Secondary Logo

Journal Logo

Muscle Activation of the Elbow Flexor and Extensor Muscles During Self-Resistance Exercises: Comparison of Unilateral Maximal Cocontraction and Bilateral Self-Resistance

Serrau, Virgile1; Driss, Tarak1; Vandewalle, Henry2; Behm, David G.3; Lesne-Chabran, Ethelle1; Le Pellec-Muller, Armande1

Journal of Strength and Conditioning Research: September 2012 - Volume 26 - Issue 9 - p 2468–2477
doi: 10.1519/JSC.0b013e31823bc0a2
Original Research

Serrau, V, Driss, T, Vandewalle, H, Behm, DG, Lesne-Chabran, E, and Le Pellec-Muller, A. Muscle activation of the elbow flexor and extensor muscles during self-resistance exercises: comparison of unilateral maximal cocontraction and bilateral self-resistance. J Strength Cond Res 26(9): 2468–2477, 2012—Muscle activation during self-resistance exercises was studied in 18 subjects performing (a) maximal unilateral isometric cocontractions of flexor and extensor muscles of the right elbow (UNI); (b) bilateral exercises consisting of maximal isometric extensions of the right elbow against the left elbow flexors (BiExt) and maximal isometric flexion of the right elbow against the left elbow extensors (BiFlex). Force production by the biceps brachii (BB), brachioradialis (BR), and triceps brachii (TB) during UNI, BiFlex, and BiExt were estimated by comparing the integrated surface electromyograms (iEMG) of BB, BR, and TB during UNI, BiExt, and BiFlex with the individual iEMG-force relationship determined from isometric contractions at 30, 60, and 100% maximal voluntary contraction during elbow flexion (MVCflex) or extension (MVCext) against a force transducer. During BiFlex for BB or BR and BiExt for TB, the values (mean ± SE) of BB-iEMG, BR-iEMG, and TB-iEMG were 74.0 ± 4.5, 76.6 ± 5.7, and 84.4 ± 4.5% iEMG at MVC (% iEMGmax). The forces were 86.0 ± 3.7% TB-Forcemax during BiExt, 74.1 ± 3.6% BB-Forcemax and 71.8 ± 4.0% BR-Forcemax during BiFlex. During UNI, BB-iEMG, BR-iEMG, and TB-iEMG were 59.9 ± 4.6, 53.4 ± 4.0, and 66.3 ± 4.7% iEMGmax, respectively. The forces during UNI (70.4 ± 4.0% TB-Forcemax, 60.4 ± 4.3% BB-Forcemax, and 49.2 ± 3.1% BR-Forcemax) were significantly lower than those during bilateral exercises. A 2-way analysis of variance (Muscle × Exercise) indicated that the effects of Muscle and Exercise upon % iEMGmax were significant (p < 0.05; p < 0.001, respectively). In conclusion, bilateral opposition exercises should be more effective in developing strength than cocontraction exercises, which correspond to a moderate activation level even for weak agonist muscle groups.

1Research Center on Sport and Movement (CeRSM), Sciences and Techniques of the Physical and Sporting Activities (UFR STAPS), University Paris Ouest Nanterre La Defense, Nanterre Cedex, France

2Laboratory of Physiology, UFR Health, Medicine, and Human Biology, University of Paris XIII, Bobigny Cedex, France

3School of Human Kinetics and Recreation, Memorial University of Newfoundland, St. John's Newfoundland, Canada

Address correspondence to Tarak Driss,

Back to Top | Article Outline


Self-resistance exercises involve either opposing body parts (e.g., the right arm against the left arm, i.e., contralateral opposition) or cocontraction exercises (coactivation of agonist and antagonist muscles). In a recent study(21), an 8.5 and 5.8% increase in maximal voluntary contraction (MVC) of the elbow extensors and flexors was observed after a program of self-resistance anisometric cocontraction training. It is implicitly assumed that muscle tension should be high during self-resistance exercises because strength improvement is mainly observed in strength training programs soliciting a high percentage of maximal voluntary force (38). However, the literature examining bilateral contractions often reports a bilateral deficit (22,36). A bilateral deficit refers to a reduced force output with bilateral contractions because compared with the sum of 2 unilateral contractions (22,36). Because the bilateral deficit is more common in nonstrength trained individuals (28), it is possible that this phenomenon could adversely affect self-resistance strength training with untrained individuals. Second, the bilateral deficit refers to a decrease in force and activation when contracting homologous contralateral muscles. To the authors' knowledge, there is no research examining activation responses to the simultaneous contraction of contralateral flexor and extensor muscles. Because there is some evidence of muscle activation facilitation with the simultaneous contraction of upper and lower limbs (12), it would be important from a physiological and practical standpoint to examine the consequences of simultaneous contralateral flexor and extensor contractions.

Cocontractions are used in both rehabilitation and strength training regimens (7). However, few investigations (10,34) have examined the activation levels during the performance of cocontractions in healthy subjects. The mean frequency of single motor unit discharges of the biceps brachii (BB) was higher during a maximum isometric contraction as compared with the cocontraction condition (10). Surface EMG of the biceps and triceps brachii (TB) during maximal voluntary cocontraction has been compared with the EMG activity of the same muscles during maximal voluntary elbow flexions and extensions against the resistance of a Cybex II Dynamometer in isometric conditions (34). Whatever the elbow angle, they observed that the EMG activity levels during maximal cocontractions were significantly lower than the maximal contractions against resistance. Unfortunately, the relationship between EMG and torque (or force) was not determined (34) and, consequently, it was not possible to estimate the force output during the maximal cocontraction. Indeed, the EMG-force relationship is either linear (31,39) or curvilinear (23,31,39). There is no study in the literature, which investigates the activation levels during self-resistance exercises involving opposing contralateral contractions and therefore, no study, which compares contralateral opposition and cocontraction exercises.

The difference (8.5 vs. 5.8%) between the effects of an anisometric cocontraction training program (21) upon the strength of the elbow flexors and extensors has also been found for other types of strength training programs such as for example training with an accentuated external resistance during the eccentric phase of elbow flexor or extensor exercises (5). This difference in the training effect for elbow flexors and extensors after anisometric cocontraction training could be the result of differences in trainability or training levels of these muscle groups. However, it could also be explained by a difference in training loads because of a difference in the strength of the antagonist muscle groups. Indeed, for a maximum isometric cocontraction “elbow flexors need only to be activated to the extent to which they cancel the torque produced by the elbow extensor” (34). Therefore, the higher strength increase in elbow extension (21) after an anisometric cocontraction program could also be explained by a higher activation of the extensors during training sessions when compared with the elbow flexors. The same principle should be applied to contralateral opposition exercises. Other things being equal, the activation (% MVC) of an agonist muscle group during a maximal self-resistance exercise should be low if the strength of the antagonist is low when related to the agonist strength. Consequently, low activation levels should correspond to low “MVC antagonist/MVC agonist” ratio.

The aims of this were (a) the comparison of muscle forces during self-resistance exercises consisting of either maximal cocontractions or contralateral oppositions; (b) the study of the relationship between the maximal activation of an agonist muscle during self-resistance exercises and the strength of its antagonists.

Back to Top | Article Outline


Experimental Approach to the Problem

The muscle force levels (percentage of MVC) during self-resistance exercises (unilateral maximal cocontractions [UNI] vs. BiFlex and BiExt) were estimated by comparing iEMG of 2 elbow flexors (BB and brachioradialis [BR]) and 1 elbow extensor (lateral head of the TB) during self-resistance exercises and iEMG during different calibrated (30, 60, and 100% MVC) isometric exercises of the flexor and extensor muscles at the same elbow angle to determine the relationships between integrated surface electromyograms (iEMGs) and force. In addition, the MVC of the elbow flexors and extensors was also measured for the left arm to estimate the strength of the antagonist muscle groups during BiFlex and BiExt protocols.

Back to Top | Article Outline


Eighteen active right-handed physical education students (age 24.1 ± 2.2 years, weight 71.3 ± 6.6 kg, height 175.7 ± 4.8 cm) were tested. None of them regularly practiced strength training exercises. They had no history of upper or lower-extremity strength training during the 6 months before the study. They were considered healthy and had no medical contraindications that could affect the results of the study. The subjects were prohibited from consuming any known stimuli or depressants that would possibly enhance or compromise the experimental results. They were requested to maintain their habitual physical activity and to avoid strenuous activity during the 24 hours before the test session. All the subjects had the same standard isocaloric meal at 12:30 (±15 minutes). Each subject was permitted to drink only 1 glass (∼200 ml) of water before the experimentation to standardize the level of hydration and to avoid the effects of postprandial thermogenesis. The data were collected at 16:00 (±30 minutes). Laboratory temperature was maintained constant at about 21° C during the experimentation. In the first session, the subjects were familiarized with the experimental procedures before signing an informed consent document approved by the Institutional Review Board of the University.

Back to Top | Article Outline



The flexor and extensor isometric MVC was measured on a special chair on which the subjects were fastened by means of straps and belts at the shoulders and waist levels. A strain gauge transducer (TME, France) was fixed on the seat of the chair between the thighs for the measurement of the MVC of the elbow flexors (MVCflex, Figure 1A) and on an adjustable horizontal bar for the elbow extensors (MVCext, Figure 1B).

Figure 1

Figure 1

The subjects exerted their force by pulling on a strap linked to the force transducer. This strap was located around the wrist (ulna styloid process for the extension and radial styloid process for the flexion).

The strain gauge output was fed into a computer by means of a data acquisition card (FastLab card, Eurosmart, France) with a sampling frequency of 1,000 Hz. For the 30 and 60% MVC exercises, the required force level and the actual force were displayed on the computer monitor.

Simultaneously with force data, EMG activity of the long head of the BB, the BR, and the lateral head of the TB were collected by means of skin surface electrodes (bipolar electrodes with preamplifiers, 2-cm interelectrode distance) and recorded by means of a MA300-6 EMG System (Motion Lab Systems, Baton Rouge, LA, USA) at a 1,000-Hz sampling frequency.

Back to Top | Article Outline


Maximal Flexion and Extension Protocol

The right elbow was positioned at 90° flexion with the arm in the vertical position along the thorax, the shoulder at 45° internal rotation (Figure 1), and the wrist in neutral position between pronation and supination. The MVC of the right elbow flexors (MVCflex-right) and MVC of the right elbow extensors (MVCext-right) were measured during a 5-second maximal isometric contraction against the resistance of the strap linked to the strain gauge (Figures 1A and B). The subjects performed 5 trials for both MVCflex-right and MVCext-right with a 2 minutes' recovery period between trials.

The same protocols were used for the left elbow to measure MVCext (MVCext-left) and MVCflex (MVCflex-left).

Back to Top | Article Outline
Submaximal (30 and 60%) Flexion and Extension Protocols

The 30 and 60% MVC exercises consisted of maintaining a force level equal to 30 or 60% MVCflex-right (or MVCext-right) during 5 seconds with the required force level and the actual force displayed on the computer monitor. The subjects performed 3 trials at 30 and 60% MVCflex-right and 3 trials at 30 and 60% MVCext-right.

Back to Top | Article Outline
Protocols of Unilateral Maximal Cocontractions

The UNI protocol (Figure 1C) corresponded to the maximal isometric efforts of both flexor and extensor elbow muscles. The instruction was: “Try to tense your arm muscles as hard as you can without moving your arm, that is, keeping the elbow in the same position (90° elbow flexion, 45° internal rotation of the shoulder, wrist in midprone position) as in the MVCflex-right exercise.”

Four trials were performed for UNI cocontractions. All the trials were performed for 5 seconds with a 90-second recovery period between trials.

Back to Top | Article Outline
Protocols of Maximal Bilateral Self-Resistance Exercises (BiFlex and BiExt)

Maximal bilateral self-resistance exercises consisted of maximal contractions against contralateral oppositions: (a) maximal efforts of isometric flexion of the right elbow against the opposition of the left elbow extensors (BiFlex, Figure 1D) and (b) maximal efforts of isometric extension of the right elbow against the opposition of the left elbow flexors (BiExt, Figure 1E).

As for UNI, 4 trials were performed for BiFlex and BiExt. All the trials were performed for 5 seconds with a 90-second recovery period between trials.

The right elbow was in the same position as in the MVCflex-right and MVCext-right or the UNI protocol. The opposition between the left and the right limbs was located at the ulna and radial styloid processes with the interposition of a thin (7-mm) foam cushion to avoid pain.

Back to Top | Article Outline
Experimental Design

After skin preparation, the subjects performed a light elbow flexion with the forearm in the midprone position to locate the EMG electrodes over the long head of the BB (in front of the center of the arm) and the BR muscle (on the top of the contracting muscle, 1 cm in front of the arm). Thereafter, the subjects performed a light elbow extension to locate the EMG electrodes over the contracting lateral head of the TB (at the back of the center of the arm, slightly laterally).

The subjects began the sessions by performing the MVC exercises (100% MVC) of the right and left arms, in a random order, for either elbow flexion or elbow extension. Then, they performed the same exercise (flexion or extension of the right elbow) at 30 and 60% MVC. Thereafter, they performed the same series of contractions (i.e., 100, 30, 60% MVC of the right and left arms) for the other exercise (flexion or extension). Finally, the subjects performed the UNI, BiFlex, and BiExt protocols, in a random order with a 5-minute interval.

Back to Top | Article Outline
Data Processing

Mechanical and EMG data were processed by means of the software package of a Vicon system (Motion Systems, Oxford, United Kingdom). All the data (except the values of MVC and the computation of 30 and 60% MVC) were processed after the completion of the session.

The MVCext (right or left) corresponded to the peak force during maximal voluntary extension, which was equal to the maximal value of a 20-millisecond moving average. The values of MVCext corresponded to the mean of the 3 best trials out of the 5 trials. The same data processing was used for MVCflex (right or left).

The raw EMG data were synchronized with the corresponding force record. Then, EMG data of the TB (TB-iEMG), BR (BR-iEMG), and BB (BB-iEMG) were rectified and integrated by digital procedures.

The TB-iEMG corresponding to MVCext-right (TB-iEMGmax) was the mean (3 best trials) of the iEMG measured on a 512-millisecond window centered on the peak force. Thereafter, TB-iEMG corresponding to the mean of 3 trials at 30 and 60% MVCext-right was measured on a 512-millisecond period corresponding to a steady signal. The forces and iEMG corresponding to 30 and 60% MVCext-right were equal to the means of the 3 trials.

Similarly, the maximal value of TB-iEMG on a 512-millisecond window was measured for each of the 4 trials with the UNI protocol. The value of TB-iEMG corresponding to UNI was equal to the average of the 3 best values. The same procedure was used for the TB-iEMG corresponding to BiExt.

Then, second-order polynomial regressions were used to determine the individual TB-iEMG/TB-force relationships from individual values of TB-iEMG and TB-force corresponding to 30, 60, and 100% MVCext-right protocols. According to the flexor (3,4) and extensor (20) equivalent concept, it was assumed that all the elbow extensors exert their maximal force at MVCext-right and the contributions of the different elbow extensors (percent of force production) are independent of force. Consequently, the estimation of force production by the TB during UNI and BiExt instruction was computed from these individual iEMG-force relationships and expressed as a percentage of the TB-force at MVCext-right (% TB-Forcemax).

The same procedures were used for the estimations of % BB-Forcemax, and % BR-Forcemax during UNI and BiFlex.

The left elbow extensors were the antagonist muscles of BB and BR during BiFlex. Similarly, the left elbow flexors were the antagonist muscles of TB during BiExt. During UNI, the right elbow flexors were the antagonists of TB. Conversely, TB was an antagonist of the right elbow flexors during UNI. Consequently, the following “MVC antagonist/MVC agonist” ratios were calculated: MVCflex-left/MVCext-right for TB during BiExt; MVCext-left/MVCflex-right for BB or BR during BiFlex; MVCflex-right/MVCext-right for TB during UNI; MVCext-right/MVCflex-right for BB or BR during UNI.

Back to Top | Article Outline

Statistical Analyses

A Student t test for paired data was used in the statistical comparisons of the force data (MVCext or MVCflex, right or left, and the different ratios MVC antagonist/MVC agonist) or the iEMG data (TB, BB, BR) collected under different conditions (UNI, BiFlex, and BiExt).

The activations of the elbow flexors ([BB + BR]/2) and elbow extensors (TB) during the maximal cocontraction exercise (UNI) and the bilateral self-resistance exercises (BiExt for TB and BiExt for TB) were compared by means of a 2-way analysis of variance (ANOVA) (Muscle × Exercise) with repeated measures.

A probability level of 0.05 was selected as the criterion for statistical significance. Statistical power was determined to be 0.80 for the sample size used at the 0.05 α level. Effect sizes were calculated as partial eta-squared ηp2 to estimate the meaningfulness of significant findings.

Back to Top | Article Outline


Maximal Voluntary Contraction during Elbow Flexion and Extension

The data of the right arm are presented in Table 1. For the left arm, MVCflex-left was equal to 266.4 ± 13.1 N, and MVCext-left was equal to 199.0 ± 10.5 N. Significant differences were observed for MVCflex-right vs. MVCext-right (p < 0.001; ηp2 = 1), MVCflex-right vs. MVCflex-left (p < 0.02; ηp2 = 0.548), MVCflex-right vs. MVCext-left (p < 0.001; ηp2 = 1), MVCflex-left vs. MVCext-right (p < 0.001; ηp2 = 1), and MVCflex-left vs. MVCext-left (p < 0.001; ηp2 = 1). The MVCext-right was not significantly different from MVCext-left (p = 0.3; ηp2 = 0.050).

Table 1

Table 1

Ratios of MVC antagonist/MVC agonist corresponding to UNI, BiFlex, and BiExt were (mean ± SE) are as follows:

  • MVCflex-right/MVCext-right = 1.44 ± 0.04 for TB during UNI,
  • MVCext-right/MVCflex-right = 0.70 ± 0.09 for BB or BR during UNI,
  • MVCflex-left/MVCext-right = 1.30 ± 0.09 for TB during BiExt,
  • MVCext-left/MVCflex-right = 0.67 ± 0.04 for BB or BR during BiFlex.

The ratio MVCflex-right/MVCext-right was significantly (p < 0.001; ηp2 = 1) different from the ratios MVCext-right/MVCflex-left, MVCext-right/MVCflex-right, and MVCext-left/MVCflex-right. The difference between MVCext-right/MVCflex-right and MVCext-left/MVCflex-right was not significant (p = 0.498; ηp2 = 0.050).

Back to Top | Article Outline

Integrated Surface Electromyogram–Force Relationships at 30, 60, and 100% Maximum Voluntary Contraction

The actual force values at the 30 and 60% instructions and the corresponding values of BB-iEMG, BR-iEMG, and TB-iEMG are given in Table 1 and Figure 2.

Figure 2

Figure 2

The TB coactivation at 100% MVCflex-right (Table 1 and Figure 2, top) was significantly higher than the BB coactivation (p < 0.01; ηp2 = 0.793) or the BR coactivation (p < 0.01; ηp2 = 0.788) at 100% MVCext-right (Figure 2, bottom). On the other hand, the BB and the BR coactivations during MVCext-right were not significantly different at 30 (p = 0.77; ηp2 = 0.050), 60 (p = 0.97; ηp2 = 0.050), and 100% MVCext-right (p = 0.77; ηp2 = 0.050) (Table 1 and Figure 2, bottom).

Back to Top | Article Outline

Integrated Surface Electromyogram and Estimated Forces Corresponding to BiFlex and BiExt

The individual values of iEMG corresponding to the agonist muscles acting during BiFlex (BB and BR) and BiExt (TB) are presented in Figure 3. The mean values and SEs of BB-iEMG, BR-iEMG, and TB-iEMG during BiFlex and BiExt are presented in Table 2. The mean values of BB-iEMG or BR-iEMG during BiFlex and TB-iEMG during BiExt were significantly lower than those of BB-iEMGmax, BR-iEMGmax, and TB-iEMGmax, respectively. Nevertheless, the individual values of BB-iEMG and BR-iEMG during BiFlex and TB-iEMG during BiExt were >100% in 2 cases for each muscle (Figure 3). The values of BR-iEMG during BiFlex were not significantly different from BB-iEMG during BiFlex (p = 0.624; ηp2 = 0.050) and TB-iEMG (p = 0.161; ηp2 = 0.163) during BiExt. The difference between BB-iEMG during BiFlex and TB-iEMG during BiExt was close to the significance level (p = 0.06; ηp2 = 0.351).

Table 2

Table 2

Figure 3

Figure 3

The mean values of BB-force, BR-force, and TB-force estimated from the individual iEMG-force relationships are given in Table 2. The TB-Force during BiExt was significantly higher than BR-Force (p < 0.001; ηp2 = 0.984) and BB-Force (p = 0.012; ηp2 = 0.702) during BiFlex. On the other hand, the difference between BB-Force and BR-Force during BiFlex was not significant.

Back to Top | Article Outline

Coactivation during BiFlex and BiExt

The mean values of TB-iEMG during BiFlex and BB-iEMG or BR-iEMG during BiExt and the estimations of the force corresponding to these levels of coactivation are given in Table 2 (bold data). The BB coactivation during BiExt was statistically lower (p = 0.01; ηp2 = 0.734) than BB coactivation measured during MVCext. On the other hand, BR-iEMG during BiExt was not statistically different from BR-iEMG (p > 0.05; ηp2 = 0.050) measured during MVCext.

Back to Top | Article Outline

Integrated Surface Electromyograms and Estimated Forces Corresponding to UNI

The individual values of the iEMG corresponding to BB, BR, and TB during UNI are presented in Figure 3. The values of the iEMG for BB, BR, and TB muscles, during UNI (Table 2) were significantly lower than the iEMGmax. The individual values of iEMG during UNI were lower than iEMGmax in all the cases except 2 subjects (Figure 3): one for BB-iEMG and one for TB-iEMG.

The BB-iEMG was not significantly different from TB-iEMG (p = 0.227; ηp2 = 0.104) or BR-iEMG (p = 0.197; ηp2 = 0.17). However, the difference between BR-iEMG and TB-iEMG (Table 2) was significant (p = 0.049; ηp2 = 0.411).

The TB-Force (Table 2) was significantly different from BB-Force (p = 0.026; ηp2 = 0.547) and BR-Force (p < 0.001; ηp2 = 0.996). In contrast with BiFlex, BB-Force was also significantly different from BR-Force during UNI (p = 0.023; ηp2 = 0.576).

Back to Top | Article Outline

Comparison between UNI, BiFlex, and BiExt

A 2-way ANOVA (Muscle × Exercise) indicated that Exercise (UNI or Bi) and Muscle (extensor or flexor) factors were both significant (p = 0.034; ηp2 = 0.459 for Muscle and < 0.001; ηp2 = 0.991 for Exercise) without interaction (p = 0.946; ηp2 = 0.050).

The iEMG of TB, BB, and BR during UNI (Table 2) was significantly lower than the iEMG of the same muscles acting as agonists during BiExt and BiFlex: TB-BiExt > TB-UNI (p = 0.001; ηp2 = 0.0997), BB-BiFlex > BB-UNI (p = 0.004; ηp2 = 0.086), and BR-BiFlex > BR-UNI (p < 0.001; ηp2 = 1).

Back to Top | Article Outline


The estimated forces during the self-resistance exercises that involved contralateral opposing contractions (BiExt and BiFlex) were high for TB (86% TB-Forcemax during BiExt) and BB (74% BB-Forcemax during BiFlex). However, these forces were significantly lower than the MVC. Moreover, the estimated forces were significantly lower for the cocontraction exercise (UNI protocol) than for the contralateral opposition exercises (BiFlex and BiExt exercises). The estimated force of the TB during UNI (70% TB-Forcemax) was not an expression of dystonia in the subjects of this study. Indeed, the antagonist activities recorded during contractions of the agonist muscles at 30, 60, and 100% MVC in this study (Figure 2) are similar to the data in the literature.(9,11,13,14,26,33)

In this study, the MVC was lower for elbow extension than that for elbow flexion: MVCext/MVCflex ratios were equal to 0.70 and 0.75 for the right and the left arm, respectively. Differences between MVCext and MVCflex have been observed in previous investigations.(16,21,32) In the first study (32), the MVCext/MVCflex ratio depended on the elbow angle, and the value of this ratio (computed by interpolation of MVC at 80 and 100°) was 0.62, that is, a lower value than in this study. The MVCext/MVCflex ratio in another study (16) was high and equal to 0.88 (young subjects) and 0.82 (old subjects). In a more recent study (21), MVCext/MVCflex ratios before training were equal to 0.75 and 0.77 for the dominant and nondominant arms, respectively. It should be mentioned that the elbow was in full supination in these studies (16,21,32), which should decrease the action of BR as an elbow flexor and, consequently, should increase the MVCext/MVCflex ratio. The 2-way ANOVA (Muscle × Exercise) showed that the activation of the elbow extensor (TB) was significantly higher than the activation of the elbow flexors (average of BB and BR). As previously mentioned (34) for a maximum isometric cocontraction of the elbow flexor and extensor muscles, agonist muscles need to be activated to the extent to which they cancel the torque produced by the antagonist muscles, only. Other things being equal, low activations should be observed for low values of MVC antagonist/MVC agonist ratio. As suggested in Figure 4, the MVC antagonist/MVC agonist ratio could partly explain the difference in muscle activation between elbow flexors and extensors.

Figure 4

Figure 4

However, the differences in MVC antagonist/MVC agonist ratio are not the only cause of the observed differences between the protocols. As shown in Figure 4, the activation for given “MVC antagonist/MVC agonist” ratios was largely ( ∼19% Forcemax) higher in bilateral exercises (BiFlex or BiExt). The estimated force during UNI (Table 2) was equal to 82% of the estimated force during bilateral exercise for TB (TB-Force UNI/TB-Force BiExt = 0.819) and BB (BB-Force UNI/BB-Force BiFlex = 0.815) although the differences in MVC antagonist/MVC agonist ratios were small and not significant. The difference in the estimated force between UNI and BiFlex (Table 2) was even larger for BR (BR-Force UNI/BR-Force BiFlex = 0.685). The 2-way ANOVA (Muscle × Exercise) confirmed that the activation during UNI was significantly lower than that during bilateral exercises.

The fact that BB-Force and BR-Force during BiFlex were significantly lower than BB-Forcemax and BR-Forcemax could be expected because the maximal resistance to the flexion of the right arm (i.e., MVCext-left) was lower than MVCflex-right. However, the TB-Force during BiExt was also significantly lower than TB-Forcemax although the resistance to extension (MVCflex-left) was significantly higher (p < 0,001) than MVCext-right. Since the study by Asmunssen and Heeboll-Nielsen (2), it is known that the MVC produced during a bilateral effort is generally lower than the sum of the maximal force of the right and the left limbs acting separately, which has been termed bilateral deficit. The average value of TB-Force (Table 2) was 16% lower than TB-Forcemax, that is, a difference comparable with the bilateral deficit, which has been found for leg extension: 25% (30), 13% (29), and 9% (35). However, bilateral deficit has been found to be lower in upper limb studies: 4.8 and 5.8% for the elbow flexion (27), 3% for wrist flexion (18) and even a 4–10% increase (37) instead of a deficit. The cause of the bilateral deficit is currently unknown, and it is not obvious that this phenomenon, which has been described for exercises involving the same muscle groups of the right and left side, also applies to the simultaneous actions of agonist muscles on one side and the antagonist muscles on the other side. On the contrary, there are spinal circuits, which could facilitate the action of contralateral antagonist muscles. For example, the stimulation of ipsilateral flexor reflex afferents results in the excitation of ipsilateral flexor and contralateral extensor motoneurons combined with the inhibition of ipsilateral extensor and contralateral flexor motoneurons (24,25).

The lower values of the iEMG during UNI compared with the iEMG during BiFlex and BiExt were probably an expression of the differences in neural processes controlling cocontractions and flexion-extension exercises instead of an expression of Sherrington's “double reciprocal innervation” as previously suggested (34). For example, the action of the central nervous system on homonymous recurrent inhibition (inhibition mediated by Renshaw cells) is different between cocontractions and contractions against an external resistance (25). Renshaw cell activity inhibits force production because of a hyperpolarization and a decrease in the firing frequency of the agonist motoneurons. In addition to the inhibition of the agonist motoneurons (recurrent inhibition), the Renshaw cells of the agonist motoneurons also inhibit the Ia interneurones of the antagonist muscles (inhibition of the reciprocal inhibition), which facilitates the firing of the antagonist motoneurons (24,25). Consequently, the disinhibition of the antagonist muscles by the Renshaw cells facilitates coactivation and the recurrent inhibition of the agonist muscles could partly explain the low level of activation during UNI. On the other hand, the inhibition of Renshaw cells during contractions against an external resistance (25) improves net torque production by a facilitation of torque production by the agonist muscles (disinhibition) and an inhibition of torque production by the antagonist muscles (lower disinhibition). The influence of corticospinal tracts is more important for distal muscles than for proximal muscles (8), and the stabilization of the proximal joints by the cocontraction of antagonistic muscles improves movement accuracy. At the distal level of the upper arm (wrist), there is probably an absence of recurrent inhibition of reciprocal Ia inhibition (25), which should limit coactivation processes mediated by Renshaw cells. Consequently, coactivation processes could be more important for arm muscles (e.g., the BB, which is inserted on the proximal part of the radius) than for forearm muscles (e.g., the BR inserted on the distal part of the radius). The underactivation of BR during UNI could be compensated by an overactivation of its agonist (BB) to counterbalance the action of the antagonist muscles (TB). This could partly explain the lack of a significant difference in the activation level between BB and TB during UNI despite there being large difference in the ratio MVC antagonist/MVC agonist (Figure 4).

The activation levels of the 3 elbow muscles were estimated according to the concept of flexor (3,4) and extensor (20) equivalent muscles. In the flexor equivalent muscle concept, it is assumed that not only are the contributions (percent of force production) of the different elbow flexors independent of force level but also that all the elbow flexors exert their maximal force at MVC. However, the concept of flexor (or extensor) equivalent muscle was proposed and verified for isometric and concentric exercises against inertia or external resistances. But its validity for cocontraction exercises is questionable as suggested by the significant difference between BR-Force and BB-Force during UNI. As a consequence, it is difficult to estimate the activation levels during UNI for the muscles, which were not studied, for example, the major elbow flexor, the brachialis muscle (15).

The variances in % iEMGmax were important for the different muscles whatever the protocol (Figure 3). Although the EMG activity during isometric contractions is generally regarded as reliable (19), the within-day coefficients of variation of EMG data are not negligible and range from 8 to 10% for isometric contractions at 30, 50, and 100% MVC (40). The iEMG data were equal to the average of 3 trials in this study to minimize the importance of within-day variability upon total iEMG variability corresponding to UNI, BiFlex, and BiExt. Moreover, the iEMG values for the different protocols were related to iEMG at MVC as normalization reference (6). The variance of a ratio (σR2) depends on the variances of its numerator (σN2) and the variance of its denominator (σD2). The value of σR2 is equal to the sum of σRN2 and σD2 if the values of the numerator and the denominator are independent. However, the value of iEMG during self-resistance exercises (numerator) and the iEMG value at MVC (denominator) were not totally independent, but both depended on common factors (skin resistance, gain of the amplifier, distances between the electrodes and the muscle fibers, etc.) and the variance of the ratio iEMG/iEMGmax should be lower than the variance of iEMG because of these common factors. The use of the twitch interpolation method (1) has showed that the elbow flexors can be maximally activated, but this is more likely in some individuals than in others. Submaximal efforts during MVC protocols result in low iEMGmax, which, in turn, should result in overestimations of % iEMGmax during UNI, BiFlex, and BiExt and should increase the intersubject variability. However, UNI, BiFlex, and BiExt were also maximal voluntary exercises. It is possible that the subjects who were unable to produce a true maximal contraction during the measurements of MVC were also unable to produce true maximal cocontractions or true maximal contractions against the contralateral arm, which should limit the overestimation of % iEMG and the intersubject variability because of submaximal MVC.

In conclusion, the activation levels during self-resistance exercises were clearly submaximal. The difference in activation levels between contralateral opposition exercises (BiFlex and BiExt) and maximal cocontraction exercise (UNI) was significant and amounted to approximately 18% MVC in favor of bilateral exercises. However, activation depended not only on the types of self-resistance exercises but also on muscle groups. In agreement with the findings of a previous study (34), activation was higher for the elbow extensor muscle (TB) than for the elbow flexor muscles (BB and BR), whatever the exercise protocols (cocontraction or contralateral opposition).

Back to Top | Article Outline

Practical Applications

Self-resistance exercises in strength training can be performed everywhere, in any circumstances, and without any devices. Although maximum intensity contractions are not necessary to achieve strength gains, sufficiently high-intensity contractions must be achieved. Strength gains can be achieved with 40–120% of 1-repetition maximum dependent upon the trained state of the population (17). Thus, the activation levels during cocontraction exercises are high enough to explain the increase in strength after a self-resistance training program, at least in the subjects who are not strength athletes. The results of this study show that the activation level during a self-resistance exercise is lower for a muscle group, which is stronger than its opponent as, for example, the elbow flexors against the elbow extensors. Moreover, contralateral opposition exercises should be more efficient for strength training because activation was higher than that during maximal cocontraction exercises (UNI). However, both types of exercises should be performed in a self-resistance training program because it is difficult to design contralateral opposition exercises for all the muscle groups. For example, it is possible to perform maximal cocontractions of diaphragm and transversus abdominis muscles but very difficult to perform voluntary cocontractions of the neck (or jaw) muscles, which can be trained against the resistance of the hands.

Back to Top | Article Outline


1. Allen GM, McKenzie DK, Gandevia SC. Twitch interpolation of the elbow flexor muscle at high forces. Muscle Nerve 21: 318–328, 1998.
2. Asmunssen E, Heeboll-Nielsen K. Isometric muscle strength of adult men and women. Commun Dan Natl Assoc Infant Paral 11: 3–43, 1961.
3. Bouisset S, Lestienne F, Maton B. The stability of synergy in agonists during the execution of a simple voluntary movement. Electroencephalogr Clin Neurophysiol 42: 543–551, 1977.
4. Bouisset S, Maton B. Muscles, Posture, et Mouvement. Bases et Applications de la Méthode Electromyographique. Paris, France: Hermann, 1995.
5. Brandenburg JP, Docherty D. The effects of accentuated loading on strength, muscle hypertrophy, and neural adaptations in trained individuals. J Strength Cond Res 16: 25–32, 2002.
6. Burden A. How should we normalize electromyograms obtained from healthy participants? What we have learned from over 25 years of research. J Electromyogr Kinesiol 20: 1023–1035, 2010.
7. Fahey TD, Insel PM, Roth WT, Wong I. Fit and Well: Core Concepts and Labs in Physical Fitness and Wellness (2nd ed.). Toronto, Canada: McGraw Hill Ryerson, 2010. pp. 87–134.
8. Ghez C. Cortical control of voluntary movement. In: Principles of Neural Science. Kandel E.R., Schwartz J.H., JH, eds. New York, NY: Elsevier-North Holland, 1981. pp. 323–333.
9. Gribble PL, Mullin LI, Cothros N, Mattar A. Role of co-contraction in arm movement accuracy. J Neurophysiol 89: 2396–2405, 2003.
10. Gydikov AA, Kostov KG, Gatev PG. Mechanical properties and discharge frequencies of single motor units in human m. biceps brachii. In: Neural and Mechanical Control of Movement. Kumamoto M., ed. Kyoto, Japan: Yamaguchi Shoten, 1984. pp. 20–36.
11. Häkkinen K, Kallinen K, Izquierdo M, Jokelainen K, Lassila H, Mälkiä E, Kraemer WJ, Newton RU, Alen M. Changes in agonist-antagonist EMG, muscle CSA, and force during strength training in middle-aged and older people. J Appl Physiol 84: 1341–1349, 1998.
12. Huang HJ, Ferris DP. Upper and lower limb muscle activation is bidirectionally and ipsilateerally coupled. Med Sci Sports Exerc 41: 1778–1789, 2009.
13. Hultborn H. Convergence on interneurones in the reciprocal Ia inhibitory pathway to motoneurons. Acta Physiol Scand Suppl 375: 1–42, 1972.
14. Hunter SK, Enoka RM. Changes in muscle activation can prolong the endurance time of a submaximal isometric contraction in humans. J Appl Physiol 94: 108–118, 2003.
15. Kawakami Y, Nakazawa K, Fujimoto T, Nozaki D, Miyashita M, Fukunaga T. Specific tension of elbow flexor and extensor muscles based on magnetic resonance imaging. Eur J Appl Physiol Occup Physiol 68: 139–147, 1994.
16. Klein CS, Rice CL, Marsh GD. Normalized force, activation, and coactivation arm muscles of young and old men. J Appl Physiol 91: 1341–1349, 2001.
17. Kraemer WJ, Fleck S. Resistance training: Exercise prescription. Phys Sportsmed 16: 69–81, 1988.
18. Kroll W. Isometric cross-transfer effect under conditions of central facilitation. J Appl Physiol 20: 297–300, 1965.
19. Larsson B, Karlsson S, Eriksson M, Gerdle B. Test-retest reliability of EMG and peak torque during repetitive maximum concentric knee extensions. J Electromyogr Kinesiol 13: 281–287, 2003.
20. Le Bozec S, Maton B, Cnockaert JC. The synergy of elbow extensor muscles during static work in man. Eur J Appl Physiol Occup Physiol 43: 57–68, 1980.
21. MacKenzie SJ, Rannelli LA, Yuchevich JJ. Neuromuscular adaptations following antagonist resisted training. J Strength Cond Res 24: 156–64, 2010.
22. Matkowski B, Martin A, Lepers R. Comparison of maximal unilateral versus bilateral voluntary contraction force. Eur J Appl Physiol 111: 1571–1578, 2011.
23. Maton B, Bouisset S. The distribution of activity among the muscle of a single group during isometric contraction. Eur J Appl Physiol 37: 101–109, 1977.
24. McCrea DA. Spinal cord circuitry and motor reflexes. Exerc Sport Sci Rev 14: 105–141, 1986.
25. Pierrot-Deseilligny E, Burke D. The Circuitry of the Human Spinal Cord. Its Role in Motor Control and Movement Disorders. Cambridge, MA: Cambridge University Press, 2005.
26. Psek JA, Cafarelli E. Behavior of coactive muscles during fatigue. J Appl Physiol 74: 170–175, 1993.
27. Sahaly R, Vandewalle H, Driss T, Monod H. Maximal voluntary force and rate of force development in humans—Importance of instruction. Eur J Appl Physiol 85: 345–350, 2001.
28. Secher NH. Isometric rowing strength of experienced and inexperienced oarsmen. Med Sci Sports 7: 280–283, 1975.
29. Secher NH, Rorsgaard S, Secher O. Contralateral influence on recruitment of type I muscle fibres during maximum voluntary extension of legs. Acta Physiol Scand 96: 20A, 1976.
30. Secher NH, Rorsgaard S, Secher O. Contralateral influence on recruitment of curarized muscle fibers during maximal voluntary extension of legs. Acta Physiol Scand 103: 456–462, 1978.
31. Simons DG, Zuniga EN. Effect of wrist rotation on the X-Y plot of mean biceps EMG and isometric tension. Am J Phys Med 49: 253–256, 1970.
32. Singh M, Karpovich PV. Isotonic and isometric forces of forearm flexors and extensors. J Appl Physiol 21: 1435–1437, 1966.
33. Solomonow M, Baratta R, Zhou BH, D'Ambrosia R. Electromyogram coactivation patterns of the elbow antagonist muscles during slow isokinetic movement. Exp Neurol 100: 470–477, 1988.
34. Tyler AE, Hutton RS. Was Sherrington right about co-contractions? Brain Res 370: 171–175, 1986.
35. Vandervoort AA, Sale DG, Moroz J. Comparison of motor unit activation during unilateral and bilateral extension. J Appl physiol 56: 46–51, 1984.
36. Van Dieen J, Ogita F, Dew Haan A. Reduced neural drive in bilateral exertions a performance limiting factor. Med Sci Sports Exerc 35: 111–118, 2003.
37. Wawrizinoszek M, Kramer H. Experimentelle Untersuchungen zur Maximalkraftentwicklungen: Maximale Handschlusskräfte bei Ein und Beidhandbetätigung in zwei unterschiedlichen Arbeitsebenen. Med Sport 24: 105–107, 1984.
38. Wernbom M, Augustsson J, Thomée R. The influence of frequency, intensity, volume and mode of strength training on whole muscle cross-sectional area in humans. Sports Med 37: 225–264, 2007.
39. Woods JJ, Bigland-Ritchie B. Linear and non-linear surface EMG/force relationship in human muscles. Am J Phys Med 62: 287–299, 1983.
40. Yang JF, Winter DA. Electromyography reliability in maximal and submaximal isometric contractions. Arch Phys Med Rehabil 64: 417–420, 1983.

iEMG-force relationship; strength; resistance training; coactivation

© 2012 National Strength and Conditioning Association