The relationship between signal amplitude and force is one of the most frequently studied topics in electromyographic (EMG) research (1,10,11,17,20,21,27). This is largely related to the desire among researchers and clinicians to assess the relative effort of individual muscles during a contraction. For example, a physical therapist might be interested in estimating the contributions of the vastus lateralis, rectus femoris, and vastus medialis to single-joint force production following anterior cruciate ligament reconstruction in his or her athletes. Unfortunately, the EMG amplitude vs. force relationship is complex, and it is influenced by both biological and technical factors (3). In a detailed study that controlled for many of the methodological factors that affect the signal, Lawrence and De Luca (20) showed linear EMG amplitude versus isometric force relationships for the first dorsal interosseous, but not for the biceps brachii or deltoid. Thus, they (20) concluded that the pattern of response for the EMG amplitude vs. force relationship is affected by the primary control scheme of the muscle (i.e., high firing rates vs. recruitment thresholds close to the maximum voluntary contraction). Furthermore, Fischer and Merhautova (11) and deVries (10) demonstrated enhanced linearity for this relationship when a monopolar electrode configuration was used, suggesting that the instrumentation used to detect the EMG signal is significant.
The data obtained from the EMG amplitude vs. force relationship have practical applications for strength and conditioning researchers. Specifically, the linear slope coefficient for this relationship was used by deVries (10) and Moritani and deVries (22) to study disuse atrophy, cross-education, and the specific time course of adaptations to training-induced strength gains. Using what they called the “efficiency of electrical activity” technique, Moritani and deVries (22) reported that the early increases in muscular strength were primarily the result of neural factors (e.g., increased motor unit synchronization, decreased antagonist activity, changes in motor unit firing rates, etc.), with hypertrophy becoming evident at 3–5 weeks into the training program. It is important to note that this conclusion was based on the analysis of monopolar EMG signals from the biceps brachii during isometric force testing, but the training program consisted of dumbbell curls using moderate (∼66% of the 1 repetition maximum [1RM]) loads.
Although Basmajian and De Luca (3) stated that the analysis of EMG signals during conditions involving changes in muscle length should be avoided, isometric force testing has somewhat limited practicality, and previous authors (3,6,25,26) have shown fairly linear EMG amplitude vs. dynamic torque relationships. Stock et al. (26) examined both the linearity and test-retest reliability of the EMG amplitude vs. concentric torque relationships for the vastus lateralis, rectus femoris, and vastus medialis. The data were examined using group mean procedures and also on an individual subject basis. It was reported that although the relationships were fairly linear, they were likely not reliable enough to be used for assessing the neural versus hypertrophic contributions to training-induced strength gains. This investigation was followed by a similar one that compared monopolar vs. bipolar recording methods during concentric and eccentric dynamic constant external resistance muscle actions of the dominant forearm flexors (25). In that study, the subjects lifted and lowered dumbbells in a controlled manner using loads that corresponded to 10–100% of the 1RM. The results indicated that there were moderate degrees of linearity for both concentric and eccentric muscle actions, and similar patterns were shown for the 2 recording methods. Thus, the EMG amplitude vs. force-torque relationship has been assessed during isometric (10,20,27), isokinetic (4,6,26), and dynamic constant external resistance (25) strength testing conditions for a variety of muscles.
A recent development in strength training technology allows subjects to perform common multiple-joint exercises while movement duration is controlled. Termed “isovelocity” training by the manufacturer, the duration for both concentric and eccentric force testing, as well as the range of motion, can be predetermined by the tester. The device is motor driven, data are collected via a load cell, and the peak and average force values are reported for each repetition. Once maximum concentric and eccentric force tests have been performed, a software program generates force curves, and visual feedback can be provided to the subject regarding his or her ability to reach or maintain the desired output. To our understanding, this technology has been used in only 2 previous investigations (16,23). Ratamess et al. (23) examined the effects of 6 weeks of chest press and seated row training. Twenty-four subjects were randomized into training and control groups. The training group visited the laboratory twice per week and performed 5 sets of 6–10 repetitions for both exercises using 3-second concentric and eccentric muscle actions. The subjects were given visual feedback and were asked to maintain force outputs corresponding to 75–85% of their maximum concentric and eccentric force values. The results indicated that 6 weeks of isovelocity training enhanced 1RM strength for the barbell bench press and bent-over row, as well as push-up endurance. The authors (23) concluded that the improvements in force production obtained via isovelocity training were in fact transferable to barbell exercises and push-up performance. Hoffman et al. (16) used maximum concentric and eccentric forces during multiple-joint testing as dependent variables in an investigation involving a nutritional intervention. Neither Ratamess et al. (23) nor Hoffman et al. (16) reported data concerning the electrical activity of the muscles involved in isovelocity training.
Although previous investigators have assessed EMG amplitude values during the squat (5), none of them have been able to control the movement duration during concentric and eccentric testings. Furthermore, many of these investigators have examined the factors that influence the kinematics of the squat (e.g., barbell location over the trapezius, foot placement, squat depth) but not the patterns of response for EMG amplitude vs. submaximal concentric and eccentric forces. Thus, the purpose of this study was to examine the linearity of the EMG amplitude vs. submaximal concentric and eccentric squat force relationships for the vastus lateralis, rectus femoris, and biceps femoris. These 3 muscles were selected for analysis because of their unique anatomical roles during multiple-joint testing (2,12,28). Based on the results from the previous work of one of the authors (25,26), we hypothesized that these relationships would demonstrate moderate-to-high degrees of linearity, especially for the vastus lateralis. In the event that strong relationships are shown, future studies in our laboratory may examine the test-retest reliability for these patterns and possibly their ability to track changes in neuromuscular adaptations with strength training.
Experimental Approach to the Problem
This investigation used a within-subjects research design. Forty-eight hours after a thorough familiarization session, 14 resistance-trained men performed concentric and eccentric squats from 10 to 90% of their maximum average force values in 10% increments. The order of testing was randomized, and a 2-minute rest period was allowed between each repetition. Bipolar surface EMG signals were detected for the right vastus lateralis, rectus femoris, and biceps femoris throughout the testing, and the middle portion of each range of motion was selected for analysis. Linear regression was used on an individual subject basis to examine the linearity for the relationships between EMG amplitude and concentric and eccentric squat forces, and repeated measures analyses of variance (ANOVAs) were used to assess mean differences among the muscles.
Fourteen resistance-trained men (mean age ± SD: 22 ± 2 years; mean height ± SD: 173.1 ± 6.6 cm; mean body mass ± SD: 82.8 ± 9 kg, mean estimated thigh muscle cross-sectional area ± SD: 221.9 ± 22.7 cm2 (22)) participated in this study. Potential subjects were required to have participated in a structured exercise program that included lower-body resistance training at least once per week over the past 6 months. All of the subjects were able to barbell back squat at least 1.5 times their body mass. The subjects refrained from lower-body exercise for at least 72 hours before data collection. Before testing, each subject was screened by the investigators for any previous or current injuries. Each subject completed a health history questionnaire, which indicated no current or recent neuromuscular or musculoskeletal disorders. The procedures for this investigation were approved by the Texas Tech University Institutional Review Board for human subjects, and all of the subjects signed an informed consent form before participation. The informed consent was written in lay terms, and the subjects understood that they could withdraw from the study at any time without penalty. This study was conducted during spring 2013.
Familiarization and Data Collection Sessions
During the initial visit to the laboratory, the subjects were familiarized with the procedures for data collection to reduce the influence of a learning effect on each of the investigation's dependent variables. Upon arrival to the laboratory, the subjects performed 10 body weight squats to warm-up the musculature involved in testing. Following the warm-up, testing with the squat device (eSQ; Exerbotics, LLC, Tulsa, OK, USA) commenced. First, the range of motion of each subject was set in accordance with the manufacturer's instructions. Range of motion was based on the knee joints and was set to 90° of flexion at the bottom of the squat and 170° between the thighs and the legs at the top (Figure 1). Once the subjects and testers were comfortable with the assigned range of motion, maximum concentric and eccentric squat force testings began. For all tests, the subjects gently grabbed onto the handles and placed the back pad over the top of their trapezius. Their feet were placed on the platform at a distance that was shoulder width apart, and the toes were placed slightly outward (∼30°). All of the subjects in this study wore tennis shoes, and weight belts or knee wraps were not permitted. The subjects were instructed to simultaneously extend at the hips and knees in a smooth, controlled manner throughout the entire range of motion. They were also instructed not to round their back by keeping their lumbar spine in extension during each repetition. Maximum force testing consisted of 3 continuous repetitions with 4 seconds of concentric and eccentric muscle actions. For the first repetition, the subjects produced forces corresponding to roughly 50% of their maximum values. For the second and third repetitions, however, the subjects provided maximum effort. The subjects were verbally encouraged to “push” and “resist” during concentric and eccentric testings, respectively. Following the 3 repetitions, a software program recorded the highest concentric and eccentric force values, and the average force output for the middle of each range of motion (i.e., 2 seconds) was determined. The subjects then performed 9 individual repetitions at force levels corresponding to each tenth percentile of the maximum average force values (i.e., 10%, 20%, 30%, etc.). A 2-minute rest period was provided between each repetition, and the order of testing was randomized. Visual feedback was not provided to the subjects. Upon completion of the 9 submaximal repetitions, the subjects exited the laboratory. No data were collected during the familiarization session.
Exactly 48 hours after the familiarization session, the subjects visited the laboratory for data collection. Upon arrival, a goniometer (Biometrics, Ltd., Ladysmith, VA, USA) was secured to the lateral aspect of the right knee to track changes in joint angle for all repetitions. The subjects then performed the same procedures for the warm-up and maximum concentric and eccentric squat force testings as described previously. Once the maximum average concentric and eccentric squat force values were determined, the subjects again performed a series of randomly ordered submaximal repetitions in 10% increments. Repetitions were repeated if the actual submaximal average force was not within ±3% of the calculated value. For example, if a subject attempted to produce an average concentric squat force output corresponding to 70.0% of his maximum value but instead produced 74.0%, a 2-minute rest period was allowed, and another repetition was performed. The subjects were asked to continue their normal dietary intake (including caffeine). Hydration status was not assessed before data collection.
Bipolar surface EMG signals were detected during the concentric and eccentric muscle actions from the right vastus lateralis, rectus femoris, and biceps femoris. The signals were detected with 3 separate DE-2.1 sensors (interelectrode distance, 10 mm [Delsys, Inc., Boston, MA, USA]) and amplified (gain = 1,000) 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 Surface EMG for the Non-Invasive Assessment of Muscles project (14), and a reference electrode (Dermatrode; American Imex, Irvine, CA, USA) was placed over the patella. The skin over the patella and the belly of the vastus lateralis, rectus femoris, and biceps femoris was prepared before testing by shaving and cleansing with rubbing alcohol. The EMG signals were digitized at a sampling rate of 2,000 samples per second and stored in a personal computer (Dell Optiplex 755, Round Rock, TX, USA) for subsequent analyses.
All signal processing was performed using custom programs written with LabVIEW programming software (version 8.2; National Instruments, Austin, TX, USA). For each repetition, a 2-second period that corresponded to the middle of the range of motion for both the concentric and eccentric muscle actions was selected for analysis. The root-mean-square (RMS) value of each signal was then calculated as a measure of EMG amplitude.
Linear regression analyses were used on an individual subject basis to examine the Pearson r values (with the corresponding 95% confidence intervals [CIs]), coefficients of determination (r 2), significance levels, linear slope coefficients (μV RMS·N−1), and y-intercepts (μV RMS) for the EMG amplitude vs. concentric and eccentric squat force relationships for the vastus lateralis, rectus femoris, and biceps femoris. Two separate 2-way (muscle [vastus lateralis vs. rectus femoris vs. biceps femoris] × muscle action [concentric vs. eccentric]) repeated measures ANOVAs were used to examine mean differences for the linear slope coefficients and y-intercepts for the EMG amplitude vs. squat force relationships. A 3-way (force [10 vs. 20 vs. 30 vs. 40 vs. 50 vs. 60 vs. 70 vs. 80 vs. 90%] × muscle [vastus lateralis vs. rectus femoris vs. biceps femoris] × muscle action [concentric vs. eccentric]) repeated measures ANOVA was used to examine the EMG amplitude values. Decomposition of a 3-way interaction was performed using the procedures described by Keppel (19 p. 445) for “analyzing simple effects.” An alpha level of 0.05 was used to determine statistical significance for all regression analyses, repeated measures ANOVAs, paired samples t-tests, and Bonferroni post hoc comparisons. The partial η2 statistic was used to evaluate the effect size for each ANOVA. Stevens (24) characterized η2 = 0.01 as corresponding to a small effect size, η2 = 0.06 to a medium effect size, and η2 = 0.14 to a large effect size. Cohen's d was used to assess the effect size for each paired samples t-test (7). Cohen's d values of 0.2, 0.5, and 0.8 are used to characterize small, medium, and large effect sizes, respectively. The 95% CI for each mean difference was also computed for each paired samples t-test.
Linearity of the Electromyographic Amplitude vs. Squat Force Relationships
The Pearson r values with corresponding 95% CIs and coefficients of determination for the EMG amplitude vs. concentric and eccentric squat force relationships are shown in Tables 1 and 2, respectively. As displayed in Table 1, all of the relationships for this study were positive, and only 1 subject (subject 14) produced patterns in which the 95% CI demonstrated values less than zero. As shown in Table 2, 4 of the 28 (14.3%) linear regression analyses performed for the biceps femoris were not statistically significant. All of the relationships for both the vastus lateralis and rectus femoris were statistically significant. For the vastus lateralis, the coefficients of determination (r 2) ranged from 0.587 to 0.957 and from 0.697 to 0.992 for the concentric and eccentric muscle actions, respectively. For the rectus femoris, the coefficients of determination (r 2) ranged from 0.512 to 0.934 and 0.531 to 0.977 for the concentric and eccentric muscle actions, respectively. The coefficients of determination (r 2) for the biceps femoris were 0.030–0.985 for concentric testing and 0.010–0.976 for eccentric testing.
Linear Slope Coefficients
The results from the 2-way (muscle × muscle action) repeated measures ANOVA for the linear slope coefficients for the EMG amplitude vs. force relationships indicated that there was a significant interaction (η2 = 0.432). For the concentric muscle actions, the follow-up 1-way repeated measures ANOVA was not significant (η2 = 0.040). However, the follow-up 1-way repeated measures ANOVA for eccentric testing was statistically significant (η2 = 0.510), and the pairwise comparisons indicated that the linear slope coefficients for both the vastus lateralis and rectus femoris were greater than those for the biceps femoris. For the biceps femoris, the mean linear slope coefficient for the concentric muscle actions was greater than that for the eccentric muscle actions (Figure 2A).
The results from the 2-way (muscle × muscle action) repeated measures ANOVA for the y-intercepts for the EMG amplitude vs. force relationships indicated that there was no significant interaction (η2 = 0.040), and no main effect for muscle action (η2 = 0.161), but there was a main effect for muscle (η2 = 0.747). When collapsed across the concentric and eccentric muscle actions, the mean y-intercept for the vastus lateralis was greater than that for both the biarticular muscles (Figure 2B).
Electromyographic Amplitude Values
The results from the 3-way (force × muscle × muscle action) repeated measures ANOVA for the EMG amplitude values indicated that there was a significant interaction (η2 = 0.285). Based on the recommendations discussed by Keppel (19), we performed 2 separate 2-way ANOVAs in which muscle action was held constant. For concentric force testing (Figure 3), the results from the 2-way (force × muscle) repeated measures ANOVA indicated that there was no significant interaction (η2 = 0.063), but there were main effects for both force (η2 = 0.914) and muscle (η2 = 0.550). The marginal mean pairwise comparisons for force (collapsed across the 3 muscles) indicated: 10% < 30-90%, 20% < 30-90%, 30% < 40-90%, 40% < 60-90%, 50% < 70-90%, 60% < 70-90%, 70% <80-90%. In addition, the marginal mean pairwise comparisons for muscle (collapsed across the 9 force levels) indicated that the EMG amplitude values for the vastus lateralis were greater than those for both the rectus femoris and biceps femoris. For the eccentric muscle actions, the results from the 2-way (force × muscle) repeated measures ANOVA indicated that there was a significant interaction (η2 = 0.404). The results from the 3 separate 1-way repeated measures ANOVAs suggested that the greatest increases in EMG amplitude occurred for the vastus lateralis. For all of the force levels examined, the mean EMG amplitude values for the vastus lateralis were greater than those for the biarticular muscles. In addition, the mean EMG amplitude values for the rectus femoris were greater than those for the biceps femoris from 20 to 90% of the maximum average force value. The results for each of the follow-up 1-way repeated measures ANOVAs have been detailed in Figure 4.
As stated previously, the purpose of this investigation was to examine the linearity of the EMG amplitude vs. concentric and eccentric squat force relationships for monoarticular and biarticular thigh muscles. When examined on an individual subject basis (Tables 1 and 2), the coefficients of determination showed fairly linear patterns for the vastus lateralis for both concentric and eccentric squat force testing. However, lower values were shown for the biarticular muscles, with low-to-moderate correlations for the biceps femoris. Additionally, as displayed in Figure 2A, the mean linear slope coefficient for the EMG amplitude vs. force relationship for the biceps femoris was significantly greater for concentric vs. eccentric testing. Close examination of the group mean data for Figures 3 and 4 shows similar patterns for the rectus femoris and biceps femoris for concentric, but not eccentric, force testing. Collectively, these findings indicated that during squats performed with a standardized movement pattern and range of motion, EMG amplitude values for the right vastus lateralis increased with force in a manner that is similar to that for unilateral, isokinetic muscle actions (26). Our findings for eccentric force testing are in agreement with previous investigations showing decreased hip extensor activity during simultaneous hip and knee extension (12,28).
Electromyography researchers have long been interested in the relationship between signal amplitude and the force output of a muscle. To our understanding, the earliest work on this topic was carried out by Inman et al. (17) and Lippold (21), both of whom reported linear patterns. By the late 1970s, it had become apparent that some of the dissimilarities for the conclusions that had been drawn were the result of differences in the signal processing and detection techniques, as well as in the statistical analyses, used to quantify the relationship. In an attempt to better understand some of the factors that may have caused these inconsistencies, Lawrence and De Luca (20) provided a detailed description of the EMG amplitude vs. isometric force relationship. Their investigation examined the following aspects: (a) Whether the relationship is affected by a subject's training background, (b) whether the relationship varies in muscles with different firing rate vs. recruitment properties, and (c) how much variability exists among the patterns. Surface EMG signals were obtained from the biceps brachii, deltoid, and first dorsal interosseous of elite powerlifters, world-class long-distance swimmers, accomplished pianists, and healthy control subjects. Elite powerlifters and swimmers were studied because it was hypothesized that these athletes exhibited different fiber-type composition for musculature of the upper limb. The results indicated that the linearity of the EMG amplitude vs. force relationship was muscle dependent, with linear patterns shown only for the first dorsal interosseous. Furthermore, it was noted that these relationships were not influenced by training status because the pianists, powerlifters, swimmers, and control subjects demonstrated similar responses. Although cross-talk from adjacent muscles and fiber-type composition was discussed, the available evidence at that time led the authors (20) to conclude that the differences among the muscles were likely related to motor unit firing rate vs. recruitment properties.
It is important to state that the majority of previous investigators who have examined the relationship between EMG amplitude and force have done so for single-joint isometric assessments (10,20,22,27). For many of these studies, investigators have restrained the joint that is being tested as much as possible to minimize the influence of antagonist coactivation or synergist muscle activity. Although isometric force testing may not be as physiologically intriguing as dynamic movements, these precautions were taken for important reasons. As discussed by authors of previous studies (3,8), testing conditions that involve changes in muscle length introduce factors other than the force generated by a muscle that may affect the amplitude of the EMG signal. Without going into great detail, these factors may include movement of the recording electrodes with respect to the active muscle fibers (3), changes in the moment arm through which the external resistance acts (13), and changes in the center of joint rotation (9). Nevertheless, several previous studies have reported positive linear relationships between EMG amplitude and dynamic torque using group mean procedures (4,6) and individual subject analyses (25,26). For example, Stock et al. (26) tested the linearity and reliability of the EMG amplitude vs. concentric torque relationships for the superficial quadriceps femoris muscles. These authors then examined this relationship for the biceps brachii during submaximal to maximal dumbbell curls while also comparing monopolar vs. bipolar recording methods (25). For both of these investigations (25,26), the primary purpose was to determine if the patterns of response were both linear and reliable enough to be used as an alternative to the efficiency of electrical activity technique described by Moritani and deVries (22). Although deVries (10) and Moritani and deVries (22) did not make specific recommendations for the levels of linearity and reliability needed to track changes in neural vs. hypertrophic adaptations, Stock et al. (25,26) concluded in both studies that these relationships were not reliable enough to be used in future investigations because the intraclass correlation coefficients for the linear slope coefficients ranged from 0.594 to 0.888. When compared with previous studies that have reported individual subject data, it appears that our findings for the vastus lateralis were comparable to unilateral, isokinetic muscle actions for the superficial quadriceps femoris muscles (26), dynamic constant external resistance muscle actions for the biceps brachii (25), and isometric force testing for the biceps brachii and vastus lateralis (27). However, our findings for the rectus femoris and biceps femoris demonstrated inconsistent patterns for EMG amplitude vs. force (see e.g., subject 14; Tables 1 and 2). Because differences in fiber-type composition for these muscles would likely not explain our results (18), we speculate that the multiple anatomical roles that these biarticular muscles played resulted in the high degree of variability. One might also speculate that inconsistencies in the control and recruitment of antagonist and synergist muscles not examined may have contributed to the variability in our data (3). In light of our findings, we believe that additional methodological studies are needed before the EMG amplitude vs. concentric and eccentric squat force relationships can be used to track changes in neuromuscular adaptations with training. Of the 3 muscles tested, only the vastus lateralis exhibited patterns that we consider consistent across all 14 subjects. Future investigators might consider performing closely related studies while also examining muscles such as the vastus medialis, gastrocnemius, and gluteus maximus.
The electrical activity of biarticular muscles has been a subject of great interest among scientists and anatomists for many years. Basmajian (2) provided the first data in humans on this topic. In his study (2), subjects were asked to extend or flex at the hip or knee joints, and electrodes were inserted into the proximal, middle, and distal portions of the rectus femoris, semitendinosus, and semimembranosus. The results showed that the greatest EMG activity occurred at the middle of the muscle regardless of whether extension or flexion occurred at the knee or hip. It was concluded that the belly of a biarticular muscle acts as an individual unit and cannot behave in isolation at either end. Fujiwara and Basmajian (12) performed a similar investigation but had subjects perform combined movements (i.e., extensionor flexion or both) at the hip and knee joints in an attempt to examine the influence of reciprocal innervation. Each subject was placed in a supine position, and the leg was supported by a sling such that the knee and the hip were held passively flexed at right angles. Bipolar wire electrodes were inserted into the rectus femoris, medial hamstrings, vastus medialis, gluteus maximus, and iliopsoas throughout testing. For monoarticular movements, the muscle activity was in agreement with many of the authors' hypotheses (e.g., only the medial hamstrings were active during knee flexion). Interestingly, the vastus medialis was active in 7 of the 10 subjects during hip extension. For biarticular movements, 2 interesting findings were noted: (a) The rectus femoris was active during hip flexion but not when hip flexion was accompanied by knee flexion and (b) the rectus femoris was active during knee extension but not when knee extension was accompanied by hip extension. Finally, noteworthy findings were demonstrated by Yamashita (28). The author (28) examined EMG signals for the vastus medialis, rectus femoris, semimembranosus, and gluteus maximus during individual or combined extension at the hip and knee joints. In one of the experiments, the subjects produced isometric hip extension force corresponding to 40% of the maximum voluntary contraction while simultaneously increasing knee extension force. As expected, the amplitude of the EMG signal for the vastus medialis increased. However, what was noteworthy was the fact that the gluteus maximus and semimembranosus became largely inhibited, but hip extension force remained constant. Yamashita (28) speculated that his results were the result of reciprocal Ia inhibition between antagonist biarticular muscles of the thigh and that the central nervous system organizes hip extension in a manner that promotes monoarticular function of the knee. Collectively, these 3 previous investigations (2,12,28) demonstrated the complex nature of biarticular muscle function and suggested that monoarticular movements involving 2-joint muscles are a result of agonist, antagonist, and synergist muscles producing force in a very coordinated pattern. Perhaps, the most novel aspect of our study was the analysis of EMG signal for monoarticular and biarticular muscles during multiple-joint concentric and eccentric muscle actions. As shown in Figure 3, during concentric force testing, the rectus femoris and biceps femoris showed similar patterns for EMG amplitude across force. For eccentric testing, however, the mean EMG amplitude values for the rectus femoris were greater than those for the biceps femoris for almost all of the force levels examined. When combined with the linear slope coefficient data shown in Figure 2A, it is evident that biceps femoris activity was depressed during eccentric, but not concentric, squats. Although we only examined the electrical activity of the 3 muscles, our results indicated that different activation patterns may exist for the hip and knee extensors during concentric vs. eccentric squat force testing.
As is the case for all research studies, the findings for the present investigation are context specific. Most importantly, it must be noted that our results are specific to squats performed with separate 4-second concentric and eccentric muscle actions. The obvious advantage of our research design is that it allowed us to standardize the range of motion, duration, and movement pattern across all of the testing performed in this study. In addition, in the future, the analysis of dependent variables obtained from force curves (e.g., concentric and eccentric peak and average force, power, the rate of force development, etc.) during multiple-joint testing will provide researchers with a great deal more information than can be provided with dynamic constant external resistance muscle actions. However, the disadvantage of this technology is that machine-based force testing may minimize the need for the recruitment of motor units for stabilizer muscles that are an essential part of free-weight training (13). Thus, although we believe that our findings have implications for functional movements involving muscles that cross 1 or 2 joints, we caution that they may not be applicable to barbell front or back squats. Nonetheless, assessing the EMG responses and force production capabilities of athletes during multiple-joint exercise could be an intriguing area of study, especially for answering mechanistic research questions related to residual force enhancement and eccentric-only strength training (15).
In summary, these data showed moderate-to-high coefficients of determination for the EMG amplitude vs. concentric and eccentric squat force relationships for the vastus lateralis. Although some of the r 2 values for the rectus femoris were high, these patterns were not consistently linear across all of the subjects and were generally lower than those for the vastus lateralis. For several of the subjects, EMG amplitude for the biceps femoris was weakly correlated with eccentric force. This was reinforced by the differences in the linear slope coefficients for the EMG amplitude vs. force relationship for concentric vs. eccentric testing. Our findings are in agreement with 2 previous studies (12,28) that have shown decreased EMG activity for the hip extensors during movements involving both the hip and knee joints. Although we examined the electrical activity for only 3 muscles, our findings indicated that different patterns of response may exist for the quadriceps and hamstrings for concentric vs. eccentric squats. We caution that these results are specific to squats performed with a standardized range of motion, duration, and movement pattern, and until additional research is performed, may not be applicable to barbell training or dynamic constant external resistance strength testing.
This study showed that when both range of motion and movement duration were accounted for, muscle activation for the vastus lateralis generally increased with force in a linear fashion during submaximal squats. With additional research, it is possible that the linear slope coefficient for the EMG amplitude vs. squat force relationships may be useful for examining neural vs. hypertrophic adaptations to strength training. However, our most important finding was the fact that the biceps femoris showed minimal increases in EMG amplitude across the entire range of forces examined for eccentric, but not concentric, testing. This would suggest that different activation strategies may exist for agonist and antagonist muscles that cross 2-joints during multiple-joint concentric vs. eccentric muscle actions. Although our EMG results should be interpreted with caution, the techniques used in this study may be useful for developing quadriceps strength and could be an intriguing testing alternative for strength and conditioning professionals that are interested in tracking changes in both concentric and eccentric force outputs.
The authors declare no conflicts of interest or professional relationships with companies who might benefit from the results of this study. These results do not constitute endorsement of the products used in this study by the authors or the National Strength and Conditioning Association.
1. Ahtiainen JP, Häkkinen K. Strength athletes are capable to produce greater muscle activation and neural fatigue during high-intensity resistance exercise than nonathletes. J Strength Cond Res 23: 1129–1134, 2009.
2. Basmajian JV. Electromyography of two-joint muscles. Anat Rec 129: 371–380, 1957.
3. Basmajian JV, De Luca CJ. Muscles Alive (5th ed.). Baltimore, MD: Williams & Wilkins, 1985.
4. Beck TW, Housh TJ, Johnson GO, Weir JP, Cramer JT, Coburn JW, Malek MH. Mechanomyographic and electromyographic time and frequency domain responses during submaximal to maximal isokinetic muscle actions of the biceps brachii. Eur J Appl Physiol 92: 352–359, 2004.
5. Clark DR, Lambert MI, Hunter AM. Muscle activation in the loaded free barbell squat: A brief review. J Strength Cond Res 26: 1169–1178, 2012.
6. Coburn JW, Housh TJ, Cramer JT, Weir JP, Miller JM, Beck TW, Malek MH, Johnson GO. Mechanomyographic and electromyographic responses of the vastus medialis muscle during isometric and concentric muscle actions. J Strength Cond Res 19: 412–420, 2005.
7. Cohen J. Statistical Power Analysis for the Behavioral Sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, 1988.
8. De Luca CJ. The use of surface electromyography in biomechanics. J Appl Biomech 13: 135–163, 1997.
9. De Luca CJ, Forrest WJ. Force analysis of individual muscles acting simultaneously on the shoulder joint during isometric abduction. J Biomech 6: 385–393, 1973.
10. deVries HA. “Efficiency of electrical activity” as a physiological measure of the functional state of muscle tissue. Am J Phys Med 47: 10–22, 1968.
11. Fischer A, Merhautova J. Electromyographic manifestations of individual stages of adapted sports technique. In: Health and Fitness in the Modern World. Chicago: The Athletic Institute, Library of Congress No. 61-12315, 1961. pp. 134–147.
12. Fujiwara M, Basmajian JV. Electromyographic study of two-joint muscles. Am J Phys Med 54: 234–242, 1978.
13. Harman E. Biomechanics of resistance exercise. In: Essentials of Strength Training and Conditioning. Baechle T.R., Earle R.W., eds. Champaign, IL: Human Kinetics, 2008. pp. 65–91.
14. Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol 10: 361–374, 2000.
15. Herzog W. Mechanisms of enhanced force production in lengthening (eccentric) muscle contractions. J Appl Physiol 2013 [epub ahead of print].
16. Hoffman JR, Ratamess NA, Kang J, Gonzalez AM, Beller NA, Craig SA. Effect of 15 days of betaine ingestion on concentric and eccentric force outputs during isokinetic exercise. J Strength Cond Res 25: 2235–2241, 2011.
17. Inman VT, Ralston HJ, Saunders JB, Feinstein B, Wright EW. Relation of human electromyogram to muscular tension. Electroencephalogr Clin Neurophysiol 4: 187–194, 1952.
18. Johnson MA, Polgar J, Weightman D, Appleton D. Data on the distribution of fibre types in thirty-six human muscles: An autopsy study. J Neurol Sci 18: 111–129, 1973.
19. Keppel G. Design and Analysis: A Researcher’s Handbook (3rd ed.). Upper Saddle River, NJ: Prentice-Hall, Inc., 1991.
20. Lawrence JH, De Luca CJ. Myoelectric signal versus force relationship in different human muscles. J Appl Physiol 54: 1653–1659, 1983.
21. Lippold LC. The relation between integrated action potentials in a human muscle and its isometric tension. J Physiol 117: 492–499, 1952.
22. Moritani T, deVries HA. Neural factors versus hypertrophy in the time course of muscle strength gain. Am J Phys Med 58: 115–130, 1979.
23. Ratamess NA, Beller NA, Gonzales AM, Spatz GE, Hoffman JR, Ross RE, Faigenbaum AD, Kang J. The effects of multiple-joint isokinetic resistance training on maximal isokinetic and dynamic muscle strength and endurance. J Strength Cond Res 26: S1–S130, 2012.
24. Stevens JP. Intermediate Statistics (3rd ed.). New York, NY: Taylor & Francis Group, 2007.
25. 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 194: 94–101, 2010.
26. 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 50: 97–106, 2010.
27. 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 71: 283–287, 1992.
28. Yamashita N. EMG activities in mono- and bi-articular thigh muscles in combined hip and knee extension. Eur J Appl Physiol Occup Physiol 58: 274–277, 1988.
Keywords:Copyright © 2014 by the National Strength & Conditioning Association.
agonist; antagonist; motor unit; quadriceps; hamstrings