Skip to main content
NCBI home page
Journal of Musculoskeletal & Neuronal Interactions logoLink to Journal of Musculoskeletal & Neuronal Interactions
. 2020;20(1):77–87.

Neuromuscular responses of the superficial quadriceps femoris muscles: muscle specific fatigue and inter-individual variability during severe intensity treadmill running

Haley C Bergstrom 1,✉, Terry J Housh 2, Taylor K Dinyer 1, M Travis Byrd 1, Nathaniel DM Jenkins 3, Kristen C Cochrane-Snyman 4, Pasquale J Succi 1, Richard J Schmidt 2, Glen O Johnson 2, Jorge M Zuniga 5,6
PMCID: PMC7104590  PMID: 32131371

Abstract

Objectives:

This study examined the time course of changes and patterns of responses in electromyographic amplitude (EMG AMP) and EMG mean power frequency (MPF) for the superficial quadriceps muscles during exhaustive treadmill runs within the severe exercise intensity zones (SIZ1 and SIZ2).

Methods:

The EMG signals for the vastus lateralis (VL), rectus femoris (RF), and vastus medialis (VM) as well as times to exhaustion (Tlim) were recorded in ten runners during two exhaustive treadmill runs (SIZ1 and SIZ2). The composite and individual responses were compared among muscles and between intensities.

Results:

The composite patterns of responses in EMG AMP (linear, quadratic, and cubic increases; r2/R2=0.684–0.848) and EMG MPF (linear, quadratic, and cubic decreases; r2/R2=0.648 – 0.852) for the VL and RF were consistent with neuromuscular fatigue in both zones, but those for the VM were not (quadratic, cubic, and non-significant relationships with responses near baseline). The RF tended to demonstrate greater fatigue (EMG MPF decreased from 80–100% Tlim). There was large inter-individual variability (only 10–60% of responses consistent with composite) in response to fatiguing treadmill running.

Conclusions:

The current findings support the examination and characterization of neuromuscular fatigue on an intensity, muscle, and subject-by-subject basis.

Keywords: Critical Velocity, Neuromuscular Fatigue, Superficial Quadriceps, Individual Responses, Treadmill Running

Introduction

Neuromuscular responses including muscle activation and action potential conduction velocity are reflected in the time and frequency domains of the electromyographic (EMG) signal and have also been used to characterize fatigue during dynamic exercises, including cycle ergometry and treadmill running[1-3]. De Luca[1] described changes in the EMG signal as myoelectric manifestations of fatigue that result in a ‘slowing’ of the surface EMG signal, causing an increase in the time, and a decrease in the frequency domains. During fatiguing cycle ergometry exercise, an increase in EMG amplitude (AMP) reflects the fatigue-induced recruitment of additional motor units, increases in firing rate, and/or synchronization[4], while a decrease in EMG mean power frequency (MPF) reflects a reduction in the muscle fiber action potential conduction velocity and changes in the shape of the waveform[5]. It has been suggested the fatigue induced changes in EMG AMP and EMG MPF are the result of the accumulation of metabolic byproducts (i.e., inorganic phosphate, hydrogen, ammonium, and potassium ions) that cause decreases in membrane excitability[6], excitation-contraction coupling involving Ca++ release and re-uptake from the sarcoplasmic reticulum, myofibrillar Ca++ sensitivity for binding with troponin, actin-myosin binding, and ATP production and breakdown[7-10]. These specific fatigue-induced responses have been shown to be intensity[2,11-15], muscle[14,16], and mode[12,14] specific.

Intensity specific responses have previously been reported that include greater positive slope coefficients for EMG AMP and greater negative slope coefficients for EMG MPF over time for exercise intensities in the severe domain compared with the heavy domain[2,11,15]. Thus, it has been suggested that the severe intensity domain is characterized by a greater increase in motor unit activation, as well as greater decreases in action potential conduction velocity than the heavy domain[2,11,15]. In addition, dissociation in neuromuscular responses for the superficial quadriceps muscles have been reported for constant power output[2] and incremental cycle ergometry[16] as well as incremental treadmill running[14], however, linear increases in EMG AMP for all three superficial quadriceps muscles have also been reported for incremental treadmill running[17]. Typically, the rectus femoris (RF) has been shown to demonstrate an earlier onset of fatigue and lower neuromuscular fatigue threshold, compared with the vastus lateralis (VL) and vastus medialis (VM)[14,16]. It has been suggested these intensity, muscle, and mode specific responses are related to differences in biomechanical properties, muscle architecture (biarticular versus monoarticular), and muscle fiber type composition[18,19].

Recently, two distinct zones have been identified within the severe exercise intensity domain based on different metabolic responses at exhaustion[20,21]. The severe intensity zone 1 (SIZ1) is defined as intensities between critical velocity (CV) and 50%∆ (∆ = difference between CV and V˙O2peak) where exhaustion may occur at V˙O2 values less than V˙O2peak. The severe intensity zone 2 (SIZ2) is defined as intensities >50%∆ but less than 175% CV where V˙O2 values reach V˙O2peak at exhaustion[20,21]. There is evidence[22] that neuromuscular responses for the VL, including EMG AMP and EMG MPF, differ between SIZ1 and SIZ2 during cycle ergometry. Specifically, an earlier onset of changes in the EMG signal was reported for SIZ2 versus SIZ1 during rides to exhaustion[22]. Currently, it is unknown if there are distinct neuromuscular responses (EMG AMP and EMG MPF) during treadmill running to exhaustion in SIZ1 and SIZ2, that are dependent on the muscle examined. In addition, although composite (i.e., mean or average) responses are typically used to draw conclusions about motor unit activation strategies and neuromuscular fatigue, there are many factors that affect the EMG signal that may result in great variability in individual responses[23]. Inter-individual variability in patterns of responses has been previously reported for cycle ergometry[12] and treadmill running[24,25], and, thus, it has been suggested EMG signals should be examined on a subject-by-subject basis. No previous studies, however, have examined the individual variability in EMG responses across the superficial quadriceps muscles during exhaustive treadmill running in the SIZ1 and SIZ2. Therefore, the purposes of this study were to: 1) examine the time course of changes and patterns of responses in EMG AMP and EMG MPF responses for the superficial quadriceps muscles (VL, RF, and VM) during treadmill runs to exhaustion within the lower (SIZ1) and upper (SIZ2) intensity zones of the severe domain; and 2) determine if the individual subject patterns of responses were consistent with the composite responses. We hypothesized that: 1) the EMG AMP and EMG MPF responses for the superficial quadriceps muscles would be consistent with fatigue in both zones, but the time course of changes analyses would indicate an earlier onset of neuromuscular fatigue in SIZ2 than SIZ1; 2) the RF, responsible for thigh flexion and leg extension, would demonstrate greater fatigue than the VL and VM; and 3) there would be variability in the individual patterns of responses (linear, quadratic, and cubic) that may not reflect the composite model, but most subjects would demonstrate characteristics of neuromuscular fatigue during exhaustive exercise in the SIZ1 and SIZ2.

Methods

Experimental procedures

The subjects visited the laboratory on 5 separate occasions, with a minimum of 24-48 hours between each visit. During the first visit, an incremental treadmill test was performed to determine the V˙O2peak and velocity at V˙O2peak (vV˙O2peak). The CV was determined from slope of the total distance versus time to exhaustion (Tlim) relationship for 4 constant velocity runs performed during visits 2 – 5. The constant velocity runs performed at the lowest and the highest velocities for the determination of the CV were within the SIZ1 and SIZ2, respectively. The EMG AMP and EMG MPF (from the VL, RF, and VM) and Tlim were examined during SIZ1 and SIZ2 runs.

Subjects

Ten moderately trained, recreational runners (4 men and 6 women; mean±SD age= 23.2±3.3 years, height= 174 ± 8 cm, weight= 69±13 kg) completed this study. Moderately trained was defined as running 16 to 48 km·wk-1 most weeks during the previous 6-months. The subjects took part in one, or a combination, of the following physical activities; running (n=11), cycling (n=7), weight lifting (n=7), and recreational sports (e.g., volleyball, badminton, Jiu Jitsu) (n=3). All subjects were instructed to abstain from caffeine consumption for 4 h prior to each session and from exercise the day of testing. These subjects were from a large data set that included multiple independent and dependent variables. There was no overlap, however, for the neuromuscular variables in this study and previously published papers[20,26,27]. The subjects had no known cardiovascular, pulmonary, metabolic, muscular and/or coronary heart disease. This study was approved by the University Institutional Review Board for Human Subjects and all subjects completed a health history questionnaire and signed a written informed consent document before testing.

Determination of the peak values

Each subject performed a graded exercise treadmill test (GXT) to exhaustion (Precor Inc., Bothell, WA USA) to determine the V˙O2peak and vV˙O2peak. Prior to the test, each subject completed a 3 min warm-up on the treadmill at a velocity of 4.8 km·h-1 and 0% grade, followed by a 3 min passive recovery. Following the warm-up, each subject was fitted with a nose clip and breathed through a 2-way valve (Hans Rudolph 2700 breathing valve, Kansas City, MO, USA). Expired gas samples were collected and analyzed using a calibrated TrueMax 2400 metabolic cart (Parvo Medics, Sandy, UT, USA). The gas analyzers were calibrated with room air and gases of known concentration prior to all testing sessions. The V˙O2, CO2, and ventilatory parameters were recorded breath-by-breath and expressed as 20 s averages[28]. The GXT began at a treadmill velocity of 6.4 km·h-1 and 0% grade. Thereafter, the velocity was increased by 1.6 km·h-1 every 2 min to 14.4 km·h-1 and 0% grade. Following the 14.4 km·h-1 stage, the velocity was no longer increased, however, the treadmill grade was increased by 2% every 2 min until the subject could no longer maintain the running velocity and grasped the handrails to signal exhaustion. The V˙O2peak was defined as the highest 20 s average V˙O2 value recorded during the test. The velocities performed at 0% grade (6.4 to 14.4 km·h-1), were plotted against V˙O2 and the regression equation derived was used to determine the vV˙O2peak.

Determination of critical velocity

Four, constant velocity, randomly ordered treadmill runs at velocities ranging from 79–103% of the vV˙O2peak were performed on separate days. This range of velocities was selected so that each subject could complete 3 to 20 min of exercise before exhaustion[29]. Prior to the start of the run, each subject practiced getting on and off the treadmill at the velocity associated with that run to become familiarized with that velocity. In addition, each subject completed a self-paced walking or jogging 5 min warm-up, followed by 3 min of passive rest. The treadmill was then set to the selected velocity at 0% grade. Timing for each treadmill run began when the subject released the handrails (usually 2–3 s after getting on the treadmill) and was terminated when the subject grasped the handrails to signal exhaustion. Critical velocity was determined from the 2-parameter, linear total distance (TD) versus Tlim model[30]. Critical velocity was defined as the slope of the linear relationship between the TD and Tlim from the four constant velocity runs. The V˙O2, EMG AMP, and EMG MPF responses were recorded during each of the constant velocity runs.

Determination of severe exercise intensity zones

Two distinct intensity zones have been previously identified within the severe domain[20]. The severe intensity domain zone 1 (SIZ1) includes intensities between critical velocity (CV) and 50%∆ (∆ = difference between CV and V˙O2peak), where exhaustion may occur below V˙O2peak. The severe intensity domain zone 2 (SIZ2) includes intensities > 50%∆ but <175% CV, where V˙O2peak is reached at exhaustion. For this study, the lowest and the highest velocities used to derive the CV were intensities within the SIZ1 and SIZ2 for each subject. Specifically, the lowest running velocity for each subject was performed between CV and approximately 50%Δ and was defined as the SIZ1 run. The highest running velocity for each subject was performed between approximately 50%Δ and 175% of CV and was defined as the SIZ2 run.

Electromyographic measurements

The EMG signals were measured from the VL, RF, and VM on the dominant leg during the SIZ1 and SIZ2 runs. Prior to electrode placement, the skin at each site was shaved, carefully abraded, and cleaned with alcohol. A bipolar surface electrode (circular 24 mm, Kendall disposable EMG electrodes, Covidien LTD; Gosport Hampshire,UK) arrangement (2.0 cm center-to-center) was placed based on the recommendations from the SENIAM Project for EMG electrodes placement[31]. Specifically, a reference line was drawn over the VL, 66% of the distance between the anterior superior iliac spine (ASIS) and the lateral superior border of the patella. In addition, the electrode-placement site was located 5 cm lateral to the reference line so that the electrodes were over the VL muscle[32]. A goniometer (Smith & Nephew Rolyan, Inc., Menomonee Falls, WI) was used to orient the EMG electrodes at a 20° angle to the reference line to approximate the pennation angle of the muscle fibers for the VL[33]. The electrodes the RF were placed on the midline of the segment at 50% of the distance between the ASIS of the pelvis and the superior part of the patella[31]. For the VM, the electrodes were placed 80% of the distance between the ASIS and the joint space in front of the anterior border of the medial collateral ligament and oriented at a 53° angle to approximate the pennation angle of the muscle fibers[31]. The reference electrode was placed over the iliac crest. The EMG signal was amplified (gain: ×1,000) using differential amplifiers (Free EMG 300, BTS, Milan, Italy, bandwidth = 10-500 Hz).

Signal processing

The raw EMG signals were digitized at 1,000 Hz, stored in a personal computer (MacBook Pro OSX, version 10.6.8, Apple Inc., Cuperino, CA) for subsequent analysis and processed with custom program, written with LabVIEW programming software (version 7.1, National Instruments, Austin, TX). The EMG signals were bandpass-filtered (zero phase shift, fourth-order Butterworth) at 10-500 Hz. Continuous 10 s epochs for the EMG AMP (microvolts root mean square, µVrms) and EMG MPF (MPF in Hz) were calculated. For the MPF analyses, each data segment was processed with a Hamming window and a discrete Fourier transform (DFT) algorithm in accordance with the recommendations of Hermens et al.[31]. The MPF was selected to represent the power spectrum on the basis of the recommendations of Hermens et al.[31] and was calculated as described by Kwatny et al.[34].

Statistical analyses

Mean differences among V˙O2peak and V˙O2 at exhaustion for the 4 constant velocity runs were examined using a one-way repeated measures ANOVA and Bonferroni corrected pairwise comparisons. Analyses were performed for the composite (defined as the mean of all of the subjects) and individual neuromuscular responses during the SIZ1 and SIZ2 runs. The neuromuscular responses were normalized as a % change from the initial 10% values to observe the pattern of responses over time. The first 10% of the EMG signals was omitted from analyses to account for changes in neuromuscular activity as the subject adjusted to treadmill velocity. Time was normalized as a percentage of the Tlim to account for the differences in times to exhaustion among the participants[2] and 10 data points were used in the analyses (10, 20, 30, 40, 50, 60, 70, 80, 90, and 100% of Tlim). Separate polynomial regression analyses (linear, quadratic, or cubic) were used for the composite and individual data to determine the normalized (% change from the initial 10% values) EMG AMP and EMG MPF responses versus %Tlim (10-100%) for the VL, RF, and VM during constant velocity runs in SIZ1 and SIZ2. For the individual and composite, normalized polynomial regression analyses, a positive relationship was defined by a 100% Tlim value above baseline and a negative relationship was defined by a 100% Tlim value below baseline. Twelve separate one-way repeated measures (RM) ANOVAs 1 (neuromuscular measurement (EMG AMP and MPF for the VL, RF, and VM)) x 10 (%Tlim for SIZ1 and SIZ2 runs (10, 20, 30, 40, 50, 60, 70, 80, 90, 100)) were used to examine changes in the neuromuscular parameters at every 10% of Tlim. Post-hoc Student Newman-Keuls (SNK) tests were then used to determine the time course of changes among the variables. The SNK test was chosen because it is designed to analyze the time course of changes in repeated measure variables[35-38]. All statistical analyses were performed with Statistical Package for the Social Sciences software (v.19.0. IBM SPSS Inc., Chicago, Illinois, USA).

Results

Characteristics of SIZ1 and SIZ2

Table 1 includes the mean (± SD) and range of values for the subject demographics as well as the V˙O2 and velocity variables from the GXT, CV test, SIZ1 and SIZ2 runs. The mean (± SD) V˙O2 values at exhaustion for the 4 runs used to determine the CV were, 2.97±0.90 L·min-1, 3.05±0.94 L·min-1, 3.11±0.98 L·min-1, and 3.19±1.02 L·min-1, respectively, from the lowest to highest velocity. The one-way RM ANOVA indicated there were significant differences (F=12.406, p<0.001, η2p=0.580) among the V˙O2peak and V˙O2 values at exhaustion for the 4 constant velocity runs. The follow-up pairwise comparisons indicated the V˙O2 values at exhaustion for the lowest two velocities were significantly less (p=0.019 and p=0.018) than V˙O2peak (3.28±1.02 L·min-1), but the V˙O2 values at exhaustion for the higher two velocities were not different (p=0.371 and p=0.103) from V˙O2peak.

Table 1.

Mean (± SD) and range of values for the subject demographics as well as the V˙O2, velocity, and times to exhaustion (Tlim) from the graded exercise test (GXT), critical velocity (CV) test, and severe intensity zones (SIZ1 and SIZ2) runs.

Variable Mean ± SD Range
Age (yrs) 23.2 ± 3.3 19 – 28
Body Mass (kg) 69.1 ± 13.2 54.9 – 92.5
Height (cm) 174.1 ± 8.0 162.6 – 190.5
V˙O2peak (L·min-1) 3.28 ± 1.02 2.14 – 5.38
vV˙O2peak (km·hr-1) 15.2 ± 1.4 13.5 – 18.0
CV (km·hr-1) 12.1 ± 1.5 9.8 – 14.3
50%∆ (km·hr-1) 13.7 ± 1.4 11.6 – 15.7
SIZ1 (km·hr-1) 13.2 ± 1.5 10.6 – 15.1
SIZ1 V˙O2end (L·min-1) 2.97 ± 0.90* 1.84 – 5.00
SIZ1 (%∆) 34 ± 8 21 – 43
SIZ1 (%CV) 109 ± 3 105 - 113
SIZ1 Tlim (min) 17.73 ± 2.63 14.06 – 22.23
SIZ2 (km·hr-1) 15.0 ± 1.6 12.6 – 17.4
SIZ2 V˙O2end (L·min-1) 3.19 ± 1.02 1.98 – 5.41
SIZ2 (%∆) 96 ± 17 70 – 124
SIZ2 (%CV) 124 ± 5 119 - 132
SIZ2 Tlim (min) 6.57 ± 0.78 5.98 – 7.92
Open in a new tab
*

Significantly lower than the V˙O2peak from the GXT (p=0.019). ∆= the difference between the CV and velocity at V˙O2peak (vV˙O2peak)

Composite Responses in SIZ1

The SIZ1 (86±5% vV˙O2peak) run occurred at 34±8 %∆ (109±3 %CV) and the Tlim was 17.73±2.63 min. The one-way RM ANOVAs indicated there were significant differences among time points for the VL EMG AMP (F=3.920, η2p=0.303, p=0.028) and RF EMG MPF (F=2.469, η2p=0.215, p=0.015), but no significant differences among time points for the VL EMG MPF (F=1.767, η2p=0.164, p=0.163), RF EMG AMP (F=1.431, η2p=0.137, p=0.266), VM EMG AMP (F=0.880, η2p=0.089, p=0.530), and VM EMG MPF (F=0.703, η2p=0.072, p=0.514). The composite polynomial regression analyses for the VL indicated there was a positive, linear relationship for EMG AMP (R2=0.846, p<0.001), that significantly increased from the 10% time point at 100% of Tlim, and a negative, linear relationship for EMG MPF (R2=0.656, p=0.004), but no differences between any of the time points and the initial 10% time point (Figure 1). For the RF, there was a positive, linear relationship for EMG AMP (R2=0.847, p<0.001), but no differences between any of the time points and the initial 10% time point, and a negative, linear relationship for EMG MPF (R2=0.852, p<0.001), which significantly decreased from the 10% time point at 80% and continued to 100% of Tlim (Figure 1). For the VM, there was a negative, cubic relationship for the EMG AMP (R2=0.848, p=0.013) and a negative, quadratic relationship for the EMG MPF (R2=0.648, p=0.024), but no differences between any of the time points and the initial 10% time point for either the AMP or MPF (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Time course of changes and polynomial regression analyses for the composite electromyographic amplitude (EMG AMP) and mean power frequency (MPF) responses of the superficial quadriceps muscles versus time to exhaustion (Tlim) for the severe intensity zone 1 (SIZ1) and 2 (SIZ2). The EMG signals were normalized as a percent change from the 10% time point and time was normalized as a percent of the Tlim. *Indicates starting of significant difference from the initial 10% time point. +Indicates ending of significant difference from the initial 10% time point.

Composite responses in SIZ2

The SIZ2 (98±4% vV˙O2peak) occurred at 96±17%∆ (124±5 %CV) and the Tlim was 6.57±0.78 min. The one-way RM ANOVAs indicated there were significant differences among time points for the VL EMG AMP (F=2.306, η2p=0.204, p=0.024) and VL EMG MPF (F=2.958, η2p=0.247, p=0.006), but no significant differences among time points for the RF EMG AMP (F=3.001, η2p=0.250, p=0.094), RF EMG MPF (F=1.954, η2p=0.178, p=0.180), VM EMG AMP (F=1.935, η2p=0.177, p=0.077), and VM EMG MPF (F=0.844, η2p=0.086, p=0.541). The polynomial regression analyses indicated there was a positive, quadratic relationship for the VL EMG AMP (R2=0.684, p=0.019), but no differences between any of the time points and the initial 10% time point, and a negative, cubic relationship for VL EMG MPF (R2=0.844, p=0.01), significant at 100% of Tlim (Figure 1). For the RF, there was a positive, cubic relationship for the EMG AMP (R2=0.953, p=0.021) and negative, linear relationship for the EMG MPF (r2=0.805, p<0.001), but no differences between any of the time points and the initial 10% time point for either the AMP or MPF. For the VM, there was a negative, quadratic relationship for the EMG AMP (R2=0.645, p–0.017), but no differences between any of the time points and the initial 10% time point, and no significant relationship for EMG MPF (p=0.604) (Figure 1).

Individual responses in SIZ1 and SIZ2

The order of the model fit (1st= linear, 2nd= quadratic, 3rd= cubic) as well as the r2 or R2 values for the VL, RF, and VM neuromuscular responses for each subject during the SIZ1 and SIZ2 runs are presented in Tables 2 and 3. The EMG AMP and EMG MPF responses versus Tlim relationships were plotted for each subject for the SIZ1 and SIZ2 runs and are presented in Figures 2 and 3.

Table 2.

The order (1st=linear, 2nd=quadratic, 3rd=cubic) and coefficients of determination (r[2] or R[2]) for the individual normalized (%change from baseline) polynomial regression analyses of the electromyographic amplitude (EMG AMP) and mean power frequency (MPF) responses of the superficial quadriceps muscles versus time to exhaustion for the severe intensity zone 1 run (SIZ1). The values are presented as order / r[2] or R[2].

SIZ1
Vasuts Lateralis Rectus Femoris Vastus Medialis
Subject EMG AMP EMG MPF EMG AMP EMG MPF EMG AMP EMG MPF
1 2nd / 0.855 2nd/ 0.914 2nd / 0.863 2nd / 0.846 NS NS
2 1st / 0.439 1st / 0.569 2nd / 0.559 3rd / 0.815 NS 1st / 0.688
3 3rd / 0.940 NS NS NS NS NS
4 NS 1st / 0.521 2nd / 0.800 NS NS 1st / 0.444
5 2nd / 0.904 NS 2nd / 0.920 2nd / 0.852 NS 2nd / 0.926
6 NS NS 1st / 0.637 3rd / 0.819 NS 1st / 0.487
7 2nd / 0.724 NS 2nd / 0.762 1st / 0.788 NS 2nd / 0.868
8 1st / 0.631 NS 1st / 0.597 3rd / 0.688 NS NS
9 NS NS NS 1st / 0.903 NS NS
10 NS NS NS 3rd / 0.872 NS NS
Open in a new tab

Bolded values indicate a positive relationship where the final value (100% of time to exhaustion) was above baseline. Italicized indicate a negative relationship negative where the final value (100% of time to exhaustion) was below baseline. NS = no significant relationship.

Table 3.

The order (1st=linear, 2nd=quadratic, 3rd=cubic) and coefficients of determination (r[2] or R[2]) for the individual normalized (%change from baseline) polynomial regression analyses of the electromyographic amplitude (EMG AMP) and mean power frequency (MPF) responses of the superficial quadriceps muscles versus time to exhaustion for the severe intensity zone 2 run (SIZ2). The values are presented as order / r[2] or R[2].

SIZ2
Vasuts Lateralis Rectus Femoris Vastus Medialis
Subject EMG AMP EMG MPF EMG AMP EMG MPF EMG AMP EMG MPF
1 NS 1st / 0.449 1st / 0.737 1st / 0.849 1st / 0.604 NS
2 NS NS NS NS NS 3rd/ 0.665
3 NS 1st / 0.453 3rd / 0.919 1st / 0.758 NS NS
4 NS NS 2nd / 0.924 3rd / 0.824 3rd / 0.962 1st / 0.723
5 1st / 0.745 1st / 0.504 1st / 0.809 1st / 0.700 3rd / 0.668 3rd /0.971
6 2nd / 0.663 NS NS 3rd / 0.819 2nd / 0.691 2nd / 0.805
7 NS NS NS 1st / 0.788 NS NS
8 NS NS NS 3rd / 0.688 NS NS
9 1st / 0.565 NS 2nd / 0.796 1st / 0.903 1st / 0.435 NS
10 1st / 0.461 NS 2nd / 0.956 3rd / 0.872 3rd / 0.663 NS
Open in a new tab

Bolded values indicate a positive relationship where the final value (100% of time to exhaustion) was above baseline. Italicized indicate a negative relationship negative where the final value (100% of time to exhaustion) was below baseline. NS = no significant relationship.

Figure 2.

Figure 2

Open in a new tab

Polynomial regression analyses for the individual electromyographic amplitude (EMG AMP) and mean power frequency (MPF) responses of the vastus lateralis (VL), rectus femoris (RF), and vastus medialis (VM) versus time to exhaustion (Tlim) for the severe intensity zone 1 (SIZ1). The EMG signals were normalized as a percent change from the 10% time point and time was normalized as a percent of the Tlim. See Table 2 for a description of the order (linear, quadratic, cubic) of the relationship. A solid line indicates a significant relationship and a dotted line indicates no significant relationship.

Figure 3.

Figure 3

Open in a new tab

Polynomial regression analyses for the individual electromyographic amplitude (EMG AMP) and mean power frequency (MPF) responses of the vastus lateralis (VL), rectus femoris (RF), and vastus medialis (VM) versus time to exhaustion (Tlim) for the severe intensity zone 2 (SIZ2). The EMG signals were normalized as a percent change from the 10% time point and time was normalized as a percent of the Tlim. See Table 3 for a description of the order (linear, quadratic, cubic) of the relationship. A solid line indicates a significant relationship and a dotted line indicates no significant relationship.

Discussion

Characteristics of SIZ1 and SIZ2

This study examined the neuromuscular responses for the superficial quadriceps muscles during exhaustive treadmill running in two distinct zones[20,21] within the severe intensity domain, SIZ1 and SIZ2. These zones are defined by specific metabolic responses, where V˙O2 may (SIZ2) or may not (SIZ1) reach V˙O2peak at exhaustion[20,21]. In this study, the V˙O2 at exhaustion for the SIZ1 run was less than V˙O2peak and was performed at 34±8% ∆, while the V˙O2 at exhaustion for the SIZ2 run, was not different from V˙O2peak and was performed at 124±5% CV. These findings were consistent with the V˙O2 responses and relative intensities (%∆ and CV) previously used20 to define the SIZ1 and SIZ2 and indicated the neuromuscular parameters examined in this study reflected responses from two distinct zones (SIZ1 and SIZ2) within the severe intensity domain.

Composite responses in SIZ1 and SIZ2

We hypothesized the EMG AMP and EMG MPF responses for each muscle would be consistent with neuromuscular fatigue in both zones, but the time course of changes analyses would indicate an earlier onset of neuromuscular fatigue in the SIZ2 than SIZ1, and the RF, a bi-articular muscle, would demonstrate greater fatigue than the VL and VM. The current findings, however, did not fully support this hypothesis. The patterns of responses in EMG AMP and EMG MPF for the VL and RF were consistent with neuromuscular fatigue[5,15,37] in the SIZ1 and SIZ2, but those for the VM were not (Figure 1). In addition, although the RF tended to demonstrate greater fatigue, the time course of changes analyses identified a limited number of significant differences among time points, likely as a result of large inter-individual variability in response to fatiguing treadmill running (Tables 2 and 3; Figures 2 and 3).

The dissociation in neuromuscular responses for the superficial quadriceps muscles is consistent with constant power output cycle ergometry[16] and treadmill running[24] as well as incremental treadmill running[14], however, linear increases in EMG AMP for all three superficial quadriceps muscles have also been reported for incremental treadmill running[17]. In the present study, the VL and RF demonstrated similar motor unit activation strategies (positive, linear relationships for EMG AMP in SIZ1 and positive, cubic and quadric relationships in SIZ2). The RF, however, demonstrated greater fatigue (significant decreases in EMG MPF from 80–100% Tlim) in SIZ1 compared to the VL or VM, while the RF and VL demonstrated a similar level of fatigue in the SIZ2 (Figure 1). The patterns of responses for the VL and RF in the SIZ1 and SIZ2 likely indicated fatigue induced increases in motor unit activation (recruitment and/or firing rate; EMG AMP) and a slowing of the motor unit action potential conduction velocity (EMG MPG)[5,6]. For the RF in particular, the negative responses in EMG MPF (motor unit action potential) corresponded to the positive responses in EMG AMP (motor unit activation) at 60% of Tlim in the SIZ2. The VM, however, did not demonstrate the same fatigue characteristics as the VL and RF, as evidenced by the cubic relationship for EMG AMP and quadratic relationship for EMG MPF that remained near baseline during the exhaustive run in SIZ1 and the non-significant relationship for EMG MPF in SIZ2. These findings suggested treadmill running at a 0% grade may be more dependent on the VL and RF for force production than the VM. Thus, the current findings indicated there were muscle specific responses to exhaustive treadmill running. The differences in fatigability among muscles may be related to the muscle architecture and fiber type characteristics[18,19] of the VL, RF, and VM.

The biarticular nature of the RF and its role in both thigh flexion and leg extension during treadmill running as well as its fiber type characteristics likely explain its greater fatigability[18,19] during the SIZ1 run. In fact, there is evidence from magnetic resonance imaging that ~74% of the total muscle volume of the RF was activated during horizontal treadmill running, while only 53% of the monoarticular vastus muscle group (VL, VM and vastus intermedius) contributed to force production[25]. Furthermore, the RF is characterized by a higher percentage, relative to the vastus muscle group[18,19], of fast-twitch glycolytic fibers, which rely heavily on anaerobic energy production and would contribute to greater metabolic byproduct accumulation[8]. Thus, the muscle specific responses and greater fatigability of the RF to exhaustive treadmill running in the SIZ1 are likely explained by differences in muscle architecture (biarticular versus monoarticular) and fiber type characteristics between the RF and vastus muscle group.

Previous studies have demonstrated intensity[2,22] specific neuromuscular responses to dynamic, constant power output exercise in the severe domain. In the present study, there were differences in the patterns of responses between exhaustive runs performed in the SIZ1 and SIZ2. Specifically, changes in the nature of the relationships for EMG AMP and EMG MPF occurred at ~60% Tlim for RF and VL during the SIZ2 run (Figure 1), versus a positive, linear relationship from the start of the SIZ1 run (Figure 1). The divergent patterns of responses between the intensity zones may be a reflection of the competing influences of metabolic byproducts and decreases in the global motor unit firing rate as a result of difference in total time accumulated at each epoch and the relative intensity at the start of the run. For the SIZ2 run, 60% of Tlim reflected ~4 minutes of exercise, compared with ~10.5 minutes in SIZ1. During the initial 60% of Tlim (~4 min) in the SIZ2, it is possible the decrease in the EMG AMP reflected a decrease in the global motor unit firing rate from the recruitment of higher threshold motor units with lower firing rates[39,40]. In addition, the shorter duration of exercise in SIZ2 likely allowed for less time for byproduct accumulation to alter motor unit activation or the action potential conduction velocity[5,15]. These findings indicated intensity specific neuromuscular patterns of responses within and between muscles of the superficial quadriceps. Therefore, the current findings indicated neuromuscular fatigue should be examined and characterized on a muscle and intensity specific basis within the severe intensity domain.

Inter-individual variability

We hypothesized there would be variability in the individual patterns of responses (linear, quadratic, and cubic) that may not reflect the composite model, but most subjects would demonstrate characteristics of neuromuscular fatigue during exhaustive exercise in the SIZ1 and SIZ2. The presence of inter-individual variability in neuromuscular responses has previously been reported for variable power output cycle ergometry[12], incremental treadmill running[17], and constant velocity running[24]. For example, during variable power output cycle ergometry, only 19 to 56% and 25 to 31% of the subjects demonstrated the same patterns of responses as the composite for EMG AMP and EMG MPF, respectively. In addition, Zuniga and Malek[17] reported that 78% of the individual responses reflected the composite responses for EMG AMP versus running velocity during incremental exercise, but all of the subjects demonstrated a positive relationship. In addition, it was reported that the individual patterns of responses were consistent across muscles[17]. The current findings, however, demonstrated greater inter-individual variability in response for EMG AMP and EMG MPF than previously reported[12,17,24] and individuals did not demonstrate consistent patterns across muscles (Tables 2 and 3; Figures 2 and 3). In this study, 10 to 20% of the subjects for the VL and RF, and none of the subjects for the VM, demonstrated the same patterns of responses as the composite for EMG AMP in the SIZ1 and SIZ2. There was, however, an overall increase in EMG AMP for 20 to 60% of the subjects for the VL, RF, and VM in both zones. For the EMG MPF, 10 to 20% of the subjects for the VL, RF, and VM in the SIZ1 and 40 to 60% of the subjects for the RF and VM, but none of the subjects for the VL, in the SIZ2 demonstrated the same patterns of responses as the composite. There was an overall negative relationship for 20 to 60% of the subjects for the VL, RF, and VM in the SIZ1 and SIZ2. These findings indicated there were dissociations between the composite and individual responses and support the recommendation of others[12,23] that in conjunction with composite results, EMG signals should be examined on a subject-by-subject basis. The greater inter-individual variability in this study compared to cycle ergometry and incremental treadmill running may be due to a number of factors including changes in stride length/frequency[24], muscle activation patterns that differ between individuals[24] as well as the influence of training status of the subject on muscle activation[41]. The large inter-individual variability reported in this study indicated that individual responses should be reported and considered in conjunction with composite findings to examine neuromuscular fatigue and make inferences about motor control strategies.

Summary

In the current study, the composite patterns of responses in EMG AMP and EMG MPF for the VL and RF were consistent with neuromuscular fatigue in the SIZ1 and SIZ2, but those for the VM were not. Thus, treadmill running may be more reliant on the VL and RF than the VM during fatiguing runs at 0% grade. In addition, the composite as well as the individual responses for the RF indicated greater neuromuscular fatigue than the VL or VM. The biarticular nature and fiber type characteristics of the RF may explain the earlier onset of neuromuscular fatigue compared with the VL and VM, especially at the lower severe intensity (SIZ1). In addition, the composite patterns of responses for each muscle differed between the SIZ1 and SIZ2, but the time course of changes analyses identified a limited number of significant differences among time points, likely as a result of large inter-individual variability in response to fatiguing treadmill running. Taken together, the current findings support the examination and characterization of neuromuscular fatigue on an intensity, muscle, and subject-by-subject basis.

There are several limitations in the current study regarding the ability to address the specific nature of the inter-individual variability in neuromuscular fatigue. This study examined a combined sample of men and women and did not include direct comparisons between sexes. In addition, the subjects in this study were all of a similar training status and there were no assessments of biomechanical differences in running technique or changes across fatigue levels. Furthermore, this study did not include any direct assessments of muscle architecture or size. Therefore, future studies should compare the neuromuscular fatigue responses between men and women to determine if there are sex differences in the onset of fatigue during severe intensity runs. In addition, these responses should be examined across a range of fitness levels and include biomechanical (e.g., stride length and frequency) and muscle architecture assessments to determine the effects of these factors on the rate of neuromuscular fatigue and their contribution to the inter-individual variability observed in this study.

Acknowledgements

HCB was a substantial contributor to data acquisition, analysis, and interpretation, and drafting the manuscript, was the primary author, and takes responsibility for the integrity of the data analysis. TKD, MTB, and PJS helped carry out data analysis and interpretation. KCS and NDMJ helped with data acquisition, analysis, and interpretation of the data. TJH, RJS, and GOJ were substantial contributors to the conception and design of the work and data analysis and interpretation. All authors contributed to revising the work and approved the final submission of this manuscript.

There were no sources of external funding for this work.

Footnotes

Dr. Nathaniel Jenkins reports grants from the American Heart Association, National Strength and Conditioning Association Foundation, and from the Center for Integrative Research on Childhood Adversity under NIGMS Award Number P20GM109097, personal fees from serving as a book reviewer for Human Kinetics, and honorarium from the National Strength and Conditioning Association as a speaker at the 2019 Annual Meeting; all outside of the submitted work. The remaining authors have nothing to disclose.

Edited by: G. Lyritis

References

  • 1.De Luca CJ. Myoelectric manifestation of localized muscular fatigue in humans. Crit Rev Biomed Eng. 1984;11:251–279. [PubMed] [Google Scholar]
  • 2.Housh TJ, Perry SR, Bull AJ, Johnson GO, Ebersole KT, Housh DJ, et al. Mechanomyographic and electromyographic responses during submaximal cycle ergometry. Eur J Appl Physiol. 2000;83:381–87. doi: 10.1007/s004210000315. [DOI] [PubMed] [Google Scholar]
  • 3.Mastalerz A, Gwarek L, Sadowski J, Szczepanski T. The influence of the run intensity on bioelectrical activity of selected human leg muscles. Acta Bioeng Biomech. 2012;14:101–107. [PubMed] [Google Scholar]
  • 4.Basmajian JV, De Luca CJ. Muscles alive: their functions revealed by electromyography. Williams & Wilkins, Baltimore, MD. 1985 [Google Scholar]
  • 5.Beck TW, Housh TJ, Johnson GO, Cramer JT, Weir JP, Coburn JW, et al. Does the frequency content of the surface mechanomyographic signal reflect motor unit firing rates? A brief review. J Electromyogr Kinesiol. 2007;17:1–13. doi: 10.1016/j.jelekin.2005.12.002. [DOI] [PubMed] [Google Scholar]
  • 6.Enoka RM, Stuart DG. Neurobiology of muscle fatigue. J Appl Physiol. 1992;72:1631–48. doi: 10.1152/jappl.1992.72.5.1631. [DOI] [PubMed] [Google Scholar]
  • 7.Gladden LB. Lactate metabolism: a new paradigm for the third millennium. J Phyiol. 2004;558:5–30. doi: 10.1113/jphysiol.2003.058701. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.MacLaren DP, Gibson H, Parry-Billings M, Edwards RH. A review of metabolic and physiological factors. Exerc Sport Sci Rev. 1989;17:29–66. [PubMed] [Google Scholar]
  • 9.Robergs RA, Dwyer D, Astorino T. Recommendations for improved data processing from expired gas analysis indirect calorimetry. Sports Med. 2010;40:95–111. doi: 10.2165/11319670-000000000-00000. [DOI] [PubMed] [Google Scholar]
  • 10.Westerblad H, Allen DG, Lannergren J. Muscle fatigue: lactic acid or inorganic phosphate the major cause? News Physiol Sci. 17:17–21. doi: 10.1152/physiologyonline.2002.17.1.17. [DOI] [PubMed] [Google Scholar]
  • 11.Camic CL, Housh TJ, Johnson GO, Hendrix CR, Zuniga JM, Mielke M, Schmidt RJ. An EMG frequency-based test for estimating the neuromuscular fatigue threshold during cycle ergometry. Eur J Appl Physiol. 2010;108:337–345. doi: 10.1007/s00421-009-1239-7. [DOI] [PubMed] [Google Scholar]
  • 12.Cochrane-Snyman KC, Housh TJ, Smith CM, Hill EC, Jenkins NDM, Schmidt RJ, Johnson GO. Inter-individual variability in the patterns of responses for electromyography and mecanomyography during cycle ergometry using an RPE-clamp model. Eur J Appl Physiol. 2016;116:1639–1649. doi: 10.1007/s00421-016-3394-y. [DOI] [PubMed] [Google Scholar]
  • 13.Hanon C, Thépaut-Mathieu C, Hausswirth JM, Le Chevalier JM. Electromyogram as an indicator of neuromuscular fatigue during incremental exercise. Eur J Appl Physiol. 1998;78:315–323. doi: 10.1007/s004210050426. [DOI] [PubMed] [Google Scholar]
  • 14.Hanon C, Thépaut-Mathieu C, Vandewalle H. Determination of muscular fatigue in elite runners. Eur J Appl Physiol. 2005;94:118–125. doi: 10.1007/s00421-004-1276-1. [DOI] [PubMed] [Google Scholar]
  • 15.Petrofsky JS. Frequency and amplitude analysis of the EMG during exercise on the bicycle ergometer. Eur J Appl Physiol. 1979;41:1–15. doi: 10.1007/BF00424464. [DOI] [PubMed] [Google Scholar]
  • 16.Housh TJ, deVries HA, Johnson GO, Housh DJ, Evans SA, Stout JR, et al. Electromyographic fatigue thresholds of the superficial muscles of the quadriceps femoris. Eur J Appl Physiol. 1995;71:131–136. doi: 10.1007/BF00854969. [DOI] [PubMed] [Google Scholar]
  • 17.Zuniga JM, Malek MH. 2013 Electromyographic responses of the superficial quadriceps femoris muscles during incremental treadmill running. Muscle Nerve. 2002;48:938–944. doi: 10.1002/mus.23842. [DOI] [PubMed] [Google Scholar]
  • 18.Edgerton VR, Smith JL, Simpson DR. Muscle fiber type populations of human leg muscles. Histochem J. 1975;7:259–266. doi: 10.1007/BF01003594. [DOI] [PubMed] [Google Scholar]
  • 19.Komi PV, Tesch P. EMG frequency spectrum, muscle structure and fatigue during dynamic contractions in man. Eur J Appl Physiol Occup Physiol. 1979;42:41–50. doi: 10.1007/BF00421103. [DOI] [PubMed] [Google Scholar]
  • 20.Bergstrom HC, Housh TJ, Cochrane KC, Jenkins NDM, Byrd MT, Schmidt RJ, Johnson GO. A model for identifying intensity zones above critical velocity. J Strength Cond Res. 2017;31:3260–3265. doi: 10.1519/JSC.0000000000001769. [DOI] [PubMed] [Google Scholar]
  • 21.Sawyer BJ, Morton RH, Womack CJ, Gaesser GA. O2max may not be reached during exercise to exhaustion above CP. Med Sci Sports Exerc. 2012;44:1533–38. doi: 10.1249/MSS.0b013e31824d2587. [DOI] [PubMed] [Google Scholar]
  • 22.Dinyer TK, Byrd MT, Cochrane-Snyman KC, Jenkins NDM, Housh TJ, Schmidt RJ, et al. Time course of changes in neuromuscular responses during rides to exhaustion above and below critical power. J Musculoskel Neuronal Inter 2019; In Press. [PMC free article] [PubMed] [Google Scholar]
  • 23.Farina D, Merletti R, Enoka R. The extraction of neural strategies from the surface EMG. J Appl Physiol. 2004;96:1486–1495. doi: 10.1152/japplphysiol.01070.2003. [DOI] [PubMed] [Google Scholar]
  • 24.Guidetti L, Gianfranco R, Francesco F. EMG patterns during running: Intra- and Inter-individual variability. J Electromyogr Kinesiol. 1996;6:37–48. doi: 10.1016/1050-6411(95)00015-1. [DOI] [PubMed] [Google Scholar]
  • 25.Sloniger MA, Cureton KJ, Prior BM, Evans EM. Lower extremity muscle activation during horizontal and uphill running. J Appl Physiol. 1997;83:2073–2079. doi: 10.1152/jappl.1997.83.6.2073. [DOI] [PubMed] [Google Scholar]
  • 26.Bergstrom HC, Housh TJ, Cochrane KC, Jenkins NDM, Buckner SL, Goldsmith JA, et al. Factors underlying the perception of effort during treadmill running above and below the critical heart rate. Eur J Appl Physiol. 2015;115:2231–2241. doi: 10.1007/s00421-015-3204-y. [DOI] [PubMed] [Google Scholar]
  • 27.Bergstrom HC, Housh TJ, Cochrane KC, Jenkins NDM, Buckner SL, Goldsmith JA, et al. Application of the critical heart rate model to treadmill running. J Strength Cond Res. 2015;29:2237–2248. doi: 10.1519/JSC.0000000000000869. [DOI] [PubMed] [Google Scholar]
  • 28.Robergs RA, Ghiasvand F, Parker D. Biochemistry of exercise-induced metabolic acidosis. Am J Physiol Regul Integr Comp Phyiol. 2004;287:R502–516. doi: 10.1152/ajpregu.00114.2004. [DOI] [PubMed] [Google Scholar]
  • 29.Bull AJ, Housh TJ, Johnson GO, Rana SR. Physiological responses at five estimates of critical velocity. Eur J Appl Physiol. 2008;102:711–720. doi: 10.1007/s00421-007-0649-7. [DOI] [PubMed] [Google Scholar]
  • 30.Pepper ML, Housh TJ, Johnson GO. The accuracy of the critical velocity test for prediciting time to exhaustion during treadmill running. Int J Sports Med. 1992;13:121–124. doi: 10.1055/s-2007-1021242. [DOI] [PubMed] [Google Scholar]
  • 31.Hermens HJ, Freriks B, Merletti R, Stegeman D, Blok J, Rau G, et al. SENIAM 8: European recommendations for surface electromyography: results of the SENIAM project. Enschede: Reoessingh Research and Development: 1999 [Google Scholar]
  • 32.Malek MH, Coburn JW, Weir JP, Beck TW, Housh TJ. The effects of innervation zone on the electromyographic amplitude and mean power frequency during incremental cycling. J Neurosci Methods. 2006;155:126–33. doi: 10.1016/j.jneumeth.2006.01.019. [DOI] [PubMed] [Google Scholar]
  • 33.Abe T, Kumagai K, Berchue WF. Fascicle length of leg muscles is greater in sprinters than distance runners. Med Sci Sports Exerc. 2000;32:1125–29. doi: 10.1097/00005768-200006000-00014. [DOI] [PubMed] [Google Scholar]
  • 34.Kwanty ED, Thomas DH, Kwanty HG. An application of signal processing techniques to the study of myoelectric signals. IEEE Trans Biomed Eng. 1970;17:303–313. doi: 10.1109/tbme.1970.4502758. [DOI] [PubMed] [Google Scholar]
  • 35.Abdi H, Williams LJ. Newman-keuls test and Tukey test. In: Salkind NJ, editor. Encyclopedia of research design. Thousand Oaks: Sage; 2010. [Google Scholar]
  • 36.Smith CM, Housh TJ, Jenkins NDM, Hill EC, Cochrane-Snyman KC, Miramonti A, et al. Combining regression and mean comparison to identify the time course of changes in neuromuscular responses during the process of fatigue. Physiol Meas. 2016;37:1993–2002. doi: 10.1088/0967-3334/37/11/1993. [DOI] [PubMed] [Google Scholar]
  • 37.Smith CM, Housh TJ, Herda TJ, Zuniga JM, Camic CL, Bergstrom HC, et al. Time course of changes in neuromuscular parameters during sustained isometric muscle actions. J Strength Cond Res. 2016;30:2697–2702. doi: 10.1519/JSC.0000000000001547. [DOI] [PubMed] [Google Scholar]
  • 38.Smith CM, Housh TJ, Hill EC, Schmidt RJ, Johnson GO. Time course of changes in neuromuscular responses at 30% versus 70% 1 repetition maximum during dynamic constant exercise resistance leg extensions to failure. Int J Exerc Sci. 2017;10:365–78. doi: 10.70252/WIVG1238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Contessa P, De Luca CJ, Kline JC. The compensatory interaction between motor unit firing behavior and muscle force during fatigue. J Neurophys. 2016;116:1579–85. doi: 10.1152/jn.00347.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.De Luca CJ, Erim Z. Common drive of motor units in regulation of muscle force. Trends in Neurosciences. 1994;17(7):299–305. doi: 10.1016/0166-2236(94)90064-7. [DOI] [PubMed] [Google Scholar]
  • 41.deVries HA. Efficiency of electrical activity as a physiological measure of the functional state of muscle tissue. Am J Phys Med. 1968;47:10–22. [PubMed] [Google Scholar]

Articles from Journal of Musculoskeletal & Neuronal Interactions are provided here courtesy of International Society of Musculoskeletal and Neuronal Interactions (ISMNI)

RESOURCES