscles were ana-lyzed in time (root mean square; RMS) and frequency(median frequency; MDF) domain. The following parame-ters were estimated:

� RMS of the EMG [lV] and MMG [mV] signals,� MDF [Hz] of the EMG and MMG,� MMG RMS/EMG RMS [mV/lV] ratio as an index of

electromechanical efficiency,� MVC force/EMG RMS [N/lV] ratio for the BB muscle

(n = 18) and TB muscle (n = 6) as an index of neuro-muscular efficiency,� The MMG RMS/MVC [mV/N] ratio for the BB muscle,

and� The MMG MDF/MVC [Hz/N] ratio for the BB muscle

as indexes of muscle stiffness changes.

Data were analyzed using SPSS version 10.0. A generallinear model for repeated-measures analyses was applied toassess differences among the two muscles and sessions(before and after ECC, and recovery days). The Wilcoxontest was used to compare results obtained in different ses-

Table 1Mean and standard deviation (±SD) of the MVC, EMG and MMG paramet

MVC (N) MVC force/EMG RMS (N/lV

Pre-ECC 163 ± 37.9 0.265 ± 0.106Immediately after 111 ± 24.5* 0.184 ± 0.072*

24 h 105 ± 23.7* 0.162 ± 0.057*

48 h 108 ± 31.6* 0.168 ± 0.070*

72 h 114 ± 31.0* 0.180 ± 0.093*

120 h 126 ± 22.7* 0.187 ± 0.075*

* P 6 0.05 differences statistically significant compared to Pre-ECC.# P 6 0.05 differences statistically significant compared to immediately after

sions as regards muscle pain. Data are presented as meanand standard deviation (SD); the level of statistical signifi-cance was set at P 6 0.05.

3. Results

3.1. Maximal isometric force

Compared to baseline, there was 32% decrease in MVC(P 6 0.05) immediately after ECC (Table 1). The forceremained at the significantly lower level until the last dayof recovery (P 6 0.05), with the lowest value at 24 h afterthe ECC (36% lower than the pre-ECC value).

There were no exercise-induced significant changes inantagonist MVC force (measured during elbow extension)immediately after ECC and during five days of recovery.Force range was 144 ± 43 N at baseline, 138 ± 33 N imme-diately after ECC (4% decrease), with the lowest value of129 ± 37 N at 72 h (10% decrease) (P > 0.05).

3.2. MVC force/EMG RMS ratio (neuromuscular

efficiency)

The MVC force/EMG RMS ratio in the BB muscledecreased significantly immediately after ECC andremained at the lower level up to 120 h post-fatigue(P 6 0.05) (Table 1). The respective ratio measured in sixsubjects during elbow extension MVC for the antagonistmuscle (TB) did not change significantly, with an oppositetendency. The biggest values were recorded immediatelyand 24 h after ECC (0.267 ± 0.124 and 0.300 ± 0.191N/lV, respectively). The pre-ECC value was 0.240 ±0.102 N/lV.

3.3. MMG RMS/MVC force and MMG MDF/MVC force

ratios of the BB

Compared to baseline, the MMG RMS calculated perunit force (MMG RMS/MVC force) did not change signif-icantly immediately after ECC (P > 0.05) (Table 1), but itdecreased during recovery compared to the value recordedimmediately post-ECC (P 6 0.05) (Table 1).

ers for the biceps brachii muscles

) MMG RMS/MVC (mV/N) MMG MDF/MVC (Hz/N)

0.148 ± 0.113 0.167 ± 0.0750.254 ± 0.168 0.221 ± 0.066*

0.153 ± 0.085# 0.271 ± 0.079*

0.129 ± 0.062# 0.256 ± 0.066*

0.134 ± 0.067# 0.249 ± 0.077*

0.136 ± 0.051# 0.206 ± 0.051*

exercise.

0

200

400

600

800

1000

1200

Pre Post 24h 48h 72h 120hTime

Time

EM

G R

MS

[μV

]

Biceps brachii Triceps brachii

0

20

40

60

80

100

120

Pre Post 24h 48h 72h 120h

EM

G M

DF

[H

z]

* * * * * * * *

Fig. 2. Values of EMG median frequency (MDF, lower panel) and RMS (upper panel) of the biceps and triceps brachii before ECC (pre), immediatelyafter (post), and 24, 48, 72, 120 h after ECC. * = P 6 0.05 differences statistically significant compared to pre.

572 A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577

Values of the MMG MDF/MVC force ratio recordedduring recovery were significantly greater than the pre-ECC value (P 6 0.05) (Table 1).

3.4. EMG RMS and median frequency

The eccentric exercise did not affect the EMG RMS ineither muscle (Fig. 2, upper panel). The ratio of theEMG RMS of the TB to BB muscle (an index of coactiva-tion) also did not change significantly after ECC: the valuesranged from 0.16 ± 0.088 (baseline) to 0.13 ± 0.054 at 72and 120 h. There was a significant (P 6 0.05) decrease inEMG median frequency in the BB immediately afterECC and during recovery (Fig. 2, lower panel). Thechanges in the EMG median frequency in the TB were sim-ilar to those observed in the BB muscle.

3.5. MMG RMS and median frequency

There were no significant changes in the MMG RMS ofthe BB and TB following the ECC. However, the differencebetween the MMG RMS recorded immediately after ECCand those at 24, 48, 72 and 120 h post-ECC was statisticallysignificant in both muscles (P 6 0.05) (Fig. 3, upper panel).The analysis revealed no statistically significant response in

MMG median frequency to ECC in either muscle (Fig. 3,lower panel).

3.6. MMG/EMG ratio (electromechanical efficiency)

Compared to the pre-ECC value, the MMG/EMGRMS ratio in the BB and TB muscles was not significantlydifferent after ECC (P > 0.05). However, there was a statis-tically significant difference between the values recordedimmediately after and those recorded during recovery inboth muscles (Fig. 4).

3.7. Pain

Sixteen subjects out of 18 reported pain measured bypalpation in the BB muscle at 24 h, 18 at 48 h, 17 at72 h, and 9 at 120 h post-ECC (P 6 0.05), while in theTB muscle there was pain in 9 subjects at 24 h and in 5at 48 h post-ECC (P 6 0.05).

Pain measured during force development occurred in allsubjects during 5 days of recovery. Compared to MVCforce at pre-ECC, the level of force at which pain wassensed was 37 ± 36% at 24 h, 32 ± 22% at 48 h, 40 ± 38%at 72 h, and 76 ± 38% at 120 h (P 6 0.05).

0

100

200

300

400

500

600

Pre Post 24h 48h 72h 120h

Time

V]

Biceps brachii Triceps brachii

# # # # # # # #

Fig. 4. Ratio of MMG RMS/EMG RMS of the biceps brachii and triceps brachii before eccentric exercise (pre), immediately after (post), and 24, 48, 72,120 h after ECC. # = P 6 0.05 differences statistically significant compared to IA.

0

10

20

30

40

50

60

Pre Post 24h 48h 72h 120hTime

Time

MM

G R

MS

[μV

]

Biceps brachii Triceps brachii

# # # # # # # #

0

5

10

15

20

25

30

35

Pre Post 24h 48h 72h 120h

MM

G M

DF

[H

z]

Fig. 3. Values of MMG median frequency (MDF, lower panel) and RMS (upper panel) of the biceps and triceps brachii before ECC (pre), immediatelyafter (post), and 24, 48, 72, 120 h after ECC. # = P 6 0.05 differences statistically significant compared to IA.

A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577 573

4. Discussion

A new finding of this study was that the ECC-inducedMVC force drop was accompanied by similar changes inelectrical as well as mechanical properties in the BB andTB. Moreover, in both muscles the changes in the EMGparameters were different from those in the MMG param-

eters immediately after submaximal ECC and during 5days of recovery.

4.1. Maximal isometric force

The present results showed that there was a significantdecrease in MVC force of elbow flexors (but not extensors

574 A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577

– the antagonists) after 25 submaximal ECC and that thesubjects still showed muscle weakness up to 5 days(120 h) after exercise; this is in agreement with the findingsof others (Clarkson et al., 1992; Kroon and Naeije, 1991;Michaut et al., 2001; Prasartwuth et al., 2005). A numberof studies have indicated that FT fibers are more suscepti-ble to damage than slow-twitch (ST) fibers during ECC(Friden et al., 1983; Byrne et al., 2004) and that it is thisfactor that contributes to force loss seen after ECC. Oneof the well-documented symptoms of muscle damage afterECC is DOMS (Newham et al., 1983; Clarkson et al.,1992). Therefore, pain recorded during five days of recov-ery in the BB in the present study can potentially affectmuscle activation by inhibition of the agonistic muscle(Graven-Nielsen et al., 1997; Birch et al., 2000) and/or exci-tation of the antagonist muscle, thereby reducing force pro-duction (Lund et al., 1991).

4.2. Eccentric exercise effect on EMG activity of the pair of

antagonist muscles

The present study showed no significant changes in theEMG RMS at MVC in the BB and TB immediately afterECC and during 5 days of recovery; this is in agreementwith the findings of other studies (Pearce et al., 1998;Michaut et al., 2001). The lack of ECC-related changesin antagonist coactivation, that could potentially contrib-ute to the agonist net isometric torque decrement, is inagreement with the results of Michaut et al. (2001). How-ever, neuromuscular efficiency (as assessed by the MVCtorque/EMG RMS ratio) did not change throughout theexperimental period in the studies by Michaut et al.(2001) and McHugh et al. (2000), while it decreased signif-icantly for 10 days following exercise in the study byDeschenes et al. (2000). Similar results were recorded forthe BB muscle in the present study. In the TB muscle, thechanges in neuromuscular efficiency were not statisticallysignificant, with a strong tendency to increase immediatelyand 24 h after ECC, but this finding needs to be interpretedwith caution because of a small number (6) of subjectstested. Neuromuscular efficiency is a measure of the capa-bility of the contractile elements to answer the neural sig-nals. Thus, it provides information on whether thereduction in force is due to neural or contractile factors.Our results support Sayers’ hypothesis (Sayers et al.,2003) that the decline in MVC force after ECC is due tocontractile impairments in agonists rather than neural fac-tors. This could include electro-mechanical coupling failure(reduced rate of calcium release from the sarcoplasmicreticulum) (Byrne et al., 2004).

Frequency domain analysis revealed a shift of the EMGmedian frequency in the BB and TB muscles towards alower value after ECC. Our data are partially in agreementwith those of Day et al. (1998), who showed that meanpower frequency decreased at day 1, 2, 3, and then returnedto the pre-ECC value. Felici et al. (1997) showed that med-ian frequency of the surface EMG decreased after ECC at

80% and 50% of MVC force. Our study extends those find-ings to 100% MVC. However, different results have beenreported. Barry et al. (1990) found that EMG amplitudeand median frequency did not change after ECC in thequadriceps muscles, while McHugh et al. (2000) showedan increase in median frequency during MVC immediatelyafter eccentric isokinetic exercise which returned to base-line on days 1, 2, 3 after exercise. The discrepancy mightbe due to different muscles and subjects tested (males,females) (Byrne et al., 2004), and/or different fatigue con-traction intensities (Nosaka and Newton, 2002). For exam-ple, the intensity of contraction used by McHugh et al.(2000) was supposedly too low because they recorded anMVC increase (not decrease) immediately after ECC(105%) which then declined 1 day after (89%) and returnedto the baseline by the second day.

The drop in the EMG median frequency reported in thepresent study may result from an exercise-related decreasein muscle fiber conduction velocity (Eberstein and Beattie,1985) and/or an increase in motor unit synchronization(Krogh-Lund and Jorgensen, 1993; Bigland-Ritchie et al.,1986; De Luca, 1984). However, it has also been shownthat the decrease in conduction velocity cannot solelyexplain the changes in the mean power frequency (MPF)as they are not always synchronous (Linssen et al., 1993).Bigland-Ritchie et al. (1983) and Enoka and Stuart(1992) suggested that sensory afferent nerves originatingwithin the fatigued muscle supply a signal to the motoneu-ron pool to cause a lower discharge rate of the motor units.

The present finding of the lower EMG median frequencyafter ECC indicates that fast fatigable motor units are prob-ably mostly involved and injured in ECC at 50% MVC; thisagrees with findings of other studies (Nardone and Schiep-pati, 1988; Enoka, 1996; Friden and Lieber, 2001; Byrneet al., 2004). Most likely the injury of the FT fiber is compen-sated for by an increased activity of the ST fibers trying tomaintain the force output, resulting in a shift in the medianfrequency toward lower values. Pain recorded during recov-ery after ECC could cause changes in voluntary activation ofmotor units (shift from fast to slow-twitch), but probably notin muscle contractile properties since Farina et al. (2004) didnot record changes in conduction velocity with the injectionof hypertonic saline which caused pain. An altered tissueconductance related to swelling and edema of damaged mus-cle could also contribute to EMG frequency changes, sincean increase (not statistically significant) in exercised arm cir-cumference was noted during recovery in our study (data notpresented). The exercise-related EMG frequency decreasecan also be associated to an impairment of the excitation-contraction coupling (Friden and Lieber, 2001; Byrneet al., 2004).

4.3. Eccentric exercise effect on MMG activity of antagonist

muscles

The present study showed significant MMG RMS dim-inution at 24, 48, 72 and 120 h in comparison to the value

A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577 575

recorded immediately post-ECC, but not to baseline (pre-ECC). This is because the value immediately post-ECCwas somewhat bigger than that at the baseline. Thisincrease in MMG RMS and MMG RMS expressed perunit of force and per unit of EMG RMS probably resultsfrom motor unit synchronization related to fatigue tremorthat was too small to change them significantly, but bigenough to affect them. The decrease in median frequencyof the EMG (P 6 0.05) and MMG (P > 0.05) signalsimmediately after ECC confirms this notion.

Explanation of the MMG RMS diminution during therecovery period most likely is related to changes in motorunit activation strategy as a result of ECC. The preferentialrecruitment of FT muscle fibers during ECC (Nardone andSchieppati, 1988; Enoka, 1996; Byrne et al., 2004) and theirdamage (Friden et al., 1983; Friden and Lieber, 2001;Byrne et al., 2004) result in a greater activation of slow-twitch fibers. The MMG amplitude of ST fibers is smallerthan that of FT fibers (Bichler, 2000) because they are char-acterized by a smaller diameter (Stein and Padykula, 1962).Moreover, ST fibers are located deeper in a muscle (Mantaet al., 1996; Bodine-Fowler et al., 1990) and can be dampedby the nearby tissues (Orizio et al., 1989).

To some extent, the post-ECC lowering of the MMGRMS during recovery could result from increased musclestiffness due to a higher intramuscular pressure (Fridenet al., 1988) and edema related to DOMS; indeed therewas a non-significant enlargement of the exercised arm cir-cumference. The greater muscle stiffness may restrict themuscle fibers’ capability to oscillate (Orizio, 1993), thusdecreasing the MMG amplitude. This hypothesis is con-firmed by a larger MMG median frequency per unit offorce (MMG MDF/MVC) during recovery (than at base-line) found in the present study for the BB muscle. In addi-tion, the increased muscle stiffness may induce an increasein the motor unit firing rate (Vaz et al., 1997), which in turncan result in an MMG frequency shift to larger values. Ifso, such an effect should result in EMG median frequencyincrease. Compared to immediately post-ECC there was anincreasing tendency in EMG median frequency at 24, 48and 72 h post-ECC but the values were still significantlylower that the pre-ECC value.

The fact that the MMG RMS increased immediatelyafter ECC was not significant in the present study, whilethe significant increase in the first dorsal interosseousobserved by Bajaj et al. (2002) can be related to a differencebetween testing protocols. In our study, 25 ECC at 50%MVC force was used while six bouts of 110% force ofMVC were applied in the study of Bajaj et al. (2002).High-intensity ECC (110% MVC) can be associated witha bigger physiologic tremor, thereby increasing MMGamplitude.

Madeleine and Arendt-Nielsen (2005) found that theMMG RMS increased after experimental pain inducedon the BB muscle during submaximal isometric contrac-tions. In the present study, there was the non-significantMMG RMS change in the BB and TB as an acute effect

of the ECC and the MMG RMS even diminished duringrecovery. It must be noted that, except for ECC relatedmuscle pain, there are other factors that could contributeto the decrease in MMG amplitude at MVC after ECC,e.g., raised intra-muscular pressure and muscle stiffness,which play bigger roles during maximal than submaximalisometric contractions. Even if the increased intra-muscu-lar pressure itself does not significantly affect the MMGamplitude (Sogaard et al., 2006), increased muscle stiffnessdoes. In addition, the changes in motor control mecha-nisms relayed on prevalence of ST-fiber recruitment duringthe recovery period (as a result of FT-fiber injury duringeccentric fatigue) will affect MMG amplitude with a maxi-mal isometric contraction more than at a submaximal levelbecause more FT fibers are involved during an MVC. Aswas already discussed, the MMG amplitude of ST fibersis smaller than that of the FT fibers (Bichler, 2000) becauseof their diameter (Stein and Padykula, 1962) and locationin a muscle (Manta et al., 1996; Bodine-Fowler et al.,1990).

4.4. Agonist–antagonist co-contraction

One of the main problems of motor control which couldbe related to muscle fatigue is the individual role of agonistand antagonist muscles while performing voluntary con-tractions and during recovery after exhausting exercise.The study of Mullany et al. (2002) reported the trend ofparallel agonist and antagonist surface EMG behavior,supporting the evidence for the existence of a commondrive controlling the entire agonist motoneuron pool andsubpopulation of the antagonist motoneuron pool (Jongenet al., 1989). The simultaneous activation of the antagonistmuscle, defined as co-contraction (also referred to as coac-tivation), could reduce the measured force. However, sim-ilar to what Michaut et al. (2001) reported, we did notrecord significant changes in coactivation level during anMVC after maximal ECC. An interesting finding fromthe present study is that not only EMG, but also MMGchanges in the BB (agonist) and TB (antagonist) muscleare similar immediately after ECC and during recovery,supporting evidence for the existence of a common drive.In our case, it seems that coactivation served to protectand stabilize the joint during forceful contractions(Solomonow et al., 1987) but it did not contribute to thedecrease in force.

5. Conclusions

The present research showed that (i) there were similarchanges in electrical and mechanical activities duringMVC after submaximal ECC in agonist and antagonistmuscles, suggesting a common drive controlling the agonistand antagonist motoneuron pool; (ii) ECC induced differ-ent changes in EMG than in MMG immediately afterECC and during 120 h of recovery that suggested anincreased tremor and contractile impairments, i.e., reduced

576 A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577

rate of calcium release from the sarcoplasmic reticulum(acute effect), and changes in motor control mechanismsof agonist and antagonist muscles, and increased musclestiffness (chronic effect).

These results (in comparison with others) show thatsimultaneous recording and analysis of EMG and MMGsignals allow a more precise assessment of the ECC-induced changes compared to the analogue analysis donewith one signal only.

References

Bajaj P, Madeleine P, Sjøgaard G, Arendt-Nielsen L. Assessment ofpostexercise muscle soreness by electromyography and mechanomy-ography. J Pain 2002;3:126–36.

Barry DT, Gordon DE, Hinton GG. Acoustic and surface EMGdiagnosis of pediatric muscle disease. Muscle Nerve 1990;13:286–90.

Bichler E. Mechanomyograms recorded during evoked contractions ofsingle motor units in the rat medial gastrocnemius muscle. Eur J ApplPhysiol 2000;83:310–9.

Bigland-Ritchie BR, Johansson R, Lippold OCJ, Woods JJ. Contractilespeed and EMG changes during fatigue of sustained maximalvoluntary contractions. J Neurophysiol 1983;50:313–24.

Bigland-Ritchie BR, Furbush F, Woods JJ. Fatigue of intermittentsubmaximal voluntary contractions: central and peripheral factors. JAppl Physiol 1986;61:421–9.

Birch L, Grave-Nielsen T, Christensen H, Arendt-Nielsen L. Experi-mental muscle pain modulates muscle activity and work performancedifferently during high and low precision use of a computer mouse. EurJ Appl Physiol 2000;83:492–8.

Bodine-Fowler S, Garfinkel A, Roy RR, Edgerton VR. Spatial distri-bution of muscle fibers within the territory of a motor unit. MuscleNerve 1990;13:1133–45.

Byrne Ch, Twist C, Eston R. Neuromuscular function after exercise-induced muscle damage. Sport Med 2004;34(1):49–69.

Clarkson PM, Byrnes WC, McCormick KM, Turcotte LP, White JS.Muscle soreness and creatine kinase activity following isometric,eccentric, and concentric exercise. Int J Sport Med 1986;7:152–5.

Clarkson PM, Nosaka K, Braun B. Muscle function after exercise-inducedmuscle damage and rapid adaptation. Med Sci Sport Exer1992;24:512–20.

Day SH, Donnelly AE, Brown SJ, Child RB. Electromyogram activityand mean power frequency in exercise-damaged human muscle.Muscle Nerve 1998;21:961–3.

De Luca CJ. Myoelectrical manifestations of localized muscular fatigue inhumans. Crit Rev Biomed Eng 1984;11(4):251–79.

Deschenes MR, Brewer RE, Bush JA, McCoy RW, Volek JS, KraemerWJ. Neuromuscular disturbance outlasts other symptoms of exercise-induced muscle damage. J Neurol Sci 2000;174:92–9.

Eberstein A, Beattie B. Simultaneous measurement of muscle conductionvelocity and EMG power spectrum changes during fatigue. MuscleNerve 1985;8:768–73.

Enoka RM. Eccentric contractions require unique activation strategies bythe nervous system. J Appl Physiol 1996;81(6):2339–46.

Enoka RM, Stuart DG. Neurobiology of muscle fatigue. J Appl Physiol1992;72:1631–48.

Falconer K, Winter DA. Quantitative assessment of co-contraction at theankle joint in walking. Electromyo Clin Neurophysiol 1985;25:135–49.

Farina D, Arendt-Nielsen L, Merletti R, Graven-Nielsen T. Effect ofexperimental muscle pain on motor unit firing rate and conductionvelocity. J Neurophysiol 2004;91:1250–9.

Felici F, Colace L, Sbriccoli P. Surface EMG modifications after eccentricexercise. J Electromyo Kinesiol 1997;7:193–202.

Friden J, Lieber RL. Eccentric exercise-induced injures to contractile andcytoskeletal muscle fibre components. Acta Physiologica Scandinavica2001;171:321–6.

Friden J, Sjostrom M, Ekblom B. Myofibrillar damage following intenseeccentric exercise in man. Int J Sport Med 1983;4:170–6.

Friden J, Sfakianos PN, Hargens AR. Residual muscular swelling afterrepetitive eccentric contractions. J Ortho Res 1988;6:493–8.

Graven-Nielsen T, Svensson P, Arendt-Nielsen L. Effects of experimentalmuscle pain on muscle activity and co-ordination during static anddynamic motor function. Electroen Clin Neurophysiol1997;105:156–64.

Heinonen A, Sievanen H, Viitasalo J, Pasanen M, Oja P, Vuori I.Reproducibility of computer measurement of maximal isometricstrength and electromyography in sedentary middle-aged women.Eur J Appl Physiol 1994;68:310–4.

Hortobagyi T, Houmard J, Fraser D. Normal force and myofibrillardisruption after repeated eccentric exercise. J Appl Physiol1998;84:492–8.

Howell JN, Chleboun G, Conatser R. Muscle stiffness, strength loss,swelling and soreness following exercise-induced injury in humans. JPhysiol 1993;464:183–9.

Jaskolska A, Kisiel K, Brzenczek W, Jaskolski A. EMG and MMG ofsynergists and antagonists during relaxation at three joint angles. Eur JAppl Physiol 2003;90:58–69.

Jaskolska A, Kisiel-Sajewicz K, Brzenczek-Owczarzak W, Yue GH,Jaskolski A. EMG and MMG of agonist and antagonist muscles as afunction of age and joint angle. J Electromyo Kinesiol 2006;16:89–102.

Jongen HAH, Dermier van der Gon D, Gielen CCAM. Inhomogeneousactivation of motoneuron pools as revealed by co-contraction ofantagonistic human arm muscles. Exp Brain Res 1989;75:555–62.

Komi PV, Viitasalo JT. Changes in motor unit activity and metabolism inhuman skeletal muscle during and after repeated eccentric andconcentric contractions. Acta Physiologica Scandinavica1977;100:246–54.

Krogh- C, Jorgensen K. Myo-electric fatigue manifestations revisited:power spectrum, conduction velocity, and amplitude of human elbowflexor muscles during isolated and repetitive endurance contractions at30% maximal voluntary contraction. Eur J Appl Physiol1993;66:161–73.

Kroon GW, Naeije M. Recovery of the human biceps electromyogramafter heavy eccentric, concentric or isometric exercise. Eur J ApplPhysiol 1991;63:444–8.

Linssen WHPJ, Stegeman DF, Joosten EMG, van’t Hof MA, BinkhorstRA, Notermans LSH. Variability and interrelationships of surfaceEMG parameters during local muscle fatigue. Muscle Nerve1993;16(8):849–56.

Lund JP, Donga D, Widmer CG, Stohler CS. The pain-adaptation model:a discussion of the relationship between chronic musculosceletal painand motor activity. Can J Physiol Pharmacol 1991;69:683–94.

Madeleine P, Arendt-Nielsen L. Experimental muscle pain increasesmechanomyographic signal activity during sub-maximal isometriccontractions. J Electromyo Kinesiol 2005;15:27–36.

Mamaghani NK, Shimomura Y, Iwanaga K, Katsuura T. Mechanomy-ogram and electromyogram response of upper limb during sustainedisometric fatigue with varying shoulder and elbow posture. J PhysiolAnthropol Appl Human Sci 2002;21(1):29–43.

Manta P, Kalfakis N, Kararizou E, Vassilopoulos D, Papageorgiou C.Size and proportion of fiber types in human muscle fascicles. ClinNeuropathol 1996;15:116–8.

McHugh MP, Connolly DAJ, Eston RG, Gleim GW. Electromyographicanalysis of exercise resulting in symptoms of muscle damage. J SportSci 2000;18:163–72.

Michaut A, Pousson M, Ballay Y, Van Hoecke J. Short-term changes inthe series elastic component after an acute eccentric exercise of theelbow flexors. Eur J Appl Physiol 2001;84:569–74.

Mullany H, O’Malley M, St. Clair Gibson A, Vaughan C. Agonist–antagonist common drive during fatiguing knee extension efforts usingsurface electromyography. J Electromyo Kinesiol 2002;12:375–84.

Nardone A, Schieppati M. Shift of activity from slow to fast motor unitsduring voluntary lengthening contractions of the triceps surae musclesin humans. J Physiol 1988;395:363–81.

A. Jaskolski et al. / Journal of Electromyography and Kinesiology 17 (2007) 568–577 577

Newham DJ, McPhail G, Mills KR, Edwards RH. Ultrastructuralchanges after concentric and eccentric contractions of human muscle.J Neurol Sci 1983;61:109–22.

Nosaka K, Newton M. Difference in the magnitude of muscle damagebetween maximal and submaximal eccentric loading. J Strength CondRes 2002;16(2):202–8.

Orizio C. Muscle sound: bases for the introduction of mechanomyo-graphic signal in muscle studies. Crit Rev Biomed Eng1993;21:201–43.

Orizio C. Surface mechanomyograam. In: Merletti R, Parker PA, editors.Electromyography, physiology, engineering and non-invasive applica-tions. New Jersey: IEEE Press, Wiley–Interscience, Wiley & Sons;2004. p. 305–22.

Orizio C, Perini R, Veicsteinas A. A muscular sound and forcerelationship during isometric contraction in man. Eur J Appl Physiol1989;58:528–33.

Pearce AJ, Sacco P, Byrnes ML, Thickbroom GW, Mastaglia FL. Theeffect of eccentric exercise on neuromuscular function of the bicepsbrachii. J Sci Med Sport 1998;1(4):236–44.

Prasartwuth O, Taylor JL, Gandevia SC. Maximal force, voluntaryactivation and muscle soreness after eccentric damage to human elbowflexor muscles. J Physiol 2005;567(1):337–48.

Sayers SP, Knight CA, Clarkson PM. Neuromuscular variables affectingthe magnitude of force loss after eccentric exercise. J Sport Sci2003;21:403–10.

Sbriccoli P, Felici F, Rosponi A, Aliotta A, Castellano V, Mazza C, et al..Exercise induced muscle damage and recovery assessed by means oflinear and non-linear sEMG analysis and ultrasonography. J Electro-myo Kinesiol 2001;11:73–83.

Sogaard K, Orizio C, Sjogaard G. Surface mechanomyogram amplitude isnot attenuated by intramuscular pressure. Eur J Appl Physiol2006;96:178–84.

Solomonow M, Baratta R, Zhou B, Shoji H, Bose W, Beck C, et al.. Thesynergistic action of the ACL and thigh muscles in maintaining jointstability. Am J Sport Med 1987;15:207–13.

Stein JM, Padykula HA. Histochemical classification of individual skeletalmuscle fibers of the rat. Am J Anat 1962;110:103–23.

Vaz MA, Herzog W, Zhang YT, Leonard TR, Nguyen H. Theeffect of muscle length on electrically elicited muscle vibrationsin the in situ cat soleus muscle. J Electromyo Kinesiol1997;7:113–21.

Warren GL, Lowe DA, Armstrong RB. Measurement tools used in thestudy of eccentric contraction-induced injury. Sport Med (Rev)1999;27:43–59.

Artur Jaskolski, received the M.Sc. from theUniversity School of Physical Education inKatowice, Poland and the Ph.D. degree fromUniversity School of Physical Education in Poz-nan, Poland with exercise physiology specializa-tion. In 1992 he received his Dr. Sci. in exercisephysiology from University School of PhysicalEducation in Krakow (Poland). During his careerhe worked as a Senior Researcher at ArizonaState University Tempe, Arizona (1992 - 1995),and as a Visiting Professor at Indiana University

Purdue University of Indiana, Indianapolis (1996 – 1997), working for theHERITAGE Family Study grant from the National Institute of Health,USA. In 1997 he was appointed Associate Professor and became Head of

the Department of Exercise Physiology at Faculty of Physical Educationin Gorzow Wlkp., Poland. Since 1999 he is Head of the Department ofKinesiology at University School of Physical Education in Wrocław,Poland and in 2003 was appointed Full Professor. He has publishedapprox. 100 scientific papers within exercise physiology and motorcontrol.

Renata Andrzejewska was born in 1975 in Poland.She received her M.Sc. degree in physiotherapyfrom the Faculty of Physiotherapy, UniversitySchool of Physical Education in Wrocław, Polandin 2000. From 2000 to 2005 she was a doctoralstudent in the Department of Kinesiology atUniversity School of Physical Education inWrocław, Poland. On the same University in 2005she received her Ph.D. degree in exercise physi-ology. Her main areas of research is a non-inva-sive assessment of muscle damage and repair

using surface electromyography and mechanomyography. Her Ph.D.thesis and the 3 papers she published relate to DOMS, eccentric exercise,and electrical and mechanical muscle activities.

Jarosław Marusiak received his M.Sc. degree inphysiotherapy from the Faculty of Physiotherapy,University School of Physical Education inWroclaw, Poland in 2002. Since october 2002, hehas worked as an Assistant and he is a doctoralcandidate in kinesilogy at Department of Kine-siology, Faculty of Physiotherapy, UniversitySchool of Physical Education in Wroclaw. In2004 he was on five-months sabbatical leave inLerner Research Institute at Cleveland Clinic,Ohio, USA. His research interests are related to

the field of Kinesiology: muscle and physiology and application of elec-tromyography and mechanomyography in diagnosis of neuromusculardisorders.

Katarzyna Kisiel-Sajewicz received her M.Sc. in1998 from the Institute of Physical Education inGorzow Wlkp., Poland. From 1998 to 2000 sheworked at the same Institute as an assistant andjunior researcher at Biomechanics Section of thePhysiology Department. From 2000 she works atthe Department of Kinesiology, Section Kinesi-ology University School of Physical Education inWrocław, Poland. In 2003 she got her Ph.D.degree from University School of Physical Edu-cation in Wrocław, Poland. Presently, she is an

Assistant Professor at the same Department. Her research interest focuseson motor control and kinesiology. She has published 12 scientific papers.

Anna Jaskolska received the M.Sc. and Ph.D.degree from the University School of PhysicalEducation in Warszawa (Poland) with an anat-omy and biomechanics specialization. In 1999 shereceived her Dr.Sci. in exercise physiology andkinesiology from the University School of Physi-cal Education in Krakow (Poland). From 1992 to1995 she was a Senior Researcher at ArizonaState University Tempe, USA, and from 1996 to1997 she had Visiting Professor Position at Indi-ana University Purdue University of Indiana,

Indianapolis, USA working for the HERITAGE Family Study grant fromthe National Institute of Health, USA. From 1997 to 1999 she had anAssistant Professor position at the Department of Exercise Physiology,

Faculty of Physical Education in Gorzow Wlkp., Poland. Since 2004 shehas a Full Professor position and she is a Head of the Kinesiology Sectionat the Department of Kinesiology, University School of Physical Educa-tion in Wrocław, Poland. She has published approx. 80 scientific paperswithin kinesiology, exercise physiology and motor control.