The Effect of Contralateral Submaximal Contraction on the Development of Biceps Brachii Muscle Fatigue

Introduction

Recently, it has been shown that during submaximal isometric knee extension, unilateral tasks demonstrate longer time to task failure compared to bilateral tasks generating the same relative force in each leg (Matkowski, Place, Martin, & Lepers, 2011). The authors hypothesized that the bilateral condition would fail earlier due to the larger absolute force being maintained, potentially altering motor unit recruitment and causing decreased perfusion. At task failure, it was also found that unilateral contractions were associated with larger decrements of voluntary activation, suggesting a greater amount of central fatigue (Matkowski et al., 2011). This study was, to our knowledge, the first comparison of time to task failure between unilateral and bilateral contractions generating the same relative level of force in each limb. It is currently unknown whether there is a similar difference in fatigability of arm muscles during unilateral and bilateral contractions.

Fatigability is an important factor to consider when assessing risk of injury in the workplace, as muscle fatigue impacts coordination (Semmler, Tucker, Allen, & Proske, 2007), injury mechanics (Potvin, 2008), proprioception (Voight, Hardin, Blackburn, Tippett, & Canner, 1996), and joint stabilization (Lin et al., 2009; Parnianpour, Nordin, Kahanovitz, & Frankel, 1988). For this reason, ergonomic studies establish recommendations for safe working conditions designed to reduce levels of muscular fatigue (Ciriello & Snook, 1999; Garg, Chaffin, & Herrin, 1978, Snook & Ciriello, 1991). These recommendations can be complicated due to muscle fatigability being task dependent, affected by type of contraction, rate, intensity, duration, motivation, and neural strategy (Enoka & Stuart, 1992). Many prolonged upper-limb tasks, including light manual materials handling, use of power tools, or operation of heavy machinery, require either bilateral or unilateral use of the arms. Often the loads on each arm are unequal, as one arm is used for stabilization or fine motor control while the other performs the more force-demanding task. A fundamental difference in arm muscle fatigability between bilateral and unilateral tasks, as demonstrated by Matkowski et al. (2011) in the leg, would suggest that workplace recommendations should take this finding into account.

Fatigue can be defined as a loss in force-generating ability during voluntary contraction (Søgaard, Gandevia, Todd, Petersen, & Taylor, 2006) and can be difficult to measure directly. As such, indirect measures, such as time to task failure, change in force production after failure, and electromyography (EMG) amplitude and frequency, can objectively quantify the development of muscle fatigue (Enoka & Duchateau, 2008). The purpose of this study was to determine if the development of fatigue during a unilateral isometric flexion task was affected by an additional contraction in the contralateral limb. It was hypothesized that a bilateral contraction would lead to greater fatigue compared to a unilateral contraction, as bilateral tasks require greater neuromuscular drive, potentially increasing the rate of central fatigue (Gandevia, 2001), and would produce larger absolute force, also shown to increase fatigability (Hunter, Critchlow, Shin, & Enoka, 2004). If the level of fatigue during bilateral contractions is greater than during unilateral contractions, it would be expected that for bilateral contractions, time to task failure would be shorter, muscles would produce less force during maximal voluntary isometric contractions (MVICs) post–task failure, and increases in EMG amplitude and decreases in EMG frequency content would be greater and occur more quickly.

Method

Experimental Protocol

This study was set as a balanced-block repeated-measures design (Figure 1). Participants completed two test conditions (unilateral and bilateral) separated by 7 days. To ensure that baseline levels of fatigue were consistent, participants were asked to continue regular daily routines but to avoid any strenuous arm or back workouts for 3 days prior to each collection day. The time of data collection was kept consistent between days to reduce diurnal variability between conditions (Martin, Carpentier, Guissard, van Hoecke, & Duchateau, 1999).

OPEN IN VIEWER

Figure 1.

Schematic overview of the data collection protocol. The order of unilateral and bilateral conditions (7 days apart) was balanced between participants. Three initial maximal voluntary isometric contractions (MVICs) were used to normalize electromyography (EMG) data for each muscle. EMG was collected from triceps brachii (TB) of the dominant arm to monitor co-contraction. Force data obtained during biceps brachii (BB) MVIC was used to calculate the 30% or 15% MVIC masses held during the fatigue protocol as well as to calculate the decrease in force production post-failure. In the unilateral condition, participants isometrically held a mass in their dominant arm corresponding to 30% of maximum force. In the bilateral condition, participants isometrically held two separate masses corresponding to 30% maximum force in dominant arm and 15% maximum force in nondominant arm. RPE = rating of perceived exertion.

Participants first performed MVICs for the dominant-arm biceps brachii (BB) and triceps brachii (TB) as well as the nondominant BB during bilateral testing. Three MVICs were performed for each muscle. Participants were instructed to slowly ramp up (over approximately 3 s) to a maximal contraction and hold it for 2 s. For the BB MVIC, the participants were seated (feet flat on the floor, shoulder and elbow positioned at 0° and 90°, respectively) with forearm fully supinated to grasp a handle mounted to a force transducer (Vernier Force Plate, Beaverton, OR; range = 0 to 850 N compression; resolution = 0.3 N) secured to the underside of an immovable bench; maximal EMG and force were recorded for EMG normalization and percentage change in force production, respectively. Further instructions during BB MVICs were to not raise the shoulder or push off the floor to ensure that force was solely generated by the elbow flexors. For the TB MVIC, participants stood with shoulder and elbow positioned at 0° and 90°, respectively, and with forearms midsupinated. The experimenter then provided manual resistance (to prevent any motion) at the forearm as participants generated a maximal elbow extensor effort. Maximal EMG was recorded for EMG normalization. Strong verbal encouragement was provided; however, visual force feedback was not given to participants. Two minutes of rest were given between each MVIC to prevent fatigue. In order to reduce between-condition variability, the order of BB and TB MVICs was kept consistent between conditions for each participant; however, the order was counterbalanced between participants.

After MVICs, participants completed an isometric fatigue task to failure while standing with shoulder and elbow at 0° and 90°, respectively. For the unilateral condition, participants held a mass in their dominant hand requiring 30% of their dominant MVIC elbow flexor force. During the bilateral condition, participants held the same mass in their dominant hand while also holding a second mass in their contralateral hand requiring 15% of their nondominant MVIC elbow flexor force. The 30% MVIC force in the dominant hand was chosen to correspond with the transition between low-level intensities and moderate- to high-level intensities, levels at which the fatigue process tends to be dominated by central/neural processes and intramuscular processes, respectively (Place, Bruton, & Westerblad, 2008). The nondominant arm was also expected to fatigue while generating 15% MVIC force; however, isometric elbow flexor contractions at this level can be sustained for >40 min (Søgaard et al., 2006). Placing less demand on the nondominant arm ensured that the failure point of the fatiguing task would be due to failure in the dominant arm, allowing comparisons to be made between conditions. Verbal encouragement was given to participants during the fatigue task. The fatigue protocol was terminated when the participant elected they could no longer continue or if they deviated from 90° elbow flexion for greater than approximately 5 s (Hunter et al., 2004). Time to task failure was not revealed to participants until both testing sessions were completed.

Prior to the fatiguing protocol, participants gave an initial rating of perceived exertion (RPE) for each arm between 0 and 10, with 0 being not fatigued at all and 10 being completely fatigued. This baseline measurement was used to confirm that participants did not perceive any fatigue prior to the fatiguing protocol. Three participants reported some fatigue (1 to 2.5 out of 10) at baseline; however, as they reported the same baseline levels on both testing days, they were included in the study. If baseline levels were >1 point different, the participant would have been excluded from the study; no participants were excluded due to this criterion. Immediately after the fatigue protocol, participants gave a final RPE for each arm with the same instructions.

Immediately after the fatigue task, a single dominant arm elbow flexor MVIC, in the identical manner as described earlier, was repeated to measure post-failure maximum force. This value was obtained to measure the drop in maximal force after the fatiguing task, an objective measure of fatigue (Enoka & Duchateau, 2008). A single force value was obtained to minimize recovery time. The final MVIC and RPE were completed within 30 s after task failure.

Data Processing and Statistical Analysis

A custom LabView program was used for data analysis. The first and last second of the fatigue protocol were removed from EMG analysis to ensure steady-state contractions. Muscle activation (average amplitude [aEMG]) and median power frequency (MdPF) were calculated for 3-s epochs at the beginning and end of the fatigue protocol as well as at 20-s intervals (Figure 2).

OPEN IN VIEWER

Figure 2.

Representative electromyography (EMG) data from right biceps brachii during a unilateral contraction. Raw EMG was fast Fourier transformed (FFT) over three 1-s windows, and the median power frequency (MdPF) was calculated for each window and averaged. Raw EMG was linear enveloped (rectified, low-pass filtered at 2.5 Hz) and expressed as a percentage of maximal voluntary isometric contraction. Average amplitude (aEMG) was averaged over 3-s windows. The last 3 s was used for the final epoch, even if it did not fall within a 20-s epoch. Data were collected and analyzed at 2048 Hz.

For aEMG, data were debiased, rectified, low-pass filtered (second-order Butterworth, f_C = 2.5 Hz), and normalized to the maximal activation obtained during MVICs (%MVIC; Winter, 2009); averages were then calculated over each 3-s epoch. For frequency analysis, the epochs were divided into three 1-s blocks, and the raw EMG signals were fast Fourier transformed (FFT). Three 1-s blocks were used rather than a single 3-s block to ensure the signal was stationary (Challis & Kitney, 1991; Cifrek, Medved, Tonković, & Ostojić, 2009). The average MdPF for each epoch was used to calculate the rate of spectral changes.

Time to task failure (s) assessed the muscle’s ability to resist fatigue during each condition, and the change in RPE (final – initial) measured subjective fatigue. Objective measures of BB fatigue were percentage change ([final – initial] / initial * 100%) in force, aEMG, and MdPF. Initial values were compared between conditions to ensure that participants had similar baseline levels for all variables. Linear regressions of aEMG and MdPF were calculated across all epochs to assess the rate of fatigue development.

To determine whether antagonist muscle activation changed throughout the fatigue protocol, aEMG of BB and TB were used to calculate co-contraction indices (TB / [BB + TB] * 100%). The co-contraction indices were calculated to determine whether an increase or decrease in TB activation affected the demand placed on the BB.

After testing for normality, paired t tests were used to determine whether there was a difference between bilateral and unilateral conditions for time to task failure and percentage change in force, aEMG, MdPF, and co-contraction. A signed-rank nonparametric test was used to test for a difference in RPE between conditions. SAS 9.3 (SAS Institute, Cary, NC) was used for all statistical tests (α = .05).

Results

There was no statistically significant difference for time to failure between unilateral and bilateral conditions (p = .6204; Table 1). There was also no difference in subjective RPE (p = .5863).

OPEN IN VIEWER

Table 1:

Means ± Standard Errors for All 12 Participants During Unilateral and Bilateral Conditions

Variable	Unilateral	Bilateral	p Value
Time to task failure (seconds)	240 ± 26	250 ± 30	.6204
RPE (unitless)	8.3 ± 0.43	8.4 ± 0.35	.5863
Change [(final – initial) / initial * 100]
Force (%)	−24 ± 2.9	−28 ± 2.9	.1698
aEMG (%)	130 ± 28	120 ± 22	.7223
MdPF (%)	−26 ± 3.3	−24 ± 3.7	.3292
Co-contraction (%)	−24 ± 5.0	−25 ± 6.1	.8657
Rate of change
aEMG (%MVIC/min)	4.1 ± 0.91	3.4 ± 0.5	.2818
MdPF (Hz/min)	−4 ± 1.2	−4 ± 1.1	.5373

Note. RPE = rating of perceived exertion; aEMG = average electromyography amplitude; MdPF = median power frequency; MVIC = maximum voluntary isometric contraction.

Elbow flexor muscles produced less force during MVIC contractions after the fatigue task compared to baseline in both unilateral and bilateral conditions (Figure 3A); however, there was no difference in the drop in force between conditions (p = .1698; Table 1). Participants also produced the same baseline force for both unilateral and bilateral conditions (230 ± 16 N and 230 ± 14 N, respectively).

OPEN IN VIEWER

Figure 3.

Initial and final values for (A) maximum voluntary isometric contraction force, (B) average linear enveloped electromyography (EMG), (C), median power frequency of raw EMG, and (D) co-contraction of elbow flexors and elbow extensors [triceps brachii / (triceps brachii + biceps brachii) * 100] for the dominant limb before and after task failure during bilateral (30% mass in dominant hand + 15% mass in nondominant hand) and unilateral (30% mass in dominant hand) conditions.

There were no statistically significant differences between unilateral and bilateral conditions for either EMG activation (p = .7222; Figure 3B) or frequency content (p = .3292; Figure 3C). For both conditions, the initial aEMG was 12 ± 1.7 %MVIC, which doubled by the end of the fatigue trial to 25 ± 3.2 %MVIC and 26 ± 1.7 %MVIC for the unilateral and bilateral conditions, respectively. The initial MdPF for both conditions was 70 ± 2.1 Hz, and this value decreased at similar rates over the course of the fatiguing task to final frequencies of 52 ± 2.5 Hz and 53 ± 3.8 Hz for unilateral and bilateral conditions, respectively.

The levels of co-contraction between BB and TB were not statistically different between conditions (p = .8657; Figure 3D). Co-contraction levels were initially 34 ± 4.6% and decreased to 24 ± 3.2% and 25 ± 4.1% at the end of the unilateral and bilateral fatigue tasks, respectively.

Discussion

The main finding of this study was that isometric BB elbow flexor fatigue was not affected by contraction of the contralateral muscles, as there were no changes in time to failure, MVIC force, or EMG profiles during unilateral compared to bilateral conditions. The hypothesis that bilaterally activated arm muscles would fatigue faster or to a greater extent compared to unilaterally activated arm muscles was rejected.

Two simple objective measures of muscle fatigability include the decrease in maximal force following task failure and the time to task failure. MVIC force loss is often associated with the severity of fatigue (Zwarts, Bleijenberg, & van Engelen, 2008), and the time to task failure is a measure associated with the level of excitatory and inhibitory input to the motor neuron pool (Hunter, Ryan, Ortega, & Enoka, 2002). Neither of these measures was affected by the lesser contraction in the nondominant arm during the bilateral fatigue task. Therefore, these task failure measures indicate that the nondominant limb does not affect the endurance limit for isometric 30% MVIC contraction of the elbow flexor muscles.

Both time to failure and MVIC force post-failure are only able to evaluate the fatigue state after failure has occurred; however, fatigue during prolonged submaximal isometric contractions is a gradual process (Søgaard et al., 2006). Therefore, fatigue should also be objectively measured throughout the task to determine if the rate of development of fatigue is different between conditions. EMG is a useful tool for this purpose. As a muscle fatigues, there is an expected increase in amplitude of surface EMG (Petrofsky, Glaser, Phillips, Lind, & Williams, 1982) mainly due to an increase in motor unit activity (Fuglevand, Zackowski, Huey, & Enoka, 1993). There is also a spectral shift toward lower frequencies with fatigue, due at least in part to decreased intramuscular pH and action potential conduction velocity (Brody, Pollock, Roy, De Luca, & Celli, 1991; Lindström, Magnusson, & Petersén, 1970). We found that in comparing unilateral and bilateral contractions, there were no differences in percentage change or the rate of change of either EMG amplitude (Figure 3B) or frequency (Figure 3C). This finding indicates that the additional contraction in the nondominant arm did not affect fatigue development at the level of the muscle, causing no observable change in motor unit recruitment or action potential conduction velocity.

There were also no differences between initial conditions for each participant, emphasizing that participants had the same level of baseline activation on each test day. Initial MdPF values are consistent with Christensen, Søgaard, Jensen, Finsen, and Sjøgaard (1995) for nonfatigued isometric contractions of BB (69 ± 8 Hz). Because the contralateral limb did not affect fatigue changes within the dominant arm muscles, it is suggested that peripheral rather than central mechanisms (which would be affected by the contralateral limb) are driving the fatigue process in this 30% MVIC isometric contraction. However, this effect is only speculative, as indicators of central fatigue were not measured in this study.

Despite that neither the task failure measures nor the EMG measures differed between conditions, subjective differences may have been present between bilateral and unilateral contractions. Fatigability depends not only on the mechanisms of fatigue but also on the perception of fatigue (Enoka, 2012; Zwarts et al., 2008). Participants rated their perception of fatigue prior to and immediately after the fatiguing task. The initial rating was taken to detect if there were differences in baseline fatigue or levels of motivation between testing days; it was found that there were not. It was expected that the addition of the extra contracting muscle mass in the bilateral condition would increase the amount of perceived fatigue. However, this was not the case, as participants did not report any differences in how fatigued they felt. Therefore, the additional effort required in the nondominant limb did not influence the psychological factors that may have determined time to task failure.

The final variable that was considered to assess differences between bilateral and unilateral conditions was the level of co-contraction between BB and TB. Increased co-contraction increases the demand placed on the agonist (BB) muscle in order to maintain a constant moment (Winter, 2009). Therefore, if co-contraction levels were different between conditions, then the demand placed on the BB may have fluctuated. EMG data were collected from TB throughout the fatiguing contraction, and the results demonstrated that the levels of BB and TB co-contraction were not different between conditions. Our data demonstrated a decreased magnitude of co-contraction throughout the fatiguing trial, which is consistent with previous fatigue studies (Gates & Dingwell, 2010; Missenard, Mottet, & Perrey, 2008) and is energetically more efficient during a prolonged isometric task. However, the activation of the nondominant arm did not affect co-contraction levels during the bilateral fatiguing task.

The nondominant limb was also expected to fatigue during the isometric task. This expectation was confirmed, as during the fatiguing task, the nondominant BB MdPF decreased and aEMG increased, both to lesser levels than in the dominant arm (results not reported). The demand placed on the left arm was halved in order to ensure that termination of the task was due to failure of the dominant arm. Authors of future studies will need to determine whether similar results are found during balanced bilateral tasks and also during dynamic lifting, pulling, or pushing tasks.

The findings regarding elbow flexor fatigability in this study are contrary to the findings in the quadriceps muscle group (Matkowski et al., 2011), which reported that bilaterally maintaining 20% MVIC knee extension leads to greater fatigue, as evidenced by a shorter time to failure than a 20% unilateral contraction of the nondominant limb. One confounding factor in the Matkowski et al. (2011) study is that during the bilateral condition, both legs were fatigued using the same absolute load, thereby raising the possibility that the shortened time to failure was induced by the weaker leg, something that we accounted for and prevented by halving the fatiguing load in the nondominant arm. Second, the failure processes may have been different between the 20% MVIC knee extension and the 30% elbow flexion tasks. If the limiting factor for force production in the 30% contraction was local blood flow, then the additional contraction in the contralateral limb may not have had an effect. Conversely, the lesser 20% MVIC contraction may have been limited by central fatigue (Place et al., 2008) and therefore affected by the contralateral limb.

A third difference between the studies is that a force target task (participant maintains specified force in a constant position) was used in Matkowski et al. (2011), whereas a position target task (participant maintains specified position while supporting constant load) was used in the current study. Force and position target tasks have been shown to result in differing motor unit recruitment strategies (Madeleine, Jørgensen, Søgaard, Arend-Nielsen, & Sjøgaard, 2002; Mottram, Jakobi, Semmler, & Enoka, 2005), which could have led to differences in fatigability. Finally, fundamental differences in the mechanical (e.g., fiber length, physiological cross-sectional area, and muscle size) and physiological (e.g., local blood flow and muscle fiber type) nature of these muscles may also explain the differing results between our study and Matkowski et al.

A limitation of the current study is that a single strictly controlled task (prolonged 30% isometric elbow flexion) was assessed. Authors of future studies will need to verify whether similar results are found during dynamic or intermittent contractions of various other muscle groups. Various intensities of contraction will also need to be tested to determine if more or less forceful contractions demonstrate similar results. The participant pool was also limited to healthy males from a student population, and therefore, authors of future studies will need to determine if the findings translate to pathological and/or working populations.

To close, many tasks (e.g., holding a plank while operating power tools or stabilizing a pressure washer while depressing the trigger) require lesser muscle activity in the contralateral limb while a more demanding task is concurrently performed. The aim of this study was to identify whether there were functional differences in the development of muscular fatigue between bilateral and unilateral tasks, specifically with the addition of a less demanding contralateral task. As there were no differences observed in time to task failure, drop in maximal force, EMG measures, or perceived exertion, the results of this study suggest that muscle fatigue, in the elbow flexor group, is not affected by lesser submaximal isometric contractions in the contralateral muscle group. Therefore, ergonomic assessments of task endurance do not need to involve consideration of the lesser demands placed on the contralateral arm. Rather, endurance limits, such as those set by Björkstén and Jonsson (1977), need only to assess the limb performing the dominant task. As well, assessments of injury risk for relatively isometric fatiguing upper-limb activities do not need to differentiate between bilateral and unilateral tasks.

Key Points