The minimum number of contractions required to examine the EMG amplitude versus isometric force relationship for the vastus lateralis and vastus medialis

Journal of Electromyography and Kinesiology 24 (2014) 827–834

Contents lists available at ScienceDirect

Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin

The minimum number of contractions required to examine the EMG amplitude versus isometric force relationship for the vastus lateralis and vastus medialis Matt S. Stock ⇑, Alexander S. Drusch, Brennan J. Thompson Muscular Assessment Laboratory, Texas Tech University, Lubbock, TX, USA

a r t i c l e

i n f o

Article history: Received 15 May 2014 Received in revised form 26 September 2014 Accepted 1 October 2014

Keywords: EMG Muscle Motor unit

a b s t r a c t Studies have demonstrated that the electromyographic (EMG) amplitude versus submaximal isometric force relationship is relatively linear. The purpose of this investigation was to determine the minimum number of contractions required to study this relationship. Eighteen men (mean age = 23 years) performed isometric contractions of the leg extensors at 10–90% of the maximum voluntary contraction (MVC) in 10% increments while surface EMG signals were detected from the vastus lateralis and vastus medialis. Linear regression was used to determine the coefficient of determination, slope coefficient, and y-intercept for each muscle and force combination with successively higher levels included in the model (i.e., 10–30%, . . . 10–90% MVC). For the slope coefficients, there was a main effect for force combination (P < .001). The pairwise comparisons showed there was no difference from 10–60% through 10–90% MVC. For the y-intercepts, there were main effects for both muscle (vastus lateralis [4.3 lV RMS] > vastus medialis [ 3.7 lV RMS]; P = .034) and force combination (P < .001), with similar values shown from 10–50% through 10–90% MVC. The linearity of the absolute EMG amplitude versus isometric force relationship for the vastus lateralis and vastus medialis suggests that investigators may exclude high force contractions from their testing protocol. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Among the most frequently studied topics in electromyography (EMG) research is the relationship between signal amplitude and force (Kamen and Gabriel, 2009; Weir et al., 1992). This is due to the fact that the proper analysis of muscle activity is dependent on a strong, positive relationship with force, particularly in clinical settings (De Luca, 1997). For example, the amplitude of the EMG signal could be used in physical therapy clinics to provide biofeedback to patients (Samani et al., 2010), or be utilized to roughly estimate force for individual muscles in situations where a load cell or an isokinetic dynamometer is not available. The most practical application of the relationship between EMG amplitude and force was demonstrated by Moritani and deVries (1979) in their assessment of neural adaptations versus muscle hypertrophy during eight weeks of unilateral strength training. After considering several methodological factors (deVries, 1968) and ultimately ⇑ Corresponding author at: Texas Tech University, Muscular Assessment Laboratory, 3204 Main Street, Box 43011, Room 103C, Lubbock, TX 79409-3011, USA. Tel.: +1 (806)834 1485; fax: +1 (806)742 1688. E-mail address: [email protected] (M.S. Stock). http://dx.doi.org/10.1016/j.jelekin.2014.10.001 1050-6411/Ó 2014 Elsevier Ltd. All rights reserved.

determining that the monopolar EMG amplitude versus isometric force relationship was linear (Moritani and deVries, 1978), these authors created an equation to estimate the percentage of neural versus hypertrophic adaptations that was based upon changes in the slope of the regression line. Decreases in the slope of this relationship were thought to be reflective of a lower number of activated motor units required to achieve a given force level (deVries, 1968). When combined with their skinfold and circumference analyses, Moritani and deVries (1979) used their ‘‘efficiency of electrical activity’’ (EEA) technique to demonstrate that the initial responses during strength training were due to neural adaptations (e.g., increased motor unit synchronization, enhanced maximal firing rates, etc.). Our previous work has focused on methodological considerations regarding the electrical/mechanical activity of skeletal muscle during isokinetic (Stock et al., 2010a,b) and dynamic constant external resistance (Stock et al., 2010c,d,e) testing. More recently, our laboratory examined the EMG amplitude versus concentric and eccentric force relationships for biarticular muscles during squatting (Luera et al., 2014). The central theme of these investigations has been to develop an objective protocol for non-invasively monitoring neural adaptations to dynamic movements. In theory,

828

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

high linearity and reliability for the EMG amplitude versus dynamic force/torque relationship would allow researchers to use similar, or even identical, training and testing methods, thereby eliminating concerns regarding assessment specificity. Unfortunately, however, these studies (Stock et al., 2010a,b,c,d,e; Luera et al., 2014) have generally shown that these patterns demonstrated only moderate reliability, and were not as linear as those for isometric force assessments (deVries, 1968; Moritani and deVries, 1978, 1979; Weir et al., 1992). Therefore, until additional studies are performed, it appears that the isometric protocol utilized by Moritani and deVries (1979) is the most useful means of assessing neuromuscular adaptations to various exercise interventions. In our experience, testing subjects for the EMG amplitude versus force/torque relationship can be somewhat laborious and time consuming. In addition, these procedures have the potential to be influenced by muscle fatigue, and isometric contractions at high forces (e.g., 90% of the maximum voluntary contraction [MVC]) are difficult for subjects to perform. Furthermore, since the assessment of muscle function for this relationship is based on its regression equation, and not the actual data points, it is currently not clear what the minimum number of contractions required is. The idea that only a few contractions may be necessary was discussed by deVries (1968, pg. 12) who stated ‘‘Since the regression of EMG voltage–time integral upon force of contraction is linear, this approach would have the added advantage of requiring only enough submaximal isometric contractions to establish the slope of the regression lines.’’ Interestingly, some investigators that have previously studied the EMG amplitude versus isometric force relationship have not utilized 10% increments up to 90% MVC. The highest force studied by Lawrence and De Luca (1983) was 80% MVC. More recently, in their assessment of mechanomyographic and EMG amplitude, Beck et al. (2004) and Coburn et al. (2005) utilized 20% increments from 20% to 80% of peak torque. If the linearity and regression equation for the EMG amplitude versus isometric force relationships are similar for 10–90% MVC versus only a few submaximal contractions (e.g., 10–30% MVC), this would save investigators time and also eliminate the need for constant-force testing at values close to the MVC. Thus, the purpose of this investigation was to determine the minimum number of contractions required to study the absolute EMG amplitude versus isometric force relationship for the vastus lateralis and vastus medialis. These muscles were selected for analysis because they cross one joint, are important in rehabilitation of the knee, and are large enough to not be highly affected by crosstalk when a bipolar electrode configuration is utilized (Basmajian and De Luca, 1985). We hypothesized that each of the relationships would show high degrees of linearity, and that there would be no mean differences for the regression coefficients among the force combinations.

consent form prior to testing. This investigation was conducted during the Spring 2014 semester. 2.2. Familiarization In attempt to minimize the influence of learning on each of the study’s variables, the subjects participated in a thorough familiarization session on a separate day prior to data collection. During the familiarization session, the subjects became comfortable performing MVCs of the right leg extensors. In addition, the subjects practiced performing submaximal isometric contractions at each tenth percentile of the MVC (i.e., 10%, 20%, 30%, etc.). The subjects were required to perform one contraction at each submaximal force level, but additional attempts were required if the subject or investigator felt that further practice was necessary. EMG signals were not detected during the familiarization session. 2.3. Isometric force testing A minimum of 24 h following the familiarization session, the subjects returned to the laboratory for data collection. Upon arrival, the subjects were seated in a chair designed specifically for isometric force testing of the right leg extensors. Restraining straps were secured around the hips, abdomen, and chest, and the subjects were asked to remain seated throughout testing. A Velcro cuff was secured around the right ankle, which was attached to a calibrated tension/compression load cell (Model SSM-AJ-500; Interface, Scottsdale, AZ) to allow for the measurement of isometric force (Fig. 1). All testing occurred at a joint angle of 60° below the horizontal plane. Prior to testing, the subjects performed a brief warm-up consisting of three, ten-second contractions at a force level corresponding to 50% of their perceived maximum. The subjects then performed two, five-second MVCs separated by two minutes of rest. The highest value from the two attempts was designated as the MVC. The subjects then performed a series of randomly ordered contractions corresponding to each tenth percentile of the MVC. To do so, the subjects performed trapezoidal isometric contractions in accordance with a visual template on a computer monitor, which was placed directly in front of them. Each template involved a linear increase in force (10%/s), a foursecond constant-force region, and a linear decrease in force (10%/ s). The subjects were asked to maintain their force output as close

2. Methods 2.1. Subjects Eighteen healthy men (mean ± SD age = 23 ± 2 years; height = 178.9 ± 7.3 cm; body mass = 86.1 ± 12.4 kg) volunteered to participate in this investigation. All subjects were healthy and active participants in some recreational activity, but none were competitive athletes. All subjects completed a health and exercise status questionnaire, which indicated no current or recent neuromuscular or musculoskeletal disorders (within the past six months). The subjects were asked to not participate in lower-body exercise 24–48 h prior to the data collection trial. The study was approved by the University Institutional Review Board for Human Subjects, and all subjects read, understood, and signed an informed

Fig. 1. An example of one subject seated in the isometric force testing chair, as well as the sensor placement over the vastus lateralis and the reference electrode over the patella (vastus medialis not displayed).

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

to the visual template as possible. The submaximal contractions were separated by two minutes of rest. The mean value of each four-second constant-force region was calculated and used for the regression analyses. Eighteen subjects originally participated in this study, but four subjects were physically unable to maintain a mean 90% MVC force value constant for four seconds. Thus, these four subjects were removed from subsequent data analyses. For the remaining fourteen subjects, the mean coefficient of determination for the relationship between the recorded isometric force versus the required % MVC was r2 = 0.999.

2.4. Surface EMG signal detection and processing Bipolar surface EMG signals were detected from the vastus lateralis and vastus medialis throughout testing. The signals were detected with two separate DE 2.1 sensors (interelectrode distance = 10 mm [Delsys, Inc., Boston, MA]) and amplified (gain = 1000) by a Bagnoli 16-channel Desktop system (Delsys, Inc.) with a band pass of 20–450 Hz. The sensors were placed over the muscles in accordance with the recommendations following the Surface EMG for the Non-Invasive Assessment of Muscles project (Hermens et al., 2000). Specifically, for the vastus lateralis, the sensor was oriented in the direction of the muscle fibers, and was placed at 2/3 on the line from the anterior superior iliac spine to the lateral aspect of the patella. For the vastus medialis, the sensor was placed at 80% of the line between the anterior superior iliac spine and the anterior border of the medial ligament. A reference electrode was placed over the patella. The skin over the patella and the belly of the vastus lateralis and vastus medialis was prepared before testing by shaving and cleansing with rubbing alcohol. The EMG signals were digitized at a sampling rate of 2000 Hz and stored in a personal computer (Dell Optiplex 755, Round Rock, TX) for subsequent analyses. Prior to data collection for each subject, signal quality was verified for a 20% MVC contraction to ensure low baseline noise, minimal line interface, and a signal-to-noise ratio >2.0. Example force and EMG signals for one subject have been displayed in Fig. 2. At the conclusion of the study, the signals were processed with EMGworks software (version 4.0.13, Delsys, Inc., Boston, MA). For

829

each contraction, a four-second period that corresponded to the constant-force portion was selected for analysis. The root-meansquare (lV RMS) value of each selected signal was then calculated as a measure of EMG amplitude. 2.5. Statistical analysis For each subject and muscle, linear regression analyses were used to determine the coefficient of determination (r2), slope coefficient (lV RMS/N), and y-intercept (lV RMS) for the EMG amplitude versus isometric force relationship for 10–30% MVC, as well as each combination with a successively higher level included in the model (i.e., 10–40%, 10–50% . . . 10–90% MVC). Two separate two-way (muscle [vastus lateralis, vastus medialis]  force combination [10–30% . . . 10–90% MVC]) repeated measures analyses of variance (ANOVAs) were used to examine mean differences for the linear slope coefficients and y-intercepts. In the event of a significant two-way interaction, the data were decomposed with two separate repeated measures ANOVAs (across each muscle) and seven separate dependent samples t-tests (between the muscles for each force combination)(Vincent and Weir, 2012). In the event of a significant main effect, Bonferroni adjustments were used to make marginal mean pairwise comparisons. The partial eta squared statistic (g9 2) was used to examine the effect size for each ANOVA and main effect. According to Stevens (2007), values of 0.01, 0.06, and 0.14 correspond to small, medium, and large effect sizes, respectively. Cohen’s d statistics were used to examine the effect size for mean differences, and values of 0.20, 0.50, and 0.80 were considered small, medium, and large, respectively (Cohen, 1988). An alpha level of p < 0.05 was considered statistically significant for all comparisons. SPSS software (version 21.0, IBM Corporation, Armonk, NY) was used for all analyses. 3. Results 3.1. Linearity Table 1 displays the mean and variability data for the coefficients of determination for both muscles and each force

Fig. 2. An example isometric force tracing (top), as well as the corresponding surface electromyographic signals for the vastus lateralis (middle) and vastus medialis (bottom), for one subject during a 50% MVC assessment. The vertical lines at eight and twelve seconds represent the portion of the contraction where the subject attempted to maintain a steady force. All statistical analyses were performed on this four-second portion of the contraction.

830

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

Table 1 Means, standard deviations (SDs), coefficients of variation (CVs), and ranges for the coefficients of determination (r2) for the electromyographic amplitude versus isometric force relationship for each force combination and muscle. 10–30%

10–40%

10–50%

10–60%

10–70%

10–80%

10–90%

Vastus lateralis Mean SD CV (%) Range

0.953 0.051 5.4 0.824–0.999

0.956 0.042 4.4 0.837–0.993

0.969 0.024 2.5 0.915–0.996

0.971 0.026 2.7 0.902–0.997

0.970 0.031 3.2 0.933–0.995

0.966 0.029 3.0 0.907–0.996

0.955 0.036 3.8 0.895–0.992

Vastus medialis Mean SD CV (%) Range

0.906 0.244 26.9 0.063–0.999

0.950 0.090 9.5 0.644–0.997

0.959 0.053 5.5 0.787–0.994

0.952 0.081 8.5 0.678–0.995

0.950 0.091 9.6 0.637–0.997

0.942 0.087 9.3 0.659–0.992

0.942 0.063 6.7 0.746–0.995

combination. Table 2 displays the mean and variability data for the absolute EMG amplitude and mean isometric force values in this study. All force combinations exhibited mean coefficients of determination >.900. The 10–30% MVC force combination for the vastus medialis exhibited the greatest variability, although this was likely due to one subject that exhibited a very poor relationship (r2 = 0.063). None of the relationships in this investigation were negative. Example regression lines for one subject have been displayed in Fig. 3.

vastus lateralis (4.3 lV RMS) was significantly greater than that for the vastus medialis ( 3.7 lV RMS). For force combination, when collapsed across the vastus lateralis and vastus medialis, the results from the marginal mean pairwise comparisons were as follows: (1) 10–30% < 10–80% and 10–90%, and (2) 10–40% < 10–60% – 10–90%. Table 4 displays the mean difference (lV RMS) and effect size for each possible force combination comparison. 4. Discussion

3.2. Linear slope coefficients The results from the two-way repeated measures ANOVA indicated that there was no significant muscle  force combination interaction (p = .428, g9 2 = .065; observed power = .200 [Fig. 4a]). There was also no main effect for muscle (p = .165, g9 2 = .143, observed power = .276). There was, however, a main effect for force combination (p < .001, g9 2 = .536, observed power = .999). When collapsed across the vastus lateralis and vastus medialis, the results from the marginal mean pairwise comparisons were as follows: (1) 10–30% and 10–40% < 10–70%, 10–80%, and 10–90%, and (2) 10–50% < 10–90%. Table 3 displays the mean difference (lV RMS/N) and effect size for each possible force combination comparison. 3.3. Y-intercepts The results from the two-way repeated measures ANOVA indicated that there was no significant muscle  force combination interaction (p = .771, g9 2 = .020, observed power = .088 [Fig. 4b]). There were, however, main effects for both muscle (p = .034, g9 = .302, observed power = .593) and force combination (p < .001, g9 2 = .494, observed power = .991). For muscle, when collapsed across the seven force combinations, the marginal mean for the

The purpose of this study was to determine the minimum number of contractions required to study the EMG amplitude versus isometric force relationship for both the vastus lateralis and vastus medialis. Our findings indicated that for both muscles, the 10–90% MVC relationships could be considered linear, with coefficients of determination that were generally greater than 0.900 (Table 1). Based on the results from our ANOVAs, as well as our evaluation of the effect size for each force combination comparison, the removal of each force level had a small influence on a relationship’s regression equation, with higher and lower linear slope coefficients and y-intercepts, respectively, demonstrated for models with additional contractions included. With each force level that was removed, the mean difference became more robust, eventually leading to statistically significant and meaningful discrepancies in the linear slope coefficients and y-intercepts. Collectively, these findings demonstrated that future investigators wishing to study these relationships could conceivably eliminate a few high force contractions (e.g., 70–90% MVC) from their testing protocol and still anticipate similar statistics. Caution should be taken, however, before a minimalist approach is adopted. For example, for both muscles, the mean difference for the y-intercepts for the 10–30% versus 10–90% MVC combination resulted in an effect size greater than 1.0 (Table 4).

Table 2 Means, standard deviations (SDs), and coefficients of variation (CVs) for the absolute electromyographic (EMG) amplitude and mean force values at each percentage of the maximum voluntary contraction (MVC) that was assessed in this study. Note that all three variables exhibited considerable variability. % MVC

10%

20%

Vastus lateralis absolute EMG Amplitude (lV RMS) Mean 35.5 58.6 SD 18.6 33.8 CV (%) 52.4 57.7 Vastus medialis absolute EMG Amplitude (lV RMS) Mean 23.3 39.4 SD 11.9 21.8 CV (%) 51.1 55.3 Mean force (N) Mean SD CV (%)

106.5 41.3 38.8

203.1 71.0 35.0

30%

40%

50%

60%

70%

80%

90%

81.7 51.5 63.0

107.2 57.8 53.9

135.3 79.7 58.9

165.1 89.6 54.3

196.9 101.2 51.4

233.4 123.9 53.1

285.6 158.9 55.6

56.1 31.1 55.4

78.9 38.8 49.2

103.3 49.6 48.1

129.2 50.6 39.2

152.8 63.2 41.4

184.9 72.2 39.1

228.5 91.4 40.0

297.8 100.0 33.6

398.2 128.6 32.3

494.8 161.1 32.6

596.5 194.6 32.6

693.5 223.6 32.2

786.1 248.6 31.6

877.1 268.8 30.6

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

831

Fig. 3. Example surface electromyographic amplitude versus isometric force relationships for the (a) vastus lateralis and (b) vastus medialis for one subject. Each graph displays the regression lines corresponding to each force combination. The regression equation and coefficient of determination for each relationship have been displayed to the right below each legend. Note the higher and lower linear slope coefficients and y-intercepts, respectively, for the combinations with greater force levels included in the model. The values displayed here are similar to the marginal mean values displayed in Fig. 4.

The EMG amplitude versus isometric force relationship has been of interest to neuromuscular researchers for decades (Basmajian and De Luca, 1985). The notion of using this relationship to study muscle function was originally described by Fischer and Merhautova (1961) and deVries (1968). deVries (1968) hypothesized that as a muscle became stronger, less motor unit activity (i.e., decreased recruitment and/or firing rates) was required to produce a given absolute force level. This author (deVries, 1968) demonstrated the value of monitoring the linear slope coefficient and y-intercept for this relationship by assessing their responses during strength training, disuse atrophy, and cross-education. On the basis that the specific contributions of neural versus hypertrophic adaptations to strength training could be estimated, these procedures were used by Moritani and deVries (1979) to establish the specific time course of traininginduced changes. Interestingly, both deVries (1968) and Weir et al. (1992) expressed concerns regarding changes in posture and coactivation associated with having subjects perform constant-force isometric contractions at high forces. We speculate that these concerns may partially explain why some authors (Beck et al., 2004; Coburn et al., 2005; Lawrence and De Luca, 1983) have avoided 90% MVC assessments. Our results showed that the linear slope coefficients and y-intercepts showed small/moderate differences among the 10–70%, 10–80%, and 10–90% MVC protocols, and the patterns were reasonably linear. Therefore, if investigators are apprehensive about the time commitment associated with a

large training study involving the EEA technique, or their subjects’ ability to achieve these high force levels, a given testing protocol could be modified. The fact that four of the subjects in the present study were unable to perform the 90% MVC assessment further supports this suggestion. A unique finding of this study was the significant main effect for the y-intercept for muscle, with the vastus lateralis demonstrating higher values than those for the vastus medialis. As there was no difference between the muscles for the linear slope coefficients, this is likely related to the higher absolute EMG amplitude values for the vastus lateralis. Interestingly, previous authors that have studied the EMG amplitude versus force/torque relationship have focused their attention on the linear slope coefficient more so than the y-intercept. deVries (1968) postulated that the y-intercept of this relationship could be useful for studying muscle relaxation/ tension, since it theoretically represents the degree of motor unit activity corresponding to a force level of zero. In a subset of his data, deVries (1968) reported that six out of 22 subjects demonstrated y-intercepts at zero, and the mean was 23.8 lV. In addition, the correlation between maximal strength and the y-intercept was .102, suggesting a poor relationship between force production and the theoretical estimation of muscle activity at a force level of zero. On the basis of the present study’s main effect, researchers should not assume that the y-intercept for the EMG amplitude versus isometric force relationship will always be exactly zero. While the muscle-specific difference for the y-intercept is intriguing, it

832

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

Fig. 4. Marginal means (collapsed across the vastus lateralis and vastus medialis) ± standard errors of the mean for the (a) linear slope coefficients and (b) y-intercepts for the electromyographic amplitude versus isometric force relationships. ⁄ = significantly greater than 10–30% and 10–40% MVC;   = significantly greater than 10–30%, 10–40%, and 10–50% MVC; à = significantly less than 10–40% MVC;  = significantly less than 10–30% and 10–40% MVC.

was likely a reflection of the higher EMG amplitude values for the vastus lateralis, and its meaning or practical application is not entirely clear. Like all studies, the present investigation was not without limitations, and future researchers should not blindly apply these

findings to all scenarios. Specifically, there are four issues that are worthy of discussion. First, we should note that our findings are specific to the vastus lateralis and vastus medialis during isometric force testing, and may not be applicable to other muscles. Both Lawrence and De Luca (1983) and Weir et al. (1992) compared the linearity of the EMG amplitude versus isometric force relationship for different muscles, and in each study, subtle discrepancies were noted. In the investigation by Lawrence and De Luca (1983), differences were noted among the first dorsal interosseous, deltoid, and biceps brachii, and the authors suggested that the motor control scheme that a muscle relies on to produce force (i.e., high firing rates versus recruitment thresholds close to the MVC) plays an important role in determining the linearity of these patterns. Furthermore, investigators should exercise caution when applying the findings of this study to biarticular muscles and/or dynamic movements, as our recent work has demonstrated differences in the linear slope coefficients for the biceps femoris during concentric versus eccentric squatting (Luera et al., 2014). A second issue worthy of consideration is the excessive amount of betweensubjects variance for both EMG amplitude and isometric force. As displayed in Table 2, all of the coefficients of variation for the absolute EMG amplitude and mean isometric force values were greater than 30.0%, with many approaching 60.0%. For this reason, EMG researchers are often encouraged to normalize data to the values corresponding to an MVC in an effort to minimize noise (Yang and Winter, 1983; Burden, 2010). In this study, absolute values were analyzed solely because the applicability of the EEA technique is based on changes in absolute EMG amplitude values at given force levels during or following an intervention (deVries, 1968; Moritani and deVries, 1979). In addition, the EEA technique is based on the regression coefficients for individual subjects, and not group means (deVries, 1968; Moritani and deVries, 1979). If investigators are interested in group mean analyses, however, normalization is likely to reduce variability of the linear slope coefficients and y-intercepts, and therefore, improve statistical power. A third topic worthy of attention is the fact that for some of the subjects in this investigation, considerable variability in the

Table 3 Mean differences (lV RMS/N) and effect sizes for the linear slope coefficients for the electromyographic amplitude versus isometric force relationships for each force combination comparison. Note that the more force levels that are added/removed, the larger the change in the effect size. Larger effect sizes were generally shown for the vastus medialis. Differences that are considered moderate or large have been displayed in bold font. Vastus lateralis

Vastus medialis

Statistical comparison

Mean difference

Cohen’s d

Statistical Comparison

Mean difference

Cohen’s d

10–30%

10–40% 10–50% 10–60% 10–70% 10–80% 10–90%

0.003 0.007 0.020 0.031 0.040 0.056

0.033 0.082 0.227 0.346 0.446 0.584

10–30%

10–40% 10–50% 10–60% 10–70% 10–80% 10–90%

0.022 0.040 0.060 0.063 0.076 0.094

0.351 0.519 0.741 0.802 0.877 1.006

10–40%

10–50% 10–60% 10–70% 10–80% 10–90%

0.010 0.023 0.034 0.043 0.059

0.121 0.282 0.414 0.526 0.669

10–40%

10–50% 10–60% 10–70% 10–80% 10–90%

0.018 0.037 0.041 0.053 0.072

0.249 0.504 0.569 0.674 0.838

10–50%

10–60% 10–70% 10–80% 10–90%

0.013 0.024 0.033 0.049

0.145 0.270 0.376 0.530

10–50%

10–60% 10–70% 10–80% 10–90%

0.019 0.023 0.035 0.054

0.229 0.282 0.404 0.583

10–60%

10–70% 10–80% 10–90%

0.011 0.020 0.036

0.134 0.247 0.420

10–60%

10–70% 10–80% 10–90%

0.004 0.016 0.035

0.048 0.189 0.390

10–70%

10–80% 10–90%

0.009 0.025

0.112 0.295

10–70%

10–80% 10–90%

0.012 0.031

0.147 0.355

10–80%

10–90%

0.016

0.192

10–80%

10–90%

0.019

0.208

833

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

Table 4 Mean differences (lV RMS) and effect sizes for the y-intercepts for the electromyographic amplitude versus isometric force relationships for each force combination comparison. Note that the more force levels that are added/removed, the larger the change in the effect size. Differences that are considered moderate or large have been displayed in bold font. Vastus lateralis

Vastus medialis

Statistical comparison

Mean difference

Cohen’s d

Statistical comparison

Mean difference

Cohen’s d

10–30%

10–40% 10–50% 10–60% 10–70% 10–80% 10–90%

2.7 4.3 7.0 8.9 11.6 17.3

0.221 0.300 0.519 0.686 0.880 1.223

10–30%

10–40% 10–50% 10–60% 10–70% 10–80% 10–90%

2.7 5.4 8.4 9.5 12.4 21.4

0.284 0.463 0.719 0.816 0.830 1.176

10–40%

10–50% 10–60% 10–70% 10–80% 10–90%

1.6 4.3 6.2 8.9 14.6

0.110 0.308 0.464 0.660 1.022

10–40%

10–50% 10–60% 10–70% 10–80% 10–90%

2.6 5.7 6.8 9.6 18.7

0.217 0.469 0.564 0.640 1.034

10–50%

10–60% 10–70% 10–80% 10–90%

2.6 4.6 7.3 12.9

0.167 0.300 0.476 0.818

10–50%

10–60% 10–70% 10–80% 10–90%

3.1 4.1 7.0 16.1

0.229 0.314 0.441 0.867

10–60%

10–70% 10–80% 10–90%

2.0 4.7 10.3

0.141 0.337 0.730

10–60%

10–70% 10–80% 10–90%

1.1 3.9 13.0

0.085 0.257 0.733

10–70%

10–80% 10–90%

2.7 8.3

0.206 0.635

10–70%

10–80% 10–90%

2.9 11.9

0.191 0.686

10–80%

10–90%

5.7

0.454

10–80%

10–90%

9.1

0.477

absolute EMG amplitude values was observed at high force levels. As this finding was not entirely consistent, the mean values displayed in Table 2 are not completely representative of this phenomenon. When subjects such as these are included within group mean analyses, some degree of systematic error may be expected if data points at high contraction levels are not included within a given statistical model. Although the reasoning for this high force variability among subjects is not clear, we speculate that differences among the contributions of the rectus femoris and vastus intermedius, as well as the antagonist musculature (Carolan and Cafarelli, 1992), may be responsible. Lastly, as demonstrated by Keenan et al. (2005) and Keenan and Valero-Cuevas (2007) and reviewed by Farina et al. (2010), the use of surface EMG to analyze neural mechanisms is not devoid of limitations. Using computer-simulated data, Keenan et al. (2005) showed that the interpretation of absolute EMG values was influenced by large degrees of amplitude cancellation. While cancelation was reduced with normalization, there was still a loss of signal information. Collectively, the idea of removing contractions from a given testing protocol in order to obtain a linear slope coefficient and y-intercept should be carefully considered within the context of a given study’s purpose. Our findings are particularly relevant to absolute EMG amplitude values for the vastus lateralis and vastus medialis during isometric testing, and other muscles responsible for extension at the knee joint were not examined. In summary, the results of this study demonstrated similar linear slope coefficients and y-intercepts for the EMG amplitude versus isometric force relationship for 10–90% MVC testing protocols versus those with high force levels removed from the model. For both variables, there were no mean differences from 10–60% MVC through 10–90% MVC. We also observed a marginal mean difference for the y-intercept for the vastus lateralis versus vastus medialis, which was likely a statistical outcome associated with higher absolute EMG amplitude values for the vastus lateralis. If the procedures utilized in the present study are precisely followed, future investigators that wish to study muscle function with the use of this relationship (e.g., EEA) can utilize force levels corre-

sponding to 10–60% MVC and expect to observe a similar pattern as that shown for 10–90% MVC. Doing so may reduce time and effort associated with data collection, minimize the potential for muscle fatigue, and alleviate concerns about postural changes and excessive coactivation during constant-force contractions at levels close to the MVC.

Conflicts of interest The authors declare no conflicts of interest. References Basmajian JV, De Luca CJ. Muscles alive. 5th ed. Baltimore: Williams & Wilkins; 1985. Beck TW, Housh TJ, Johnson GO, Weir JP, Cramer JT, Coburn JW, et al. Mechanomyographic and electromyographic time and frequency domain responses during submaximal to maximal isokinetic muscle actions of the biceps brachii. Eur J Appl Physiol 2004;92(3):352–9. Burden A. How should we normalize electromyograms obtained from healthy participants? What we have learned from over 25 years of research. J Electromyogr Kinesiol 2010;20(6):1023–35. Carolan B, Cafarelli E. Adaptations in coactivation after isometric resistance training. J Appl Physiol 1992;73(3):911–7. Coburn JW, Housh TJ, Cramer JT, Weir JP, Miller JM, Beck TW, et al. Mechanomyographic and electromyographic responses of the vastus medialis muscle during isometric and concentric muscle actions. J Strength Cond Res 2005;19(2):412–20. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale: Lawrence Erlbaum Associates Publishers; 1988. De Luca CJ. The use of surface electromyography in biomechanics. J Appl Biomech 1997;13(2):135–63. deVries HA. Efficiency of electrical activity as a physiological measure of the functional state of muscle tissue. Am J Phys Med 1968;47(1):10–22. Farina D, Holobar A, Merletti R, Enoka RM. Decoding the neural drive to muscles from the surface electromyogram. Clin Neurophysiol 2010;121(10):1616–23. Fischer A, Merhautova J. Electromyographic manifestations of individual stages of adopted sports technique. In: Health and fitness in modern world. Chicago, The Athletic Institute, Library of Congress, 1961 [chapter 13]. Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 2000;10(5):361–74. Kamen G, Gabriel DA. Essentials of electromyography. Champaign: Human Kinetics; 2009.

834

M.S. Stock et al. / Journal of Electromyography and Kinesiology 24 (2014) 827–834

Keenan KG, Valero-Cuevas FJ. Experimentally valid predictions of muscle force and EMG in models of motor-unit function are most sensitive to neural properties. J Neurophysiol 2007;98(3):1581–90. Keenan KG, Farina D, Maluf KS, Merletti R, Enoka RM. Influence of amplitude cancellation on the simulated surface electromyogram. J Appl Physiol 2005;98(1):120–31. Lawrence JH, De Luca CJ. Myoelectric signal versus force relationship in different human muscles. J Appl Physiol Respir Environ Exerc Physiol 1983;54(6): 1653–9. Luera MJ, Stock MS, Chappell AD. Electromyographic amplitude vs. concentric and eccentric squat force relationships for monoarticular and biarticular thigh muscles. J Strength Cond Res 2014;28(2):328–38. Moritani T, deVries HA. Reexamination of the relationship between the surface integrated electromyogram (IEMG) and force of isometric contraction. Am J Phys Med 1978;57(6):263–77. Moritani T, deVries HA. Neural factors versus hypertrophy in the time course of muscle strength gain. Am J Phys Med 1979;58(3):115–30. Samani A, Holtermann A, Søgaard K, Madeleine P. Advanced biofeedback from surface electromyography signals using fuzzy system. Med Biol Eng Comput 2010;48(9):865–73. Stevens JP. Intermediate Statistics. 3rd ed. New York: Lawrence Erlbaum Associates; 2007. Stock MS, Beck TW, Defreitas JM, Dillon MA. Linearity and reliability of the mechanomyographic amplitude versus dynamic torque relationships for the superficial quadriceps femoris muscles. Muscle Nerve 2010a;41(3): 342–9. Stock MS, Beck TW, Defreitas JM, Dillon MA. Linearity and reliability of the EMG amplitude versus dynamic torque relationships for the superficial quadriceps femoris muscles. Electromyogr Clin Neurophysiol 2010b;50(2): 97–106. Stock MS, Beck TW, Defreitas JM, Dillon MA. An examination of the linearity and reliability of the electromyographic amplitude versus dynamic constant external resistance relationships using monopolar and bipolar recording methods. J Neurosci Methods 2010c;194(1):94–101. Stock MS, Beck TW, Defreitas JM, Dillon MA. Linearity and reliability of the mechanomyographic amplitude versus dynamic constant external resistance relationships for the biceps brachii. Physiol Meas 2010d; 31(11):1487–98. Stock MS, Beck TW, Defreitas JM, Dillon MA. Linearity and reliability of the mechanomyographic amplitude versus concentric dynamic constant external resistance relationships for the bench press exercise. J Strength Cond Res 2010e;24(3):785–95. Vincent WJ, Weir JP. Statistics in kinesiology. 4th ed. Champaign: Human Kinetics; 2012. Weir JP, Wagner LL, Housh TJ. Linearity and reliability of the IEMG v torque relationship for the forearm flexors and leg extensors. Am J Phys Med Rehabil 1992;71(5):283–7. Yang JF, Winter DA. Electromyography reliability in maximal and submaximal isometric contractions. Arch Phys Med Rehabil 1983;64(9):417–20.

Matt S. Stock is an Assistant Professor at Texas Tech University where he co-directs research within the Human Performance and Muscular Assessment Laboratories. Dr. Stock completed his Ph.D. at the University of Oklahoma in 2012. His research interests include adaptations to strength training, muscle fatigue, and the firing characteristics of individual motor units.

Alex Drusch is a Graduate Assistant at Texas Tech University in the Department of Health and Exercise Sports Sciences, where he also teaches courses dealing with diet and exercise. His research has focused on the areas of electromyography and ultrasound on neuromuscular adaptations and hypertrophic growth, respectively. Alex received an undergraduate degree from Oklahoma State University in Nutrition with emphasis in Allied Health.

Brennan J. Thompson is an Assistant Professor in the Department of Health, Exercise, and Sport Sciences at Texas Tech University. Dr. Thompson received his Bachelors and Master’s degree in Exercise Science from Weber State University and Utah State University, respectively. He completed his Ph.D. from Oklahoma State University in Applied Exercise Physiology. His research interests examine the noninvasive assessment of neuromuscular function across the lifespan, and in athletic and clinical populations.