Journal of Electromyography and Kinesiology 21 (2011) 33–40

Contents lists available at ScienceDirect

Journal of Electromyography and Kinesiology

journal homepage: www.elsevier .com/locate / je lek in

The influence of the muscle fiber pennation angle and innervation zone onthe identification of neuromuscular fatigue during cycle ergometry

Clayton L. Camic ⇑, Terry J. Housh, C. Russell Hendrix, Jorge M. Zuniga, Haley C. Bergstrom,Richard J. Schmidt, Glen O. JohnsonDepartment of Nutrition and Health Sciences, Human Performance Laboratory, University of Nebraska-Lincoln, Lincoln, NE 68583-0806, United States

a r t i c l e i n f o a b s t r a c t

Article history:Received 12 July 2010Received in revised form 28 September 2010Accepted 28 September 2010

Keywords:ElectromyographyEMG amplitudeMean power frequencyLinear electrode arrayPhysical working capacity

1050-6411/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.jelekin.2010.09.010

⇑ Corresponding author. Address: Department of N110 Ruth Leverton Hall, University of Nebraska-LincUnited States. Tel.: +1 (402) 472 2690; fax: +1 (402)

E-mail address: clcamic@unlserve.unl.edu (C.L. Cam

The purpose of the present investigation was to compare the electromyographic (EMG) responses and theestimated physical working capacity at the fatigue threshold (PWCFT) values recorded from electrodearrangements placed: (1) parallel to the muscle fiber pennation angle (MFPA), (2) parallel to the long axisof the femur, and (3) over the innervation zone (IZ) during incremental cycle ergometry. Thirteencollege-aged males and females (mean age ± SD = 22.4 ± 3.4 years) performed an incremental test toexhaustion on a cycle ergometer. A linear electrode array was utilized to determine the MFPA and loca-tion of the IZ of the vastus lateralis (VL). For determination of the PWCFT values, EMG signals wererecorded from three bipolar electrode arrangements at different locations over the VL. The results of aone-way repeated measures ANOVA indicated there were no significant (p < 0.05) mean differences inPWCFT values among the electrode arrangements (parallel to the MFPA = 190 ± 36 W; parallel to the longaxis of the femur = 194 ± 40 W; and over the IZ = 199 ± 51 W) or the EMG amplitude and MPF values atthe common power outputs. There were also significant correlations (r = 0.75–0.91) among the threeelectrode arrangements for PWCFT values. These findings suggested that the PWCFT, like absolute EMGamplitude and MPF, is robust to the influence of electrode placement over the IZ as well as the orientationwith respect to the MFPA during cycle ergometry.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Theoretically, the physical working capacity at the fatiguethreshold (PWCFT) test estimates the highest power output duringcycle ergometry that can be maintained without neuromuscularevidence of fatigue. Based on increases in electromyographic(EMG) amplitude, the PWCFT provides information related, in part,to the fatigue-induced recruitment of additional motor units andincreased firing rates. To determine the PWCFT, EMG amplitudevalues are recorded, quantified, and plotted versus time for each2-min power output of an incremental cycle ergometer test.Specifically, the PWCFT is then defined as the average of the highestpower output that results in a non-significant slope coefficient forthe EMG amplitude versus time relationship, and the lowest poweroutput that results in a significant positive slope coefficient (Fig. 1).Previous studies (deVries et al., 1987, 1990) have shown that thePWCFT is moderately correlated with the onset of blood lactate,can be determined using only submaximal workbouts, and is sen-

ll rights reserved.

utrition and Health Sciences,oln, Lincoln, NE 68583-0806,472 0522.

ic).

sitive to training adaptations in older populations. The PWCFT testhas also been utilized to assess physical fitness (deVries et al.,1987), differentiate fatiguing from non-fatiguing work (deVrieset al., 1987), demarcate the moderate and heavy exercise intensitydomains (Camic et al., 2010a), examine the physiological mecha-nisms associated with neuromuscular fatigue (Housh et al.,1991), and evaluate the efficacy of creatine, beta-alanine, and argi-nine as ergogenic aids (Camic et al., 2010b; Stout et al., 2000,2008). Furthermore, it has been demonstrated (deVries et al.,1990; Housh et al., 1996) that the PWCFT test is not affected by dif-ferences in protocol (i.e. continuous versus discontinuous incre-mental cycle ergometry), the muscle examined (i.e. vastuslateralis, vastus medialis, or rectus femoris), or EMG electrodearrangements (i.e. unipolar versus bipolar).

Recent investigations (Farina et al., 2001; Hermens et al., 1999,2000; Rainoldi et al., 2000, 2004; Sacco et al., 2009) have addressedthe importance of proper electrode placement over the muscle ofinterest for the recording of surface EMG signals. For example, withelectrode placement over the innervation zone (IZ), the non-propagating and incomplete action potentials that occur in thisarea contribute high frequency power to the EMG signal (Farinaet al., 2004b). In addition, previous studies (Rainoldi et al., 2000,2004) have demonstrated that the placement of electrodes over

Fig. 1. Illustration of the method used for estimating the physical working capacity at fatigue threshold (PWCFT). The PWCFT in the current example (155 W) was determinedby averaging the highest power output (140 W) that resulted in a non-significant (p > 0.05) slope coefficient for the EMG amplitude versus time relationship, with the lowestpower output (170 W) that resulted in a significant (p < 0.05) positive slope coefficient. *Slope coefficient significantly greater than zero at p < 0.05.

34 C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40

the IZ results in a large amount of signal cancellation due to differ-ential amplification. In conjunction, these factors are likely respon-sible for the findings of recent investigations that havedemonstrated that the placement of a bipolar electrode arrange-ment over the IZ results in significantly lower EMG amplitude(Beck et al., 2008c,d; Farina et al., 2001; Rainoldi et al., 2000) andhigher EMG mean power frequency (MPF) values (Beck et al.,2008c; Farina et al., 2001) compared to a bipolar electrode arrange-ment located away from the IZ. Thus, the EMG amplitude and MPFvalues recorded over the IZ may not accurately represent the elec-trical status of the overall muscle.

One of the most common methods used to non-invasively iden-tify the location of the IZ is with a linear electrode array (Merlettiet al., 1999). In particular, a linear electrode array detects multipleEMG signals from a number of electrodes (typically 4, 8, or 16) lo-cated in a straight line (Merletti et al., 2003). The IZ is then identi-fied as the EMG channel that demonstrates minimum amplitudeand the horizontal axis from which motor unit action potentialswill propagate bilaterally in both proximal and distal directions(Merletti et al., 1999, 2003). Recent investigations (Beck et al.,2008d; Rainoldi et al., 2000, 2004) have also utilized linear elec-trode arrays to determine the muscle fiber pennation angle(MFPA). Like identification of the IZ, the determination of the MFPAis of practical importance because it has been demonstrated(Andreassen and Rosenfalck, 1978; Zuniga et al., 2009) that the ori-entation of a bipolar electrode arrangement with respect to theMFPA can affect the amplitude and frequency contents of theEMG signal. For example, Zuniga et al. (2009) recently found signif-icant mean differences in absolute and normalized EMG amplitudeand MPF values for the vastus lateralis during cycle ergometry be-tween an electrode arrangement placed parallel to the MFPA andan electrode arrangement with the same center point, but placedperpendicular to the MFPA. Furthermore, according to the SENIAMProject (Hermens et al., 1999), it is recommended that bipolarelectrode arrangements be oriented parallel to the MFPA for thevastus lateralis muscle. Thus, based on the effects of the IZ andMFPA on the amplitude and frequency values of the EMG signal,as well as the recommendations of SENIAM (Hermens et al.,1999), these factors should be considered when placing bipolarelectrodes over the muscle of interest.

The majority of investigations (deVries et al., 1990; Housh et al.,1996; Stout et al., 2000, 2008) that have examined the PWCFT have

utilized bipolar electrode arrangements placed parallel to the longaxis of the femur, approximately 20� from the MFPA of the vastuslateralis. The findings of these studies (deVries et al., 1990; Houshet al., 1996) have shown that the PWCFT test is not affected by anumber of factors including protocol design, muscle examined,and residual fatigue. The effect of electrode placement over theIZ on the PWCFT, however, has not been examined. Furthermore,it is unclear if electrode placement oriented parallel to the MFPA,as recommended by SENIAM (Hermens et al., 1999), will influencethe outcome of the PWCFT test. Therefore, the purpose of the pres-ent investigation was to compare the EMG responses and the esti-mated PWCFT values recorded from electrode arrangementsplaced: (1) parallel to the MFPA, (2) parallel to the long axis ofthe femur, and (3) over the IZ during incremental cycle ergometry.

2. Methods

2.1. Subjects

Thirteen college-aged males (n = 8) and females (n = 5)(mean ± SD age = 22.4 ± 3.4 years; body weight = 74.5 ± 11.3 kg;height = 176.7 ± 9.3 cm) volunteered to participate in this investi-gation. The training status of the subjects ranged from untrainedto moderately-trained (approximately 4–5 resistance training oraerobic training sessions per week). The study was approved bythe University Institutional Review Board for Human Subjects,and all participants completed a health history questionnaire andsigned a written informed consent document prior to testing.

2.2. Determination of _VO2peak

Each subject performed an incremental test to exhaustion on aCalibrated Lode (Corival V3) electronically-braked cycle ergometerat a pedal cadence of 70 rev min�1. Seat height was adjusted sothat the subject’s legs were near full extension during each pedalrevolution. In addition, toe clips were used to ensure that eachsubject maintained pedal contact throughout the ride. All subjectswore a nose clip and breathed through a 2-way valve (2700; HansRudolph, Kansas City, MO). Expired gas samples were collected andanalyzed using a calibrated metabolic cart (TrueMax 2400, ParvoMedics, Sandy, UT) with O2, CO2, and ventilatory parameters

C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40 35

expressed as 30-s averages. The metabolic cart was calibrated priorto each test. Each subject was fitted with a Polar Heart Watch sys-tem (Polar Electro Inc., Lake Success, NY) to monitor heart ratethroughout the test. The test began at 50 W and the power outputincreased by 30 W every 2 min until voluntary exhaustion or thesubject could no longer maintain a pedal cadence of 70 rev min�1

despite strong verbal encouragement. _VO2peak was the highest_VO2 value in the last 30 s of the exercise test that met at leasttwo of the following three criteria (Day et al., 2003): (1) 90% ofage-predicted maximum heart rate; (2) respiratory exchange ratio>1.1; and (3) a plateau of oxygen uptake (less than 150 mL min�1

in _VO2 over the last 30 s of the test). From previous work, thetest–retest reliability for _VO2peak testing (n = 20) from our labora-tory indicated the intraclass correlation coefficient (ICC) wasR = 0.95, with no significant (p > 0.05) mean difference betweentest and re-test values.

2.3. Determination of the IZ and MFPA

Surface EMG signals were examined from the vastus lateralismuscle of the dominant thigh (based on kicking preference) witha 16-channel linear electrode array and surface EMG16 data acqui-sition system (EMG16, LISiN-Prima Biomedical and Sport, Treviso,Italy). The linear electrode array was placed over the vastus latera-lis according to the procedures described in the EMG16 User’sManual (2006). Specifically, the skin over the belly of the vastuslateralis was carefully shaved and wetted with a damp cloth. Aprobe of 16 silver bar electrodes (5 mm � 1 mm, 10 mm interelec-trode distance, Ottino Bioelettronica, Torino, Italy) was then placedover the belly of the vastus lateralis and oriented at a 20� angle(relative to a reference line between the anterior superior iliacspine (ASIS) and superior lateral border of the patella) to approxi-mate the MFPA (Abe et al., 2000). The reference electrode strap waslocated around the subject’s wrist (EMG16 User’s Manual, 2006, p.25). The subject was then asked to contract the dominant legextensors at approximately 50% of their maximum, and the EMGsignals from the probe were differentially amplified (gain2000�), analog filtered (fourth-order Bessel, bandwidth =10–500 Hz), and displayed on a computer screen. The location ofthe IZ for the vastus lateralis was identified by the EMG channelthat had minimum amplitude and phase reversal (Fig. 2) (Merlettiet al., 1999; Rainoldi et al., 2000, 2004) and marked with ink. Theprobe was then moved along the muscle in a direction approxi-mately parallel to the fibers until the IZ was near the middle ofthe electrode array (EMG16 User’s Manual, 2006, p. 26). This pro-cedure ensured that all electrodes from the linear array were overthe vastus lateralis muscle. To align the electrode array with thelong axis of the muscle fibers, the probe was then rotated aroundthe IZ until the slopes of the two lines connecting the EMG wave-forms from channels above and below the IZ were approximatelyequal (EMG16 User’s Manual, 2006, p. 26) (Fig. 2). The outline ofthe probe was then traced with a marker and the MFPA was mea-sured with a standard goniometer (Smith & Nephew Rolyan, Inc.,Menomonee Falls, WI) as the angle of the probe outline relativeto the reference line between the anterior superior iliac spineand superior lateral border of the patella.

2.4. EMG measurements

Three bipolar (30 mm center-to-center) surface electrode (cir-cular 4 mm diameter silver/silver chloride, BIOPAC Systems, Inc.,Santa Barbara, CA) arrangements were placed on the dominantleg over the vastus lateralis muscle (Fig. 3). One electrode arrange-ment was placed on the vastus lateralis at one-third of the distancebetween the superior lateral border of the patella and the anteriorsuperior iliac spine (ASIS) and oriented parallel to the measured

MFPA. These points were measured with the subject in the stand-ing position and the dominant leg fully extended. In addition, theelectrode-placement site was located 5 cm lateral to the referenceline so that they would lie over the vastus lateralis muscle (Maleket al., 2006). A second electrode arrangement utilized the samecenter point as the first electrode arrangement, but was orientedparallel to the long axis of the femur (deVries et al., 1990; Houshet al., 1996). A third electrode arrangement was placed over theIZ and oriented parallel to the MFPA. The location of the IZ wasmeasured as the distance from the superior lateral border of thepatella to the center point of the bipolar electrodes over the IZ.Three reference electrodes were placed over the iliac crest. Priorto electrode placement, the skin at each electrode site was shaved,carefully abraded, and cleaned with alcohol. Interelectrode imped-ance was less than 2000 X. The EMG signals were amplified (gain:1000�) using differential amplifiers (EMG 100, Biopac Systems,Inc., Santa Barbara, CA, bandwidth = 10–500 Hz).

2.5. Signal processing

The raw EMG signals were digitized at 1000 Hz and stored in apersonal computer (Inspiron 1501, Dell, Inc., Round Rock, TX) forsubsequent analyses. All signal processing was performed usingcustom programs, which had been written with LabVIEW pro-gramming software (version 8.5, National Instruments, Austin,TX). The EMG signals were digitally bandpass filtered (fourth-orderButterworth) at 10–500 Hz.

2.6. Estimated amplitude and mean power frequency

Amplitude (microvolts root mean square, lVrms) and MPF (Hz)values were calculated for each power output. Specifically, the con-traction period for each pedal thrust (descending phase of the ped-aling cycle) was determined from the EMG bursts according to theprocedure of Shinohara et al. (1997). The amplitude and MPFvalues for the EMG signals were computed using 0.45 epochs with-in a burst and were estimated as the average from the last three.For the MPF analyses, each data segment was processed with aHamming window and a discrete Fourier transform (DFT) algo-rithm. The MPF was selected to represent the power spectrum inaccordance with the recommendations of Hermens et al. (1999)and was calculated with the following equation:

MPF ¼Pfc

f¼f0fPðf Þ

Pfcf¼f0

Pðf Þ

where f is frequency, f0 is 0 Hz, fc is the cutoff frequency (i.e. the lastfrequency in the discrete summation), and P(f) is the power density(V2/Hz) of the EMG signal.

2.7. Determination of PWCFT

The method used to determine the PWCFT values was consistentwith the protocol of deVries et al. (1987). Specifically, during each2 min power output of the incremental cycle ergometer test toexhaustion, six, 10-s EMG segments were recorded from the vastuslateralis muscle for each electrode arrangement. That is, EMGamplitude values were calculated for 10-s epochs during each 2-min power output (50, 80, 110 W, etc.) for seconds 10–20, 30–40,50–60, 70–80, 90–100, and 110–120. The EMG amplitude valuesfor each of the 10-s epochs were then plotted versus time for eachpower output of the test (Fig. 1). The relationship between EMGamplitude and time (seconds) for each power output was exam-ined with linear regression to determine if there was a significantincrease in EMG amplitude across time. The PWCFT was defined asthe average of the highest power output that resulted in a non-sig-

Fig. 2. Example of the method used to determine the muscle fiber pennation angle (MFPA) and the innervation zone (IZ) for the vastus lateralis muscle with a linear electrodearray (EMG16 User’s Manual, 2006, p. 26). The location of the IZ (channel 7) was identified by the EMG channel that had minimum amplitude and phase reversal. The MFPAwas determined from the position of the array when the slopes of the two lines connecting the EMG waveforms from the channels above and below the IZ wereapproximately equal.

Fig. 3. Example of the locations for the bipolar EMG electrode arrangements placedparallel to the muscle fiber pennation angle (1), parallel to the long axis of thefemur (2), and over the innervation zone (3).

36 C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40

nificant (p > 0.05; single-tailed t-test) slope coefficient for the EMGamplitude versus time relationship, and the lowest power outputthat resulted in a significant (p < 0.05) positive slope coefficient(Fig. 1). From previous work, the test–retest reliability for PWCFT

testing (n = 20) from our laboratory indicated that the ICC wasR = 0.95, with no significant mean difference between test andre-test values.

2.8. Statistical analyses

One-way repeated measures ANOVAs were used to determine ifthere were significant mean differences among the three electrodearrangements for PWCFT, as well as absolute EMG amplitude andMPF values at each of the six power outputs (50, 80, 110, 140,170, and 200 W) that were common to all of the subjects. Compar-isons between absolute values were justified because of the with-in-subjects design utilized in the current study. A zero-ordercorrelation matrix was used to determine the relationships amongelectrode arrangements for PWCFT values. An alpha of p < 0.05 wasconsidered statistically significant for all comparisons.

3. Results

Table 1 provides mean, standard deviation, and range values forthe physical characteristics and PWCFT values of the subjects. Fig. 2provides an example that the IZ and the MFPA were indeed obser-vable. The results of the one-way repeated measures ANOVAs indi-cated there were no significant mean differences among the threeelectrode arrangements for PWCFT values (parallel to theMFPA = 190 ± 36 W; parallel to the long axis of the fe-mur = 194 ± 40 W; and over the IZ = 199 ± 51 W) (Table 1) or abso-lute EMG amplitude and MPF at the common power outputs of 50,80, 110, 140, 170, and 200 W (Table 2). In addition, there were sig-nificant correlations (r = 0.75–0.91) among the PWCFT values for allthree electrode arrangements (Table 3).

Table 1Physical characteristics and PWCFT values of the subjects (n = 13).

Variable Mean ± SD Range

Age (years) 22.4 ± 3.4 20.0–29.0Body weight (kg) 74.5 ± 11.3 52.2–88.5Height (cm) 176.7 ± 9.3 162.6–193.5Innervation zone of vastus lateralis (cm)* 11.0 ± 2.0 8.8–15.0Muscle fiber pennation angle of vastus

lateralis (�)20.8 ± 0.9 20.0–22.5

PWCFT (W)Parallel to the muscle fiber pennation angle 190 ± 36 125–245Parallel to the long axis of the femur 194 ± 40 125–245Over the innervation zone 199 ± 51 95–275

* Distance from the superior lateral border of the patella to the center point of thebipolar electrodes over the innervation zone. There were no significant (p > 0.05)mean differences among the electrode arrangements for PWCFT values.

Table 2Mean (±SEM) absolute EMG amplitude and mean power frequency for the threeelectrode arrangements over the vastus lateralis.

Poweroutput

Electrode arrangement

Parallel to the muscle fiberpennation angle

Parallel to the longaxis of the femur

Over theinnervationzone

Absolute EMG amplitude (lVrms)50 92.6 ± 7.9 80.2 ± 5.5 85.1 ± 11.880 132.0 ± 12.6 119.1 ± 9.0 127.4 ± 15.7

110 160.0 ± 17.0 141.9 ± 13.1 153.9 ± 17.4140 193.5 ± 14.8 170.4 ± 12.4 195.8 ± 20.4170 244.9 ± 25.7 210.2 ± 18.7 239.2 ± 31.5200 306.8 ± 34.3 262.5 ± 25.7 307.9 ± 32.1

Absolute EMG mean power frequency (Hz)50 72.4 ± 4.6 76.0 ± 4.2 69.1 ± 2.780 71.5 ± 3.9 74.0 ± 3.6 70.4 ± 2.7

110 73.4 ± 4.7 77.4 ± 4.2 71.2 ± 2.8140 73.7 ± 3.8 77.7 ± 3.7 72.8 ± 2.3170 71.6 ± 3.7 76.9 ± 3.3 73.0 ± 3.4200 70.2 ± 3.9 75.5 ± 3.8 73.6 ± 4.4

There were no significant (p > 0.05) mean differences among electrode arrange-ments for absolute EMG amplitude or mean power frequency values for any of thepower outputs.

Table 3Correlations for PWCFT values among electrode placements.

Parallel to the musclefiber pennation angle

Parallel to thelong axis of thefemur

Over theinnervationzone

Parallel to the muscle fiber pennation angleParallel to the

long axis of thefemur

0.91*

Over theinnervationzone

0.84* 0.75*

* p < 0.05.

C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40 37

4. Discussion

In the present investigation, the mean (±SD) MFPA of the vastuslateralis at a knee joint angle of 120� was 20.8 ± 0.9� during an iso-metric muscle action at approximately 50% MVC. Previous studies(Abe et al., 1998, 1999, 2000; Fukunaga et al., 1997) that haveexamined the muscle architecture of the vastus lateralis havefound similar findings. For example, Abe et al. (1999, 2000) re-ported mean MFPA values for the vastus lateralis that ranged from18.5� to 23.7�. In addition, Fukunaga et al. (1997) demonstrated a

mean increase in the MFPA (14–21�) for the vastus lateralis withan increase in the angle of the knee joint from 70� to 180� (fullextension). Thus, the findings of the present study and those ofothers (Abe et al., 1998, 1999, 2000; Fukunaga et al., 1997) sug-gested that the mean MFPA of the vastus lateralis may typicallyrange from approximately 14� to 23.7� depending on the angle ofthe knee joint and level of muscular contraction.

The results of this investigation indicated that the IZ for the vastuslateralis was located 11.0 ± 2.0 cm proximal from superior lateralborder of the patella and 2–3 cm lateral from the reference line be-tween the superior lateral border of the patella and the ASIS. This va-lue was slightly greater than those of Beck et al. (2008d) and Rainoldiet al. (2004) who reported mean (±SD) values of 9.8 ± 1.8 and9.4 ± 1.3 cm, respectively, for the location of the IZ along the sameline of reference. It is possible that the slightly greater values for IZlocation in the present investigation compared to those of Becket al. (2008d) and Rainoldi et al. (2004) were due to differences inmeasurement procedures. That is, in the current study, we measuredthe distance from the superior lateral border of the patella directly tothe center point of the bipolar electrode arrangement over the IZ,whereas Beck et al. (2008d) and Rainoldi et al. (2004) describedthe location of the IZ as the distance along the line of reference be-tween the superior lateral border of the patella and the ASIS. Thus,the measurements of Beck et al. (2008d) and Rainoldi et al. (2004)were taken approximately 2–3 cm medial to the location of the elec-trode arrangement over the IZ used in the present study.

The results of the current study indicated that there were nosignificant mean differences among the three electrode arrange-ments for absolute EMG amplitude at the common power outputsof 50, 80, 110, 140, 170, and 200 W (Table 2). Based on the findingsof previous investigations (Beck et al., 2008a,c,d; Farina et al.,2001; Rainoldi et al., 2000), however, we hypothesized that theelectrode arrangement over the IZ would result in significantlylower absolute EMG amplitude values than the electrode arrange-ments away from the IZ. For example, Rainoldi et al. (2000) re-ported lower absolute EMG amplitude values for bipolarelectrodes located over the IZ (IED = 10 mm) compared to siteson the muscle that were proximal or distal to the IZ (IED = 10 mm)during submaximal isometric muscle actions of the leg extensors.It has been suggested (Rainoldi et al., 2004) that the lower absoluteEMG amplitude values recorded over the IZ are likely due to actionpotentials from the motor end plate reaching each electrode in abipolar configuration simultaneously. Theoretically, these actionpotentials would be cancelled by the differential amplifier andnot contribute to the recorded EMG signal. In the current study,it is possible that the electrode arrangement located over the IZdid not result in significantly lower absolute EMG amplitude val-ues compared to the electrode arrangements located away fromthe IZ due to movement in the underlying muscle and IZ in relationto the bipolar electrodes (Farina et al., 2001). In particular, it hasbeen demonstrated that a change in the joint angle at the kneeof 90� (75–165�, with 180� being full extension) can result in aproximal shift of the IZ for the vastus lateralis muscle of approxi-mately 10 mm relative to the center point of a bipolar electrodeconfiguration (Farina et al., 2001). During the cycle ergometer testused in the current investigation, the approximate range of the an-gle at the knee joint was also 90�, from 70� at top-dead center to160� at bottom-dead center. Therefore, it is likely that the IZ wasrarely directly under the center point of the bipolar electrodearrangement. Theoretically, the proximal shift of the IZ during dy-namic muscle actions would result in the IZ not being an equal dis-tance from each electrode in a bipolar arrangement, therebyreducing the amount of the differential signal cancellation thathas been demonstrated during isometric muscle actions. Recently,Beck et al. (2008b) examined the effects of the IZ during isometricmuscle actions on absolute EMG amplitude from bipolar electrode

38 C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40

arrangements (IED = 30 mm) with center points located directlyover the IZ, 10 mm distal to the IZ, and 10 mm proximal to theIZ. The authors (Beck et al., 2008b) reported that compared tothe electrode arrangement placed directly over the IZ, the electrodearrangements with the center point shifted 10 mm proximal ordistal to the IZ resulted in significantly greater absolute EMGamplitude values. Thus, the findings of the present study in con-junction with those of others (Beck et al., 2008b; Farina et al.,2001), demonstrated that the absolute EMG amplitude values re-corded over the IZ may not differ from those recorded proximalor distal to the IZ during dynamic muscle actions due to a shiftin the underlying muscle relative to the surface electrodes.

In the present study, there were also no significant mean differ-ences among the three electrode arrangements for the absoluteEMG MPF values at the common power outputs (Table 2). Previousstudies (Beck et al., 2008c; Farina et al., 2001) that have examinedisometric muscle actions, however, have shown that the frequencycontent of the EMG signal can be affected by electrode placementover the IZ. Specifically, from 10% to 100% MVC, significantly great-er absolute EMG MPF values have been reported (Beck et al.,2008c) from the vastus medialis for an electrode arrangement(IED = 30 mm) with the center point located over the IZ versus anelectrode arrangement (IED = 30 mm) with the center point placed20 mm distal of the IZ. Thus, the findings of Beck et al. (2008c) sup-ported the suggestion that the frequency content of the EMG signalmay be affected by the IZ shift in relation to the center point of abipolar electrode arrangement that is associated with cycle ergom-etry. It has been suggested (Farina et al., 2004b) that compared tolocations away from the IZ, the non-propagating action potentialscontribute more to the signal when measured at the motor endplate. Due to their high temporal frequency and low spatial (i.e.non-traveling) frequency content, non-propagating action poten-tials recorded over the IZ can contribute high-frequency power tothe EMG signal, resulting in greater absolute EMG MPF values (Far-ina et al., 2004b). Similar to the results of the current study, how-ever, EMG MPF measured over the IZ is not always greater thanwhen measured away from the IZ (Beck et al., 2008b,d). One possi-ble explanation for this finding is that the IED (30 mm) for thebipolar arrangement used in the current study resulted in a pick-up area that was not selective enough to be affected only by theIZ (Beck et al., 2009). That is, the EMG signal recorded from thebipolar electrode arrangement over the IZ may have been influ-enced by non-propagating as well as propagating actionspotentials.

The PWCFT test involves calculation of the slope coefficients ofEMG amplitude fatigue curves at each power output during anincremental cycle ergometer test (deVries et al., 1987). The PWCFT

is then defined as the average of the highest power output that re-sults in a non-significant slope coefficient for the EMG amplitudeversus time relationship, and the lowest power output that resultsin a significant positive slope coefficient. In the present investiga-tion, there were no significant mean differences among the threeelectrode arrangements for PWCFT values (parallel to theMFPA = 190 ± 36 W; parallel to the long axis of the fe-mur = 194 ± 40 W; and over the IZ = 199 ± 51 W) (Table 1). Thesefindings indicated that estimation of the PWCFT was not affectedby an electrode placement that was over the IZ. It is possible thatthe PWCFT values recorded over the IZ were similar to those re-corded away from the IZ due to the proximal shift of the underly-ing muscle. Therefore, the current findings indicated that, like themean absolute EMG amplitude and MPF values, estimation of thePWCFT was robust to an electrode placement over the IZ during cy-cle ergometry.

Previous studies (Housh et al., 1996; deVries et al., 1990) thathave determined the PWCFT from the vastus lateralis muscle haveutilized electrode arrangements placed parallel to the long axis of

the femur. Recently, however, a number of sources (Farina et al.,2004a; Hermens et al., 1999, 2000; Mesin et al., 2009; Merceret al., 2006; Rainoldi et al., 2004; Sacco et al., 2009) have describedthe importance of standardizing electrode location for recordingthe EMG signal from specific muscles. For example, Hermenset al. (1999) provided recommendations regarding the placementof surface EMG electrodes for 27 muscles to minimize the influenceof factors that may affect the amplitude and frequency contents ofthe EMG signal. In particular, the authors (Hermens et al., 1999)suggested that for the vastus lateralis, bipolar electrodes shouldbe placed one-third of the distance from the lateral border of thepatella to the anterior superior iliac spine and ‘‘. . .in the directionof the muscle fibers’’ (p. 45). Therefore, one purpose of the presentinvestigation was to compare the EMG amplitude and MPF re-sponses, as well as the estimated PWCFT values, from an electrodearrangement placed parallel to the MFPA as recommended by Her-mens et al. (1999, 2000) to an electrode arrangement with thesame center point, but oriented parallel to the long axis of the fe-mur. The current findings indicated that the absolute EMG ampli-tude and MPF responses, as well as PWCFT values, were not affectedby the orientation of the electrode arrangement with respect to theMFPA. One possible explanation for this finding is that the elec-trode arrangement oriented parallel to the MFPA and the electrodearrangement oriented parallel to the long axis of the femur re-sulted in similar pick-up areas. For example, Andreassen andRosenfalck (1978) demonstrated that compared to an electrodearrangement placed perpendicular (i.e. at 90�) to the MFPA, anelectrode arrangement located parallel to the MFPA would resultin a larger pick-up area and thus, record the EMG amplitude andMPF responses from a greater number of muscle fibers. In the cur-rent study, however, the electrode arrangement placed parallel tothe long axis of the femur was oriented approximately 20� with re-spect to the MFPA (Fig. 3). Therefore, unlike the findings of Andre-assen and Rosenfalck (1978) for a perpendicular (90�) electrodearrangement, it is possible that the 20� difference in orientationsbetween the electrode arrangements in the present study wasnot large enough to significantly affect the absolute EMG ampli-tude and MPF responses.

In summary, the results of the current investigation indicatedthat during incremental cycle ergometry, there were no significantmean differences among electrode arrangements located over theIZ versus away from the IZ for absolute EMG amplitude and MPFas well as PWCFT values. In contrast to previous studies that haveutilized isometric muscle actions, these findings were likely attrib-utable to the proximal shift in the IZ in relation to the bipolar elec-trodes that is associated with concentric shortening of the activemuscle fibers and changes in joint angle during dynamic muscleactions. It is also possible that the IED (30 mm) for the electrodearrangement located over the IZ in the current study resulted ina pick-up area that was not selective enough to be affected onlyby the IZ and may have also been influenced by propagating ac-tions potentials away from the IZ. These findings suggested thatestimation of the PWCFT, like absolute EMG amplitude and MPF,is robust to the influence of electrode placement over the IZ duringcycle ergometry. Furthermore, the absolute EMG amplitude andMPF responses, as well as PWCFT values, were not affected by theorientation of the electrode arrangements with respect to theMFPA. These findings were likely due to a relatively small differ-ence (20�) in orientation angles between the electrode arrange-ments, thereby resulting in similar pick-up areas.

References

Abe T, Brechue WF, Fujita S, Brown JB. Gender differences in FFM accumulation andarchitectural characteristics of muscle. Med Sci Sport Exerc 1998;30:1066–70.

C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40 39

Abe T, Brown JB, Brechue WF. Architectural characteristics of muscle in black andwhite college football players. Med Sci Sport Exerc 1999;31:1448–52.

Abe T, Kumagai K, Brechue WF. Fascicle length of leg muscles is greater in sprintersthan distance runners. Med Sci Sport Exerc 2000;32:1125–9.

Andreassen S, Rosenfalck A. Recording from a single motor unit during strong effort.IEEE Trans Biomed Eng 1978;25:501–8.

Beck TW, Housh TJ, Cramer JT, Malek MH, Mielke M, Hendrix R. The effects of theinnervation zone and interelectrode distance on the patterns of responses forelectromyographic amplitude and mean power frequency versus isometrictorque for the vastus lateralis muscle. Electromyogr Clin Neurophysiol2008a;48:13–25.

Beck TW, Housh TJ, Cramer JT, Malek MH, Mielke M, Hendrix R, et al. Electrode shiftand normalization reduce the innervation zone’s influence on EMG. Med SciSports Exerc 2008b;40:1314–22.

Beck TW, Housh TJ, Cramer JT, Mielke M, Hendrix R. The influence of electrode shiftover the innervation zone and normalization on the electromyographicamplitude and mean power frequency versus isometric torque relationshipsfor the vastus medialis muscle. J Neurosci Methods 2008c;169:100–8.

Beck TW, Housh TJ, Cramer JT, Weir JP. The effects of electrode placement andinnervation zone location on the electromyographic amplitude and meanpower frequency versus isometric torque relationships for the vastus lateralismuscle. J Electromyogr Kinesiol 2008d;18:317–28.

Beck TW, Housh TJ, Cramer JT, Weir JP. The effects of interelectrode distance overthe innervation zone and normalization on the electromyographic amplitudeand mean power frequency versus concentric, eccentric, and isometric torquerelationships for the vastus lateralis muscle. J Electromyogr Kinesiol2009;19:219–31.

Camic CL, Housh TJ, Johnson GO, Hendrix CR, Zuniga JM, Mielke M, et al. An EMGfrequency-based test for estimating the neuromuscular fatigue thresholdduring cycle ergometry. Eur J Appl Physiol 2010a;108:337–45.

Camic CL, Housh TJ, Zuniga JM, Hendrix CR, Mielke M, Johnson GO, et al. Effects ofarginine-based supplements on the physical working capacity at the fatiguethreshold. J Strength Cond Res 2010b;24:1306–12.

Day JR, Rossiter HB, Coats EM, Skasick A, Whipp BJ. The maximally attainable Vo2during exercise in humans: the peak vs. maximum issue. J Appl Physiol2003;95:1901–7.

deVries HA, Housh TJ, Johnson GO, Evans SA, Tharp GD, Housh DJ, et al. Factorsaffecting the estimation of physical working capacity at the fatigue threshold.Ergonomics 1990;33:25–33.

deVries HA, Tichy MW, Housh TJ, Smyth KD, Tichy AM, Housh DJ. A method forestimating physical working capacity at the fatigue threshold. Ergonomics1987;30:1195–204.

EMG 16. . .16 channels surface electromyographic signal amplifier (user’s manual).Turin (Italy): LISiN Bioengineering Center Polytechnic of Turin – Department ofElectronics; 2006.

Farina D, Merletti R, Enoka RM. The extraction of neural strategies from the surfaceEMG. J Appl Physiol 2004a;96:1486–95.

Farina D, Merletti R, Nazzaro M, Caruso I. Effect of joint angle on EMG variables inleg and thigh muscles. IEEE Eng Med Biol Mag 2001;20:62–71.

Farina D, Merletti R, Stegeman DF. Biophysics of the generation of EMG signals.In: Merletti R, Parker PA, editors. Electromyography: physiology,engineering, and noninvasive applications. Hoboken: John Wiley & Sons,Inc.; 2004b. p. 81–105.

Fukunaga T, Ichinose Y, Ito M, Kawakami Y, Fukashiro S. Determination of fasciclelength and pennation in a contracting human muscle in vivo. J Appl Physiol1997;82:354–8.

Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendationsfor SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol2000;10:361–74.

Hermens HJ, Freriks B, Merletti R, Stegeman D, Blok J, Rau G, et al. SENIAMEuropean recommendations for surface electromyography: results of theSENIAM project. Enschede, The Netherlands: Roessingh Research andDevelopment; 1999.

Housh TJ, deVries HA, Johnson GO, Evans SA, Housh DJ, Stout JR, et al.Neuromuscular fatigue thresholds of the vastus lateralis, vastus medialis,and rectus femoris muscles. Electromyogr Clin Neurophysiol1996;36:247–56.

Housh TJ, deVries HA, Johnson GO, Evans SA, McDowell S. The effect of ammoniumchloride and sodium bicarbonate ingestion on the physical working capacity atthe fatigue threshold. Eur J Appl Physiol 1991;62:189–92.

Malek M, Coburn JW, Weir JP, Beck TW, Housh TJ. The effects of innervation zone onelectromyographic amplitude and mean power frequency during incrementalcycle ergometry. J Neurosci Methods 2006;155:126–33.

Mercer JA, Bezodis N, DeLion D, Zachry T, Rubley MD. EMG sensor location: does itinfluence the ability to detect differences in muscle contraction conditions? JElectromyogr Kinesiol 2006;16:198–204.

Merletti R, Farina D, Gazzoni M. The linear electrode array: a useful tool with manyapplications. J Electromyogr Kinesiol 2003;13:37–47.

Merletti R, Farina D, Granata A. Non-invasive assessment of motor unit propertieswith linear electrode arrays. Electroencephalogr Clin Neurophysiol1999;50(Suppl.):293–300.

Mesin L, Merletti R, Rainoldi A. Surface EMG: the issue of electrode location. JElectromyogr Kinesiol 2009;19:719–26.

Rainoldi A, Melchiorri G, Caruso I. A method for positioning electrodes duringsurface EMG recordings in lower limb muscles. J Neurosci Methods2004;134:37–43.

Rainoldi A, Nazzaro M, Merletti R, Farina D, Caruso I, Gaudenti S. Geometricalfactors in surface EMG of the vastus medialis and lateralis muscles. JElectromyogr Kinesiol 2000;10:327–36.

Sacco ICN, Gomes AA, Otuzi ME, Pripas D, Onodera AN. A method for betterpositioning bipolar electrodes for lower limb EMG recordings during dynamiccontractions. J Neurosci Methods 2009;180:133–7.

Shinohara M, Kouzaki M, Yoshihisa T, Fukunaga T. Mechanomyography of thehuman quadriceps muscle during incremental cycle ergometry. Eur J ApplPhysiol Occup Physiol 1997;76:314–9.

Stout J, Eckerson J, Ebersole K, Moore G, Perry S, Housh T, et al. Effect ofcreatine loading on neuromuscular fatigue threshold. J Appl Physiol2000;88:109–12.

Stout JR, Graves BS, Smith AE, Hartman MJ, Cramer JT, Beck TW, et al. The effect ofbeta-alanine supplementation on neuromuscular fatigue in elderly (55–92 years): a double-blind randomized study. J Int Soc Sports Nutr 2008;5:21.

Zuniga JM, Housh TJ, Hendrix CR, Camic CL, Mielke M, Schmidt RJ, et al. The effectsof electrode orientation on electromyographic amplitude and mean powerfrequency during cycle ergometry. J Neurosci Methods 2009;184:256–62.

Clayton L. Camic received a BS (2001) degree in Exer-cise Science from Morehead State University, Morehead,KY and MS (2003) degree in Exercise Physiology fromthe University of Wyoming. He is currently a doctoralstudent in Exercise Physiology at the University ofNebraska-Lincoln and his main research interestsinclude evaluation of muscle function and fatigue usingelectromyography as well as growth and developmentin young athletes.

Terry J. Housh received a BA (1977) degree in PhysicalEducation from Doane College, Crete, NE, and MPE(1979) and Ph.D. (1984) degrees from the University ofNebraska-Lincoln. He is a Fellow of the American Col-lege of Sports Medicine, Fellow in the research Consor-tium of AAHPERD, and received the 1998 OutstandingSport Scientist Award from the National Strength andConditioning Association. Presently, he is a Professor inthe Department of Nutrition and Health Sciences,Director of the Exercise Physiology Laboratory, and Co-Director of the Center for Youth Fitness and SportsResearch at the University of Nebraska-Lincoln. His

main areas of research are muscle function, fatigue, and growth and development inyoung athletes.

C. Russell Hendrix received his B.S.E. (1992) in Biologyfrom Emporia State University, an M.S.E. (2002) inPhysical Education from Northwest Missouri StateUniversity. He is currently a doctoral student in ExercisePhysiology at the University of Nebraska-Lincoln andhis main research interests include evaluation of musclefunction using electromyography and mechanomyog-raphy, especially as they relate to fatigue thresholds. Heis a member of the American College of Sports Medicine,an ACSM-certified Health Fitness Specialist, and amember of the National Strength and ConditioningAssociation.

Jorge M. Zuniga received a BS (2003) degree in PhysicalEducation from ‘‘Cardenal R. Silva Henrriquez’’ CatholicUniversity; Santiago, Chile, and MS (2007) from theUniversity of Nebraska at Omaha. He is a member of theAmerican College of Sports Medicine and the NationalStrength and Conditioning Association. He is currently adoctoral student in Exercise Physiology at the Univer-sity of Nebraska-Lincoln and his main research interestsinclude evaluation of muscle function and fatigue usingelectromyography and mechanomyography duringcycle ergometry, endurance performance, and growthand development in young athletes.

40 C.L. Camic et al. / Journal of Electromyography and Kinesiology 21 (2011) 33–40

Haley C. Bergstrom received a BS (2009) degree inExercise Science from Doane College, Crete, Nebraska.She is currently a master’s student in Exercise Physiol-ogy at the University of Nebraska-Lincoln and her mainresearch interests include evaluation of muscle functionand fatigue using electromyography and mechanomy-ography.

Richard J. Schmidt received a BS (1969), M.Ed. inPhysical Education (1964) and a Ph.D. (1972) in Edu-cation (Psychological and Cultural Studies) from theUniversity Nebraska-Lincoln. He is a Fellow in theresearch Consortium of AAHPERD. He is currently anAssociate Professor of Nutrition and Health Sciences atthe University of Nebraska-Lincoln. His main areas ofresearch are Exercise testing, Clinical Exercise Physiol-ogy, Occupational Physiology, Military & Law Enforce-ment Sports Medicine.

Glen O. Johnson received a BS (1960) and MS (1964)degrees from Winona State University, Winona, MN,and a Ph.D. (1972) from the University of Iowa. He is aFellow of the American College of Sports Medicineand a Fellow in the research Consortium of AAHPERD.He is currently a Professor of Nutrition and HealthSciences at the University of Nebraska-Lincoln.