7
Original article The neuromotor effects of transverse friction massage Haris Begovic a, * , Guang-Quan Zhou a , Snje zana Schuster b , Yong-Ping Zheng a a The Hong Kong Polytechnic University, Interdisciplinary Division of Biomedical Engineering, Hung Hom, Kowloon, Hong Kong, SAR 999077, China b University of Applied Health Science, Mlinarska Street 38, HR-10000, Zagreb, Croatia article info Article history: Received 20 February 2016 Received in revised form 14 May 2016 Accepted 12 July 2016 Keywords: Transverse friction massage Electromechanical delay Excitation-contraction coupling Ultrafast ultrasonography abstract Background: Transverse friction massage (TFM), as an often used technique by therapists, is known for its effect in reducing the pain and loosing the scar tissues. Nevertheless, its effects on neuromotor driving mechanism including the electromechanical delay (EMD), force transmission and excitation-contraction (EC) coupling which could be used as markers of stiffness changes, has not been computed using ultrafast ultrasound (US) when combined with external sensors. Aim: Hence, the aim of this study was to nd out produced neuromotor changes associated to stiffness when TFM was applied over Quadriceps femoris (QF) tendon in healthy subjcets. Methods: Fourteen healthy males and fteen age-gender matched controls were recruited. Surface EMG (sEMG), ultrafast US and Force sensors were synchronized and signals were analyzed to depict the time delays corresponding to EC coupling, force transmission, EMD, torque and rate of force development (RFD). Results: TFM has been found to increase the time corresponding to EC coupling and EMD, whilst, reducing the time belonging to force transmission during the voluntary muscle contractions. Conclusions: A detection of the increased time of EC coupling from muscle itself would suggest that TFM applied over the tendon shows an inuence on changing the neuro-motor driving mechanism possibly via afferent pathways and therefore decreasing the active muscle stiffness. On the other hand, detection of decreased time belonging to force transmission during voluntary contraction would suggest that TFM increases the stiffness of tendon, caused by faster force transmission along non-contractile elements. Torque and RFD have not been inuenced by TFM. © 2016 Elsevier Ltd. All rights reserved. 1. Introduction Transverse friction massage, described by James Cyriax in the 1940 (Kesson and Atkins, 1998), has often been used in chronic inammatory conditions such as lateral epicondylitis, iliotbial band friction syndrome or patellar tendinitis. TFM promotes local hy- peremia, analgesia and the reduction of adherent scar tissue to ligaments, tendons, and muscles (Yoo et al., 2012). It makes scar tissue more mobile in sub-acute and chronic inammatory condi- tions by realigning the normal soft tissue bers (Brosseau et al., 2002). Therefore, it is expected that TFM reduces the stiffness (extensibility increases) of the soft tissue. Given its often clinical effectiveness, TFM has not been scrutinized enough in order to nd out its effect on the neuromotor excitability or neuromotor driving mechanism which has a neural control over stiffness changes (Bennett et al., 2014). It was shown that TFM reduces the excitability of the moto- neuron pool when tested via H-Reex, carried out as an electro- myographic (EMG) response to a mild electrical shock to the nerve (Lee et al., 2009). The petrissage massage, performed as a rhythmic grasping and releasing of the muscle tissue reduces motor-neuron excitability as well via muscle spindles and golgi tendon organs, likely to be a centrally mediated inhibition from higher motor cen- ters (Sullivan et al.,1991). Basically, a number of studies have agreed on this reduction of neuromotor excitability (Lee et al., 2009; Roberts, 2011; Goldberg et al., 1992; Newham and Lederman, 1997; Kassolik et al., 2009) and muscle stiffness when massage was applied over the muscle (Eriksson Crommert et al., 2015). Albeit a limited number of studies investigated the effect of massage Abbreviations: EC, excitation-contraction; EMD, electromechanical delay; EMG, electromyography; MTJ, myotendinous junction; QF, quadriceps femoris; RF, rectus Femoris; RFD, rate of force development; RMS, root mean square; ROI, region of interest; SD, standard deviations; TFM, transverse friction massage; US, ultrasound. * Corresponding author. E-mail addresses: [email protected] (H. Begovic), 09903101r@connect. polyu.hk (G.-Q. Zhou), [email protected] (S. Schuster), yongping. [email protected] (Y.-P. Zheng). Contents lists available at ScienceDirect Manual Therapy journal homepage: www.elsevier.com/math http://dx.doi.org/10.1016/j.math.2016.07.007 1356-689X/© 2016 Elsevier Ltd. All rights reserved. Manual Therapy 26 (2016) 70e76

Begovic et al.2016

Embed Size (px)

Citation preview

Page 1: Begovic et al.2016

lable at ScienceDirect

Manual Therapy 26 (2016) 70e76

Contents lists avai

Manual Therapy

journal homepage: www.elsevier .com/math

Original article

The neuromotor effects of transverse friction massage

Haris Begovic a, *, Guang-Quan Zhou a, Snje�zana Schuster b, Yong-Ping Zheng a

a The Hong Kong Polytechnic University, Interdisciplinary Division of Biomedical Engineering, Hung Hom, Kowloon, Hong Kong, SAR 999077, Chinab University of Applied Health Science, Mlinarska Street 38, HR-10000, Zagreb, Croatia

a r t i c l e i n f o

Article history:Received 20 February 2016Received in revised form14 May 2016Accepted 12 July 2016

Keywords:Transverse friction massageElectromechanical delayExcitation-contraction couplingUltrafast ultrasonography

Abbreviations: EC, excitation-contraction; EMD, elelectromyography; MTJ, myotendinous junction; QF, qFemoris; RFD, rate of force development; RMS, rootinterest; SD, standard deviations; TFM, transverse fric* Corresponding author.

E-mail addresses: [email protected] (H. Bpolyu.hk (G.-Q. Zhou), [email protected]@polyu.edu.hk (Y.-P. Zheng).

http://dx.doi.org/10.1016/j.math.2016.07.0071356-689X/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

Background: Transverse friction massage (TFM), as an often used technique by therapists, is known for itseffect in reducing the pain and loosing the scar tissues. Nevertheless, its effects on neuromotor drivingmechanism including the electromechanical delay (EMD), force transmission and excitation-contraction(EC) coupling which could be used as markers of stiffness changes, has not been computed using ultrafastultrasound (US) when combined with external sensors.Aim: Hence, the aim of this study was to find out produced neuromotor changes associated to stiffnesswhen TFM was applied over Quadriceps femoris (QF) tendon in healthy subjcets.Methods: Fourteen healthy males and fifteen age-gender matched controls were recruited. Surface EMG(sEMG), ultrafast US and Force sensors were synchronized and signals were analyzed to depict the timedelays corresponding to EC coupling, force transmission, EMD, torque and rate of force development(RFD).Results: TFM has been found to increase the time corresponding to EC coupling and EMD, whilst,reducing the time belonging to force transmission during the voluntary muscle contractions.Conclusions: A detection of the increased time of EC coupling from muscle itself would suggest that TFMapplied over the tendon shows an influence on changing the neuro-motor driving mechanism possiblyvia afferent pathways and therefore decreasing the active muscle stiffness. On the other hand, detectionof decreased time belonging to force transmission during voluntary contraction would suggest that TFMincreases the stiffness of tendon, caused by faster force transmission along non-contractile elements.Torque and RFD have not been influenced by TFM.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Transverse friction massage, described by James Cyriax in the1940 (Kesson and Atkins, 1998), has often been used in chronicinflammatory conditions such as lateral epicondylitis, iliotbial bandfriction syndrome or patellar tendinitis. TFM promotes local hy-peremia, analgesia and the reduction of adherent scar tissue toligaments, tendons, and muscles (Yoo et al., 2012). It makes scartissue more mobile in sub-acute and chronic inflammatory condi-tions by realigning the normal soft tissue fibers (Brosseau et al.,

ectromechanical delay; EMG,uadriceps femoris; RF, rectusmean square; ROI, region oftion massage; US, ultrasound.

egovic), [email protected] (S. Schuster), yongping.

2002). Therefore, it is expected that TFM reduces the stiffness(extensibility increases) of the soft tissue. Given its often clinicaleffectiveness, TFM has not been scrutinized enough in order to findout its effect on the neuromotor excitability or neuromotor drivingmechanism which has a neural control over stiffness changes(Bennett et al., 2014).

It was shown that TFM reduces the excitability of the moto-neuron pool when tested via H-Reflex, carried out as an electro-myographic (EMG) response to a mild electrical shock to the nerve(Lee et al., 2009). The petrissage massage, performed as a rhythmicgrasping and releasing of the muscle tissue reduces motor-neuronexcitability as well via muscle spindles and golgi tendon organs,likely to be a centrally mediated inhibition from higher motor cen-ters (Sullivan et al.,1991). Basically, a number of studies have agreedon this reduction of neuromotor excitability (Lee et al., 2009;Roberts, 2011; Goldberg et al., 1992; Newham and Lederman,1997; Kassolik et al., 2009) and muscle stiffness when massagewas applied over themuscle (Eriksson Crommert et al., 2015). Albeita limited number of studies investigated the effect of massage

Page 2: Begovic et al.2016

H. Begovic et al. / Manual Therapy 26 (2016) 70e76 71

application over the tendon, a decrease in H-Reflex was observed,showing a reduction of neuromotor excitability in normal (Kukulkaet al., 1986) and hemiparetic patients (Leone& Kukulka,1988), afterapplied pressure on the Achilles tendon. This likely to be a centrallymediated inhibition has also been suggested to be a possible causefor the observed force reduction after a series of massage applica-tions to the iliotibial band (Hunter et al., 2006). A decline in thepower reduction was also reported after the massage of thegastrocnemius muscle (Shin & Sung, 2015). On the other hand, acomprehensive study has shown that massage reduces neuromotorexcitability without affecting twitch contractile properties wheninterpreted by analyzing the peak torque and derived parameters(Behm et al., 2013). Therefore, to get some information about thechanges of the contractile properties of the muscle after massage,the combined approach of using surface sensors became a plausiblemethod depicting EC coupling of active contractile properties ofmuscle and force transmission of parallel elastic components ofmuscle, both contributors to the overall EMD (Esposito et al., 2011;Esposito et al., 2009). In the mentioned approach, combined sen-sors such as EMG,MMG and Strain gaugewere timely synchronizedwhere EMG provides important information of motor unit neuralactivation during muscle contraction, MMG provides informationabout dimensional changes of the transverse diameter of themusclefibers and Strain gauge provides information about force outputduring amuscle contraction. The time lag between the onset of EMGandMMG signal generation during contractionwas attributed to ECcoupling while the time lag between MMG and Strain gauge signalgeneration was attributed to force transmission along parallelelastic components. In this approach, anMMGsensorwas used as anexternal sensor placed over the muscle belly and a signal displayedwas the summation of the surface oscillations including muscle,sub-cutaneous tissue and the skin itself. Very recently this methodhas been renewed where instead of an MMG sensor; an ultrafastultrasound was used depicting muscle oscillations from a certaindepth of the muscle tissue itself, therefore, excluding tissues abovethe muscle and possible artifacts produced thereby (Begovic et al.,2014). In this method, combined and timely synchronized EMG,Ultrafast ultrasound and Strain gauge prompted an investigationwhich would unveil potential changes of the contractile propertiesand force transmission after TFM applied over the quadriceps fem-oris tendon. These changes are displayed in terms of time delaysbetween generations of each signal where time delay between EMGandUS (muscle disturbance depicted from inside) onset provides anindication of EC coupling duration, while a time between US andStrain Gauge signal onset provides information about force trans-mission and stiffness of series elastic components. Hence the aim ofour study was to find out what effects are produced in the muscle-tendon complex as a result of TFM applied over mechanoreceptor-rich tendon and myotendinous junction (MTJ). The time delays,belonging to EC coupling and force transmission, both contributingto EMD were computed during voluntary muscle contractionsbefore and after TFM. It was hypothesized that TFM may producetwitch contractile changes (EC coupling) inside the muscle whendetected by centrally mediated voluntary muscle contraction.

2. Methods

2.1. Subjects

Fifteen healthy male subjects and fifteen age-gender matchedcontrol subjects were recruited for the randomized control trial.Upon enrollment, participants were randomly assigned to TFM andcontrol group using simple computer program to generaterandomness. One of the subjects from TFM group was notcompliant with experimental procedure; therefore, the subject was

excluded from the study and final number of subjects included inTFM group was fourteen. All subjects were recruited from the Di-vision of Interdisciplinary Biomedical Engineering with a veryidentical age. They were without any history of previous injury,metabolic or neurologic disease. Not one of them was involved inany vigorous exercise on a daily basis. Not one of subjects wasfamiliar with particular group assignment or experimental proce-dure until the familiarization session. The human subject ethicalapproval was obtained from The Hong Kong PolytechnicUniversityHSEARS20140215001.

2.2. Experimental design

Before visiting the laboratory of the Interdisciplinary BiomedicalEngineering Division at the Hong Kong Polytechnic University forexperimental procedure, subjects participated in a familiarizationsession. The tested dominant leg including TFM was chosen as theleg with which the subject preferred to kick a ball. After theanthropometric measurements, subject was seated on the cali-brated dynamometer (Humac/Norm Testing and RehabilitationSystem, Computer Sports Medicine, Inc., MA,USA). The 30� of kneeflexion was chosen to activate the muscle with minimum pre-stretching of the muscle fibers (Sasaki et al., 2011).

The experimental procedure was well standardised with aminimised risk of experimental bias as data acquisition and anal-ysis were performed independent of each other at different times.The experimental procedure and signal analysis were performedmainly by authors, while TFM was performed by only one physicaltherapist.

Muscle activity during voluntary isometric contractions wasrecorded simultaneously by sEMG, Strain gauge and ultrafast US,while the subject was seated on the calibrated dynamometer withthe knee flexion angle adjusted at 30� (Begovic et al., 2014) (Fig 1A).The test procedure consisted of 4 repeated isometric contractionsfollowed bymassage in TFM group and resting in control group andended up again with 4 repeated contractions of the QF muscle.Between each contraction, a resting period of 2 min was allowed topreventmuscle fatigue (Shi et al., 2007). During the test, the subjectwas asked to apply maximum isometric contraction as quickly aspossible in 1 s and to keep it approximately for 3 s. Verbal orderwasgiven to the subject about the start and termination of the musclecontraction. The order “start” was given immediately after startingthe collection of A-mode signals in the ultrafast US device. After thetermination of each contraction, the position of the US probe waschecked to ensure that there was no displacement of the probecaused by the movement artifact of the muscle during contraction(Begovic et al., 2014; Shi et al., 2008; Shi et al., 2007; Strasser et al.,2013; Guo et al., 2010; Wang et al., 2014).

2.3. Electromyography

Two surface EMG bipolar AgeAgCl electrodes (Axon System,Inc., NY, USA) for differential EMG detection were attached on theRF muscle belly, approximately at the 50e60% of the distance be-tween the spina iliaca anterior superior and superior patellarmargin. To reduce the skin impedance, skin was cleaned with iso-propyl alcohol and abraded with fine sandpaper. The ground elec-trode was placed over the tibial crest. Interelectrode distancebetween two sEMG electrodes was 30 mm. The sEMG signal wasamplified by a custom designed amplifier with a gain of 2000,filtered by 10e1000 Hz bandpass analog filter within the amplifier,and digitized with a sampling rate of 4 KHz (Begovic et al., 2014;Guo et al., 2010).

Page 3: Begovic et al.2016

Fig. 1. Experimental set-up. A: Simultaneous recording of the Force signal, sEMG and Ultrafast US while subject performs an isometric contraction of the QF muscle. B: Applicationof TFM by experienced physiotherapist. C: Position of the fingers above the QF tendon checked on US image before application of TFM.

H. Begovic et al. / Manual Therapy 26 (2016) 70e7672

2.4. Torque and rate of force development

A dynamometer with a back inclination of 80� and knee flexionangle of 30� below the horizontal plane was used to measure tor-que Tti . The torque signals were digitized with a sampling rate of4 KHz, and stored on a personal computer. The RFD was also esti-mated using the torque:

RFD ¼ ðTt1 � Tt0Þ=ðt1 � t0Þ

Where t0 is the force onset time, and t1 is the time of reaching 10%maximum force. The average torque during the period of voluntaryisometric contraction was manually identified from the torquewave. (Shimose et al., 2014; Zhou et al., 1995)

2.5. Ultrasound

The US recording was made by a custom program installed in aprogrammable ultrasound scanner (Ultrasonix Touch, AnalogicCorporation, Massachusetts, USA) with a 7.5 MHz linear array ul-trasound probe (Ultrasonix L14-5/35) to achieve a very high frameultrasound scanning at a selected location. The US probewas placedas close as possible to the sEMG electrodes and longitudinally(Fig.1A,B). The A-mode US signal was collected at a frame rate of 4 kframes/s during 10s. After the first frame of A-mode signal wascollected, a signal was generated by the US scanner and outputtedas an external trigger signal, which was inputted into the device forEMG/FORCE signal collection. The recorded US signal was pro-cessed to detect the root mean square value (RMS) of the selectedregion of interest (ROI). This RMS value obtained from each frame ofUS signal was then substracted by the RMS value of the first frame

and the result was used to form a new signal representing the USsignal disturbance induced by the muscle contraction (Begovicet al., 2014; Hug et al., 2011a; Mannion et al., 2008).

2.6. Transverse friction massage

TFM was applied transversely to the pull line of the QF muscletendon by the top of the index andmiddle fingers (Fig 1C). Durationof TFM was about 15 min (Loew et al., 2014; Brosseau et al., 2002).After TFM, the lower leg was detached from the dynamometer leverarm because of uncomfortable feeling of tightness produced by thestraps over the lower part of the lower leg. The subject was advisednot to place the knee into deep flexion angle during TFM so to avoidstretching effect of the QF muscle and tendon.

2.7. Data acquisition and analysis

Collected signals were processed off-line using a programwritten in MatLab (version 2008a, USA). The EMG signal wasrectified and condition of three standard deviations (SD) from themean baseline noise was observed to detect the onset of eachsignal. In order to define crossing time as the onset time, a condi-tion for signal to stay 10ms above the threshold level was set by theprogram and visually examined. The time delays between onsets(DtEMG-US, DtUS-FORCE and DtEMG-FORCE), TORQUE and RFDwere calculated off-line for each contraction. The same experi-mental procedure was repeated immediately after an applied TFMin massage group and a resting period in control group (Fig 2).

Page 4: Begovic et al.2016

H. Begovic et al. / Manual Therapy 26 (2016) 70e76 73

2.8. Statistical analysis

The data were analyzed with a software package SPSS V.19(IBM,SPSS, Statistics for Windows, Version 19.0. Armonk,NY,USA).The normal distribution of the data was analyzed by the Kolmo-goroveSmirnov test. The One-Way Analysis of Variance (ANOVA)for repeated measures was used to check whether there are anydifferences between repeated contractions within the pre-measurements of both TFM and control groups. An average andSD were calculated if there was no difference between contractionsand this was used for the comparison between pre-TFM and post-TFM in the massage group and between pre-resting and post-resting in the control group. The paired sample T-test was usedfor comparison of the means within groups, whilst, 2-way ANOVAwas used for comparisons between groups.

3. Results

The results about demographic characteristics of the subjectsare demonstrated in Table 1. All data was normally distributed,therefore, One-Way ANOVA for repeated measures revealed thatthere was no difference at all between 4 contractions in all datafrom the pre and post tests (DtEMG-US, DtUS-FORCE, DtEMG-FORCE, TORQUE and RFD, p > 0.05), thus, averages were calculatedand used for comparisons.

After TFM, a significant increase in time delays of DtEMG-US(pre:19.2 ± 9.0 ms, post:28.5 ± 8.8 ms, p < 0.01) and DtEMG-FORCE(pre:49.7 ± 9.9 ms, post:54.1 ± 11.8 ms, p < 0.05) was found, whilstthe time delay of DtUS-FORCE (pre:30.5 ± 11.9 ms,post:25.5 ± 11.3 ms, p < 0.01) showed a decrease after TFM (Fig. 3).

In the control group, no significant difference between anyaverage of time delays was foundwhen compared to the before andafter resting period of 15 min (DtEMG-US; pre 16.8 ± 6.9 ms,post:16.6 ± 8.7 ms, p > 0.05: DtUS-FORCE, pre:31.2 ± 8.5 ms,post:30.7 ± 9.3 ms, p > 0.05: DtEMG-FORCE; pre 48.0 ± 8.4 ms,post:47.5 ± 9.9 ms, p > 0.05). However, expectedly, there was also

Fig. 2. An example of signal processing off-line using a program written in MatLab. Simulbefore and after TFM. As onsets of each signal were coming out in a sequential order, time dsEMG signal in designed MatLab program. B: disturbance signal acquired by ultrafast ultrascontraction of QF muscle.

no difference found between pre-measurements of TFM and thecontrol groups (DtEMG-US, p > 0.05: DtUS-FORCE, p > 0.05:DtEMG-FORCE, p > 0.05).

Since comparison before and after resting measurements in thecontrol group did not reveal any significant difference, the pre-resting results were compared with after-TFM results and signifi-cant increase of time delay was found in DtEMG-US, p < 0.01 andDtEMG-FORCE, p < 0.01 and significant decreasewas found inDtUS-FORCE, p < 0.01) (Fig. 4).

Comparison of the TORQUE and RFD between pre and postmeasurements did not reveal any significant difference in bothcontrol (Torque, pre:86.6 ± 21 Nm, post:83.8 ± 16 Nm: RFD,pre:182.6 ± 68.1 Nm/s, post:164.9 ± 49.0 Nm/s) and TFM groups(Torque, pre:93.7 ± 33 Nm, post:90.2 ± 34 Nm: RFD,pre:181.3 ± 108.8 Nm/s, post:167.7 ± 114.5 Nm/s) (Fig. 5).

4. Discussion

The TFM has produced an increase of both the time required forthe EC coupling (DtEMG-US] and an overall time delay named asthe EMD [DtEMG-FORCE] after TFM and when compared with thecontrol. A force transmission through non-contractile elements,when measured as a time delay between the onset of the fibermotion and the onset of the force output (DtUS-FORCE], has dis-played a significant decrease after TFM and also when comparedwith the control group.

To justify the effect of manual therapy, such as TFM, priorstudies had often been focused on the measurement of a restingEMG activity and muscle inhibition signs (Bennett et al., 2014;Bialosky et al., 2009). The neurophysiological mechanisminvolved in manual therapy is likely to originate from a peripheralmechanism, spinal cordmechanism and/or supraspinal mechanismbut an advanced modification of research methodology, such asours, was needed to find out peripheral changes first within con-tractile and non-contractile elements.

taneous recording of sEMG, US and Force was acquired during isometric contractionselays [Dt] were calculated off-line and statistically analyzed. A: automatically rectifiedound from the selected region of interest. C: force signal acquired by active isometric

Page 5: Begovic et al.2016

Table 1Physical and anthropometric characteristics of the participants of both massage andcontrol group.

TFM group (n ¼ 14) Control group (n ¼ 15)

Age (years) 28.2 ± 3.25 29 ± 4.69Weight (kg) 71.8 ± 10.16 73.6 ± 11.1Height (cm) 172.4 ± 5.84 175.6 ± 7.5BMI 24.1 ± 2.62 23.8 ± 2.72Mid-sagital

thickness ofthe RF (mm)

20.4 ± 2.40 12.3 ± 3.95

Fig. 3. Comparison of the Mean ± SD of 4 isometric contractions before and afterapplication of the TFM.

Fig. 4. Comparison of the means ± SD between pre-resting measurements of thecontrol and post-massage measurements of the TFD group.

Fig. 5. Comparison of the means ± SD torque and RFD results

H. Begovic et al. / Manual Therapy 26 (2016) 70e7674

4.1. Changes in electromechanical delay

It is likely that TFM, applied predominantly over the tendon andMTJ, increases the EMD between the moment of the action po-tential generation measured by surface EMG and force generationmeasured by dynamometer. This time course has been used as areliable measure under different experimental circumstances,when muscle dynamics were a matter of investigation (Espositoet al., 2011; Begovic et al., 2014; Zhou, 1996; Grosset et al., 2009;Hug et al., 2011(a); Hug et al., 2011(b)). The EMD has been attrib-uted to the inside changes emerging from the musculotendinouscompliance and stiffness (Shin & Sung, 2015), where the key roleplayers are contractile (active) and non-contractile (passive) com-ponents of the musculotendinous complex (Esposito et al., 2011;Hug et al., 2011(a); Hug et al., 2011(b)). It was reported that subjectswho showed the greatest increase in EMD also showed the greatestdecrease in musculotendinous stiffness, and vice versa (Grossetet al., 2009) but without detailed investigation of the stiffnesschanges in contractile and non-contractile elements.

4.2. Changes in EC coupling

EC coupling is one of the initial events inside the EMD, wherepresented time is thought to be the time between action potentialgeneration and cross-bridge formation between contractile com-ponents (Esposito et al., 2011; Hug et al., 2011(a); Hug et al.,2011(b)). When a muscle is stretched or massaged as a wholeanatomic structure without focused application over the muscle ortendon, detachment of the cross-bridges happens (Esposito et al.,2011; Reisman et al., 2009) and the time required for the ECcoupling is expected to be increased. In our study, some measuresof precaution have been taken not to massage the muscle tissueitself, relying on a popping feeling caused by the tendon motionbeneath the fingers, and ensuring that massage is applied over thetendon. In this case, without massaging the muscle itself, detectedincrease of time delay belonging to the EC coupling suggests thatTFM applied over the tendon may indirectly affect the twitchcontractile properties, via some afferent pathways and thereforedecreasing the active muscle stiffness. Since EC coupling has beendetected incorporating an US signal from inside the muscle,detecting the onset of the muscle tissue motion, we could suggestthat increase in both EC coupling and EMD could be the conse-quence of the reduction in the excitation of a-motor neurons. It islikely that this inhibition is being initiated by stimulated Golgi-tendon organs, which are predominantly located in the MTJ(Guissard and Duchateau, 2006), causing presynaptic inhibition ofthe a-motor neurons. It was also established that deep pressure

between PRE and POST in both control and TFD groups.

Page 6: Begovic et al.2016

H. Begovic et al. / Manual Therapy 26 (2016) 70e76 75

may produce synaptic changes in the brain, so that leads to thereduction of the reflex activity (Roberts, 2011). This may likely be anadditional mechanism for the increased time of EC couplingbecause of the involvement of the cortical areas during voluntarilyproduced muscle contraction in our study.

With indirect evidence, it has been demonstrated that pressure-sensitive and stretch-sensitive free nerve endings, connect toinhibitory neurons, and therefore play a role in reducing motor-neuron pool excitability (Lee et al., 2009; Ackermann PW, 2013).On the other hand, it has been established that some group-IIIafferent fibers, responsible for mediating mechanoreception,terminate in intramuscular connective tissue and within musclespindles, which implies that the mechanoreceptors within themuscle might respond to tendon stretch or pressure applied overtendon (Nakamoto and Matsukawa, 2007).

Given the fact that our results were detected during voluntarycontraction; a remaining question is whether increased EC couplingis caused by central mediation involving cortical area or some reflexmechanisms inside a muscle? To unveil the central mediation, avery recent study using different methodological approachesdetected the mean time delay between corticomotoneuronal cellfiring and the onset of facilitation of distal forelimb muscle activity.Many results have been reported this time course ranging from 60to 122 ms while a very recent study implementing an invasiveprocedure applied spike and stimulus triggered averaging of EMGactivity and found this time delay ranging from 6.7 to 9.8 ms,depending on the muscle group tested (Van Acker et al., 2015).However, we are not able to bring any explanation as to the possiblevariation of this time delay along the corticospinal pathway aftertransverse friction massage because an onset time in our mea-surement method was not from the cortical area. The contractionsin our experiment were voluntary muscle contractions, therefore,we measured the onsets of muscle dynamics (peripheral actions)by EMG, US and FORCE sensors relative to onset of the cortical area.The time delay between EMG and FORCE is well known anddescribed as EMD including EC coupling and force transmission,therefore, under our experimental circumstances, an increase of ECcoupling after TFM applied over tendon and excluding muscle itselfmay show a decrease in the active stiffness of the muscle itself.

4.3. Changes in force transmission

Since TFM was applied over the tendon, it is expected that themechanical properties (visco-elasticity) of the tendon tissue areaffected (Hunter, 1998) and consequently finding an increasedelongation of the tendon, possibly leading to an increase of the timeneeded for the force transmission during voluntary muscle con-tractions. Therefore, given the expectation to find out an increasedtime belonging to the force transmission, and therefore affectedviscoelasticity, we have found a significant decrease, which iscontradictory to the known effect of TFM. This decrease also be-comes statistically significant when compared with a control groupsuggesting that TFM applied over a tendon actually may stiffen thetendon material. Therefore, faster force transmission (an increasedloading of the tendon) may have happened during voluntarymuscle contraction. This time was calculated between the onset ofthe fibermotion detected by the US signal and the onset of the forceoutput. During the testing procedure, subjects were asked to pro-duce an isometric contraction within the first second which meansthat the tendon underwent very fast loading. As contraction in-tensity increases, the time needed for force transmission decreasestogether with an increasing in the intramuscular pressure (Espositoet al., 2009; Pfefer et al., 2009; Earp et al., 2014). This also causes acurvilinear increase of tendon stiffness as the force acting on the

tendon increases, resulting in a decreased tendon lengthening athigher force levels (Earp et al., 2014).

Very recently, a similar finding was the reason to raise a ques-tion, whether the application of transverse force componentsapplied over the muscle could be the cause of an increase of thelongitudinal loading of the MTU (Rushton & Spencer., 2011). Eventhough in that study results were obtained at the same time whilemassage was being performed, it is difficult to make an argumentbecause our results were detected several minutes after TFM.

In the present study, contraction intensity was standardised insuch a way that subjects were asked to apply maximal voluntarycontraction within the first second after the onset of thesynchronised recording. An investigation of the relationship ofloading (contraction) intensity and muscle-tendon unit behaviorshowed that increasing intensity caused a decrease in tendonlengthening in humans (Earp et al., 2014). The lesser tendonlenghtening during high contraction intensities means lessertendon compliance, faster force transmission, increased RFD, andtherefore increased loading of the tendon (Blackburn et al., 2009).With increasing rate of loading, the muscle gets stiffer and a shorterdisplay of the EMD is expected (Blackburn et al., 2009). On the otherhand, it has also been reported that during a sustained isometriccontraction, the intramuscular fluid and acid accumulate, andtherefore, intramuscular pressure and muscle volume increase(Crenshaw et al., 2000). This finding might be seen as disadvantagewhen investigating the effect of massage because of prevailing highcontraction intensity that may attenuate effect produced by themassage (Hunter et al., 2006). Therefore, in our study, fastcontraction within 1 s should be discussed as a possible method-ological limitation. On the other hand, there should be paidattention to our results because there was found a significantdecrease of time of force transmission along series elasticcomponents after TFM. It means that TFM is causing the stiffeningof the series elastic components (tendon). If found decreased timeof force transmission was caused by the TFM, then TFM should beapplied carefully before applying it on some sportsman who per-forms plyometric exercises or any form of forceful musclecontraction at any state of the pathological condition. This is animportant issue because we do not desire the overloading of thejoint in certain pathological conditions, particularly in the chronicstage of tendinitis.

4.4. Rate of force development and torque

In the present study, there has not been observed any significantchange, neither in Torque nor in RFD after TFM. This non-significantchange should be reinvestigated with redesigned methodology, inorder to find out the answer whether or how much a contractionintensity, including RFD and torque, is important when investi-gating a massage effect.

On the other hand, our results could be quite convincing sinceTFM, even at high muscle contraction intensities, shows thepersistent influence on the neuro-motor driving mechanism (sug-gested to be via afferent pathways) and eventually increases thetime within the EC coupling. Since this time has been detectedduring voluntary muscle contraction, it could be suggested thatdynamic stiffness of the muscle is most likely modified by theeffected activation patterns of the muscle. According to our results,suggested decreased stiffness after TFM, mainly preceded by timelyincreased EC coupling, is convincingly supported by our additionalfinding of an increased EMD, and eventually decreased activestiffness of the muscle. On the other hand, increased force trans-mission could suggest an increasing stiffness of the tendon.

Author’s contributions HB conceptualized the study. HB andGQZ conducted the experiments and analyzed the data. GQZ

Page 7: Begovic et al.2016

H. Begovic et al. / Manual Therapy 26 (2016) 70e7676

redesigned analyzing software. HB wrote the manuscript. YPZprovided experimental materials, conceptualized ultrafast ultra-sound, supervised research process and integrated the systemwithGQZ. SS reviewed the manuscript and provided feedbacks. All au-thors read and approved the final manuscript.

References

Ackermann PW. Neuronal regulation of tendon homoeostasis. Int J Exp Pathol2013;94:271e86.

Begovic H, Zhou G-Q, Li T, Wang Y, Zheng Y-P. Detection of the electromechanicaldelay and its components during voluntary isometric contraction of thequadriceps femoris muscle. Front Physiol 2014;5:494. http://dx.doi.org/10.3389/fphys.2014.00494.

Behm DG, Peach A, Maddigan M, Aboodarda SJ, DiSanto MC, Button DC, et al.Massage and stretching reduces spinal reflex excitability without affectingtwitch contractile properties. J Electromyogr Kinesiol 2013;23:1215e21.

Bennett A, Watson J, Simmonds J. The efficacy of the use of manual therapy in themanagement of tendinopathy: a systematic review. Br J Sports Med 2014;48:A11e2. http://dx.doi.org/10.1136/bjsports-2014-094114.17.

Bialosky EJ, Bishop DM, Price DD, Robinson EM, George ZS. The mechanisms ofmanual therapy in the treatment of musculoskeletal pain: a comprehensivemodel. Man Ther 2009;14:531e8. http://dx.doi.org/10.1016/j.math.2008.09.001.

Blackburn JT, Bell DR, Norcross MF, Hudson JD, Engstrom LA. Comparison ofhamstring neuromechanical properties between healthy males and females andthe influence of musculotendinous stiffness. J Electromyogr Kinesiol 2009;19:e362e369.

Brosseau L, Casimiro L, Milne S, Robinson V, Shea B, Tugwell P, et al. Deep transversefriction massage for treating tendinitis. Cochrane Database Syst Rev 2002;4:CD003528.

Crenshaw AG, Gerdle B, Heiden M, Karlsson S, Frid�en J. Intramuscular pressure andelectromyographic responses of the vastus lateralis muscle during repeatedmaximal isokinetic knee extensions. Acta Physiol Scand 2000;170:119e26.

Earp JE, Newton RU, Cormie P, Blazevich AJ. The influence of loading intensity onmuscle-tendon unit behavior during maximal knee extensor stretch shorteningcycle exercise. Eur J Appl Physiol 2014;114:59e69. http://dx.doi.org/10.1007/s00421-013-2744-2.

Eriksson Crommert M, Lacourpaille L, Heales LJ, Tucker K, Hug F. Massage inducesan immediate, albeit short-term, reduction in muscle stiffness. Scand J Med SciSports 2015;8. http://dx.doi.org/10.1111/sms.12341.

Esposito F, Ce E, Rampichini S, Veicsteinas A. Acute passive stretching in a previ-ously fatigued muscle: electrical and mechanical response during tetanicstimulation. J Sports Sci 2009;27:1347e57. http://dx.doi.org/10.1080/02640410903165093.

Esposito F, Limonta E, C�e E. Passive stretching effects on electromechanical delayand time course of recovery in human skeletal muscle new insights from anelectromyographic and mechanomyographic combined approach. Eur J ApplPhysiol 2011;111:485e95. http://dx.doi.org/10.1007/s00421-010-1659-4.

Goldberg J, Sullivan SJ, Seaborne ED. The effect of two intensities of massage on H-reflex amplitude. Phys Ther 1992;72:449e57.

Grosset JF, Piscione J, Lambertz D, P�erot C. Paired changes in electromechanicaldelay and musculo-tendinous stiffness after endurance or plyometric training.Eur J Appl Physiol 2009;105:131e9.

Guissard N, Duchateau J. Neuralaspects of muscle stretching. Exerc Sport Sci Rev2006;34:154e8.

Guo JY, Zheng YP, Xie HB, Chen X. Continuous monitoring of electromyography(EMG), mechanomyography (MMG), sonomyography (SMG) and torque outputduring ramp and step isometric contractions. Med Eng Phys 2010;32:1032e42.http://dx.doi.org/10.1016/j.medengphy.2010.07.004. Epub 2010 Aug 4.

Hug F, Gallot T, Catheline S, Nordez A. Electromechanical delay in biceps brachiiassessed by ultrafast ultrasonography. Muscle Nerve 2011a;43:441e3. http://dx.doi.org/10.1002/mus.21948.

Hug F, Lacourpaille L, Nordez A. Electromechanical delay measured during avoluntary contraction should be interpreted with caution. Muscle Nerve2011b;44:838e9. http://dx.doi.org/10.1002/mus.22139.

Hunter AM, Watt JM, Watt V, Galloway SD. Effect of lower limb massage on elec-tromyography and force production of the knee extensors. Br J Sports Med2006;40:114e8.

Hunter G. Specificsofttissue mobilisation in the management of softtissue-dysfunction. Man Ther 1998;3:2e11.

Kassolik K, Jask�olska A, Kisiel-Sajewicz K, Marusiak J, Kawczy�nski A, Jask�olski A.Tensegrityprinciple in massage demonstrated by electro- and mechanomyog-raphy. J Body Work Mov Ther 2009;13:164e70.

Kesson M, Atkins E. Orthopaedic medicine: a practical approach. Oxford, UK:Linacre House Butterworth-Heinemann; 1998. p. 63e101.

Kukulka CG, Beckman SM, Holte JB, Hoppenworth PK. Effects of intermittent tendonpressure on alpha motoneuron excitability. Phys Ther 1986;66:1091e4.

Lee HM, Wu SK, You JY. Quantitative application of transverse friction massage andits neurological effects on flexor carpi radialis. Man Ther 2009;14:501e7.

Leone JA, Kukulka CG. Effects of tendon pressure on alpha motoneuron excitabilityin patients with stroke. Phys Ther 1988;68:475e80.

Loew LM, Brosseau L, Tugwell P, Wells GA, Welch V, Shea B, et al. Deep transversefriction massage for treating lateral elbow or lateral knee tendinitis. CochraneDatabase Syst Rev 2014 Nov 8;11:CD003528. http://dx.doi.org/10.1002/14651858.CD003528.pub2.

Mannion AF, Pulkovski N, Schenk P, Hodges PW, Gerber H, Loupas T, et al. A newmethod for the non-invasive determination of abdominal muscle feedforwardactivity based on tissue velocity information from tissue Doppler imaging.J Appl Physiol 2008;104:1192e201. http://dx.doi.org/10.1152/japplphysiol.00794.2007.

Nakamoto T, Matsukawa K. Muscle mechanosensitive receptors close to the myo-tendinous junction of the Achilles tendon elicit a pressor reflex. J Appl Physiol2007;102:2112e20.

Newham DJ, Lederman E. Effect of manual therapy techniques on the stretch reflexin normal human quadriceps. Disabil Rehabil 1997;19:326e31.

Pfefer MT, Cooper SR, Uhl NL. Chiropractic management of tendinopathy: a litera-ture synthesis. J Manip Physiol Ther 2009;32:41e52.

Reisman S, Allen TJ, Proske U. Changes in passive tension after stretch of unexer-cised and eccentrically exercised human plantarflexor muscles. Exp Brain Res2009;193:545e54.

Roberts L. Effects of patterns of pressure application on resting electromyographyduring massage. Int J Ther Massage Bodyw 2011;30:4e11.

Rushton A, Spencer S. The effect of soft tissue mobilisation techniques on flexibilityand passive resistance in the hamstring muscle-tendon unit: a pilot investi-gation. Man Ther 2011;16:161e6.

Sasaki K, Sasaki T, Ishii N. Acceleration and force reveal different mechanisms ofelectromechanical delay. Med Sci Sports Exerc 2011;43:1200e6. http://dx.doi.org/10.1249/MSS.0b013e318209312c.

Shi J, Zheng YP, Chen X, Huang QH. Assessment of muscle fatigue using sono-myography: muscle thickness change detected from ultrasound images. MedEng Phys 2007;29:472e9.

Shi J, Zheng YP, Huang QH, Chen X. Continuous monitoring of sonomyography,electromyography and torque generated by normal upper arm muscles duringisometric contraction: sonomyography assessment for arm muscles. IEEE TransBiomed Eng 2008;55:1191e8. http://dx.doi.org/10.1109/TBME.2007.909538.

Shimose R, Ushigome N, Tadano C, Sugawara H, Yona M, Matsunaga A, et al. In-crease in rate of force development with skin cooling during isometric kneeextension. J Electromyogr Kinesiol 2014;24:895e901. http://dx.doi.org/10.1016/j.jelekin.2014.08.002.

Shin MS, Sung YH. Effects of massage on muscular strength and proprioceptionafter exercise-induced muscle damage. J Strength Cond Res 2015;29:2255e60.

Strasser EM, Draskovits T, Praschak M, Quittan M, Graf A. Association between ul-trasound measurements of muscle thickness, pennation angle, echogenicityand skeletal muscle strength in the elderly. Age 2013;35:2377e88. http://dx.doi.org/10.1007/s11357-013-9517-z.

Sullivan SJ, Williams LR, Seaborne DE, Morelli M. Effects of massage on alphamotoneuron excitability. Phys Ther 1991;71:555e60.

Van Acker GM, Luchies CW, Cheney PD. Timing of cortico-muscle transmissionduring active movement. Cereb Cortex 2015:1e10. http://dx.doi.org/10.1093/cercor/bhv151.

Wang CZ, Li TJ, Zheng YP. Shear modulus estimation on vastus intermedius ofelderly and young females over the entire range of isometric contraction. PLoSOne 2014;3(9):e101769. http://dx.doi.org/10.1371/journal.pone.0101769.

Yoo YS, Yu KP, Lee KJ, Kwak SH, Kim JY. Development and application of a newlydesigned massage instrument for deep cross-friction massage in chronic non-specific low back pain. Ann Rehabil Med 2012;36:55e65.

Zhou S, Lawson DL, Morrison WE, Fairweather I. Electromechanical delay in iso-metric muscle contractions evoked by voluntary, reflex and electrical stimula-tion. Eur J Appl Physiol Occup Physiol 1995;70:138e45.

Zhou S. Acute effect of repeated maximal isometric contraction on electrome-chanical delay of knee extensor muscle. J Electromyogr Kinesiol 1996;2:117e27.http://dx.doi.org/10.1016/1050-6411(95)00024-0.