Neuromuscular recovery from severe- and extreme-intensity exercise in men and women

Abstract

Maximal voluntary contraction force (MVC), potentiated twitch force (Q_pot), and voluntary activation (%VA) recover to baseline within 90s following extreme-intensity exercise. However, methodological limitations masked important recovery kinetics. We hypothesized reductions in MVC, Q_pot, and %VA at task failure following extreme-intensity exercise would be less than following severe-intensity exercise, and Q_pot and MVC following extreme-intensity exercise would show significant recovery within 120s but remain depressed following severe-intensity exercise. Twelve subjects (6 men) completed two severe-intensity (40, 50%MVC) and two extreme-intensity (70, 80%MVC) isometric knee-extension exercise bouts to task failure (T_lim). Neuromuscular function was measured at baseline, T_lim, and through 150s of recovery. Each intensity significantly reduced MVC and Q_pot compared to baseline. MVC was greater at T_lim (p<0.01) and at 150s of recovery (p=0.004) following exercise at 80%MVC compared to severe-intensity exercise. Partial recovery of MVC and Q_pot were detected within 150s following T_lim for each exercise intensity; Q_pot recovered to baseline values within 150s of recovery following exercise at 80%MVC. No differences in %VA were detected pre- to post-exercise or across recovery for any intensity. Although further analysis showed sex-specific differences in MVC and Q_pot, future studies should closely examine sex-dependent responses to extreme-intensity exercise. It is clear, however, that these data reinforce that mechanisms limiting exercise tolerance during extreme-intensity exercise recover quickly.

Introduction

Severe-intensity exercise has been well described as a domain of work rates that are associated with progressive metabolic and neuromuscular responses (Burnley et al. 2010; Monod and Scherrer 1965; Poole et al. 1988), such as V̇O₂, blood lactate, blood and intramuscular H⁺ and inorganic phosphate, which do not reach steady-state, but increase until task failure (i.e., inability to maintain target force or power) is reached (Burnley et al. 2010; Jones et al. 2008). Exercise tolerance in the severe-intensity domain can be expressed as a hyperbolic power-duration relationship when plotting time as a function of work rate, where the asymptote is termed critical power (CP), and the curvature constant is termed W’. This has resulted in a widely accepted interpretation that the work done in the severe-intensity domain (i.e., W’) is independent of work rate (Moritani et al. 1981; Poole et al. 1988). There is further evidence suggesting that the physiological mechanisms responsible for exercise cessation are likely peripheral, as contraction-induced metabolites associated with decreases in force production (e.g., inorganic phosphate and H⁺) are not different at task failure across severe-intensity work rates (Black et al. 2017). The power-duration relationship of the severe-intensity domain, however, only remains accurate for constant-load exercises which terminate at V̇O_2max, (i.e., where time to task failure (T_lim) is two minutes or longer (Hill et al. 2002; Monod and Scherrer 1965). Constant-load exercises performed at an intensity that would lead to task failure before V̇O_2max is able to be reached (i.e., T_lim less than two minutes) would be defined as supra-severe-intensity exercise, or the extreme-intensity domain (Hill et al. 2002). We have previously shown that T_lim during extreme-intensity exercise occurs earlier than predicted by the severe-intensity power-duration relationship (Alexander et al. 2019), and is characterized by a separate, smaller W’ (W’_ext). Although limited research exists regarding the physiological responses to extreme-intensity exercise, by utilizing the same interpretation of the severe-intensity domain, the W’_ext suggests that the mechanisms limiting extreme-intensity exercise, although likely different from those seen during severe-intensity exercise, are common throughout the extreme-intensity domain, and currently remain unknown.

Previous studies have shown that much of the loss in force at T_lim remains unrestored in the first few minutes (< 2 min) following prolonged exercise (> 15 min) (Baker et al. 1993; Sahlin and Seger 1995). Previous data from our lab extended this finding, showing reductions in maximal voluntary contraction force (MVC), potentiated twitch force (Q_pot; indicator of peripheral fatigue), and percent voluntary activation (%VA; indicator of central fatigue) were still evident 90 s into recovery following severe-intensity exercise compared to baseline values. However, no residual reductions in MVC, Q_pot, or %VA compared to baseline were found 90 s into recovery following extreme-intensity exercise (Alexander et al. 2019) suggesting that by then the muscles were fully recovered. Therefore, it is likely that the mechanisms limiting exercise tolerance also show differences in the kinetics during recovery from fatigue, being much faster following extreme-intensity exercise compared to severe-intensity exercise. Procedural limitations in our previous study masked the condition of the muscle at T_lim and throughout early recovery (90 s) following task failure. First, there was a short (< 30 s) delay while transferring subjects from the exercise ergometer to the force transducer following task failure. Second, the protocol we utilized recommended discarding the first 2 neuromuscular measurements (at 30 and 60 s) due to incomplete potentiation (Kufel et al. 2002). This resulted in our first data point being 90 s into recovery. From these results, it became clear that in order to determine the contributions of peripheral and central factors limiting extreme-intensity exercise tolerance, neuromuscular measurements would need to be made at task failure and throughout early recovery.

It is well established that the relative contributions to fatigue by peripheral and central factors are intensity-dependent (see refs: (Burnley and Jones 2018)). These differences may also be apparent with maximal exercise, including extreme-intensity exercise (Alexander et al. 2019; Hill et al. 2002). However, to date, no studies have compared the relative neuromuscular contributions to fatigue in the context of severe- and extreme-intensity domains. Therefore, the purpose of this study was to examine neuromuscular conditions immediately following task failure and to investigate the immediate rate of recovery following severe- and extreme-intensity exercise. We hypothesized that 1) the changes in MVC, Q_pot, and %VA immediately following task failure of extreme-intensity exercise would be significantly less than following severe-intensity exercise, and 2) Q_pot and MVC following extreme-intensity exercise would begin to recover within 2 minutes but remain depressed following severe-intensity exercise. In order to make measurements at T_lim and early into recovery, we moved from an isotonic knee-extension ergometer to one incorporating isometric contractions, thus permitting us to measure the relevant neuromuscular responses immediately at task failure and early recovery. Finally, although the main purpose of this study was not to examine specific differences between men and women, we tested for a main effect of sex to determine if the groups should be analyzed separately.

Methods

Subjects

Six men (mean ± S.D.: 23 ± 3 yr; 87.5 ± 16.1 kg; 179.3 ± 2.3 cm) and six women (mean ± S.D.: 22 ± 1 yr; 58.3 ± 3.4 kg; 158.8 ± 4.2 cm) participated in this study. All participants were free from cardiovascular, pulmonary, and metabolic disease as determined by a medical history questionnaire. Prior to participation in this study, subjects were informed of all procedures, and associated potential risks and benefits. Written informed consent was obtained from all participants prior to participation. Subjects were instructed to refrain from vigorous exercise 24 hours, alcohol consumption 12 hours, and food and caffeine 2 hours prior to each session, as well as to maintain current exercise habits in order to avoid any training or detraining effect. All research components were reviewed and approved by the Institutional Review Board of Human Subjects (#8369) at Kansas State University, Manhattan, KS.

Experimental Design:

Subjects visited the laboratory a minimum of 4 times, with at least 48 hours between sessions. On the first visit, subjects were familiarized with all testing procedures and equipment prior to testing. All exercise testing and measurements were performed on the right leg using a custom-built isometric knee extension ergometer and force transducer (LBG1, BLH Electronics, Waltham, MA, USA). Subjects were seated upright and the backrest was adjusted so that hip flexion angle was 105° and knee flexion angle was 90°. Force transducer position on the lower leg was adjusted for each subject such that a 90° angle of pull was maintained. The position of the force transducer and backrest setting were recorded and replicated for each subject and exercise sessions.

Constant-Load Tests:

During each subsequent visit, subjects performed one of 4 intermittent isometric constant-load tests to T_lim at a target force of 40, 50, 70, or 80% MVC in random order on separate visits. Based on our previous findings that intensities greater than ~ 60%MVC would elicit task failure in < 2 min (Alexander et al. 2019) reflecting extreme-intensity exercise, intensities were chosen to ensure two severe- (40 and 50% MVC, T_lim = 2 – 15 min) and two extreme- (70 and 80% MVC, T_lim < 2 min) intensities would be tested. MVC was measured prior to each exercise test to examine day-to-day variability. Subjects performed all knee-extension exercise at a 60% duty cycle (3 s contraction; 2 s relaxation) at a rate of 12 contractions/min as per a pre-recorded audio cue. Target force was displayed as a line generated on the display. Subjects were required to reach and maintain target force for the entire 3 s of contraction as indicated by visual and verbal feedback. Subjects were verbally encouraged throughout the test. Task failure was defined as the first failure to reach and maintain > 90 – 95% target force for > 66% (2 s) of the required contraction time. Upon task failure, the post-exercise neuromuscular assessment began at the subsequent audio cue.

Neuromuscular Function:

Neuromuscular function testing was conducted similar to previous extensively used protocols (Bigland-Ritchie et al. 1986; Broxterman et al. 2015; Burnley et al. 2012; Hammer et al. 2020). Briefly, testing was performed on the right leg prior to and immediately following each constant load exercise test. Adhesive electrodes (4 × 6 cm) were used to electrically stimulate the knee extensor muscles of the right leg via the femoral nerve. The anode was attached to the gluteal fold. The cathode was first positioned over the approximate location of the femoral nerve, and minor adjustments were made based on verbal feedback from the participants, who were instructed to report the greatest feeling of a knee-extension motion to remove activation of other thigh muscles. Once the cathode was placed over the approximate location of the femoral nerve, maximal stimulation was performed to ensure optimal placement. Maximal stimulation was assessed prior to each exercise bout. Stimulation intensity was initiated at 50 mA and was increased in 25 mA increments until the measured force and compound muscle action potential (M-wave) ceased to increase. The stimulator current was then increased an additional 30% to ensure the stimuli were supramaximal. This current averaged 575 mA across subjects, with a range of 400 – 800 mA. Force was sampled at 1000 Hz and displayed on a computer screen (LabVIEW, National Instruments, Austin, TX, USA). The femoral nerve was stimulated using a high-voltage constant-current electrical stimulator (DS7AH, Digitimer, Welwyn Garden City, UK). Paired stimuli (doublets) were delivered at 400 V with 100 μs square-wave pulse durations and a 10 ms pulse interval. Prior to each exercise test, subjects performed a series of 6, 3 s maximal voluntary contractions (MVCs), beginning every 30 s. Doublet muscle stimulations were delivered 5 s prior to each MVC, 1.5 s into the MVC, and 5 s after each MVC to obtain measurements of unpotentiated, superimposed, and potentiated doublet forces, respectively. MVC was determined as the greatest force attained prior to the superimposed muscle doublet stimulation. A series of 6 of these neuromuscular assessments was completed immediately upon task failure (T_lim) and for the next 2.5 min.

Data analysis:

Maximal voluntary contraction (MVC) was identified as the average force attained prior to the superimposed electrical stimulation. Potentiated twitch force (Q_pot) has been shown to increase after the first neuromuscular measurements, becoming maximal after the third neuromuscular measurement (Kufel et al. 2002), likely due to incomplete potentiation. Therefore, the average of the last three of six Q_pot forces during baseline neuromuscular measurements was used as baseline Q_pot, as done previously (Kufel et al. 2002). However, the post-exercise measurements were assumed to be fully potentiated as subjects were exerting maximal effort to reach target force at task failure. Therefore, individual neuromuscular measurements were analyzed as representing the current state of the muscle. In addition to investigating the change in absolute Q_pot, Q_pot was normalized relative to the decrease in force across exercise, i.e., baseline = 100%; force at end exercise = 0%.

Voluntary activation (%VA) was calculated using doublet interpolation corrected for when the superimposed stimulation did not occur during the MVC plateau force:

$$\%VA=\left[1-\left(\mathit{\text{force}}\mathit{\text{prior}}\mathit{\text{to}}\mathit{\text{superimposed}}\mathit{\text{doublet}}/MVC\right)*\left(\mathit{\text{superimposed}}\mathit{\text{doublet}}\mathit{\text{force}}/\mathit{\text{Qpot}}\right)\right]$$

Statistical Analysis:

All of the data were tested using a 3-way ANOVA (sex × time point in recovery × intensity) with 2 repeated measures (time point in recovery × intensity). There was a significant main effect for sex (p < 0.001), which suggested data should be separated by sex for subsequent analyses. The primary dependent variables analyzed were MVC, Q_pot, and %VA. Normal distribution was tested using the Shapiro-Wilk test and a logarithmic transformation was performed in the case of non-normality. A two-way ANOVA was used to test for potential training adaptations (i.e., day-to-day differences in baseline MVC) as well as between sexes (intensity × sex). Differences in MVC, Q_pot, and %VA were tested using two-way ANOVAs with repeated measures to compare 1) each work rate at each time point during recovery within men and women (work rate × time) and 2) at each time point during recovery between men and women within each work rate (time point during recovery × sex). Post hoc analyses using Tukey’s test were performed when appropriate Differences were considered statistically significant when p < 0.05. Data were reported as means ± standard deviation (SD) unless otherwise noted.

Results

Resting isometric MVC force among all subjects was 79.0 ± 17.2 kg. There were no significant day-to-day differences in MVC (p = 0.32). Mean ± SD force and T_lim between sexes for each intensity are shown in Table 1. Women had a lower MVC and resistances at each relative work rate than men (p = 0.002). Women exercised longer at 40% MVC compared to men (p < 0.001), but no other sex differences in T_lim were present.

Table 1.

Mean ± SD resistance and time to task failure (T_lim) for all intensities.

	Resistance (N)		Tlim (s)
	Men	Women	Men	Women
MVC	92.7 ± 11.4	65.4 ± 9.5 *
80% MVC	73.8 ± 8.8	51.9 ± 7.9 *	62 ± 15	72 ± 18
70% MVC	64.2 ± 9.2	45.7 ± 7.8 *	86 ± 16	134 ± 25
50% MVC	47.0 ± 5.7	33.9 ± 3.6 *	148 ± 24	190 ± 35
40% MVC	37.2 ± 4.2	25.5 ± 3.3 *	240 ± 55	350 ± 113 *

^*Different from men (p < 0.05).

Neuromuscular Function:

Maximal voluntary contraction (MVC):

When analyzed as a single group (n = 12) MVC was significantly lower at task failure compared to baseline following exercise intensities (p < 0.05). There was a significant interaction between MVC and time point in recovery (F_18,180 = 3.52, p < 0.001). MVC at task failure was not decreased to the same degree following exercise at 80% MVC compared to exercise at 40 and 50% MVC (p < 0.01). MVC remained depressed throughout 150 s of recovery following each exercise intensity. MVC was significantly higher at 90 s of recovery following severe-intensity exercise (40 and 50% MVC) compared to at task failure (p < 0.001), and at 60 s of recovery following exercise at 70% MVC (p = 0.004), demonstrating significant, albeit partial, recovery from fatigue. Maximal force following exercise at 80% MVC was not different from force at task failure until 150 s of recovery (p = 0.004), however, this likely results from less reduction in MVC at this intensity, suggesting less fatigue. Indeed, MVC was significantly greater at 150 s of recovery following exercise at 70 and 80% MVC compared to at 150 s of recovery following exercise at 40% MVC (p < 0.05).

MVC forces at baseline, task failure, and up to 150 s into recovery for both men and women are shown in Fig 1. Men had greater baseline MVC compared to women (Fig 1, p = 0.001). There was a main effect of sex following exercise at 40% (F_1,60 = 4.53, p = 0.05), 50% (F_{1, 60} = 7.90, p = 0.018), 70% (F_1,60 = 6.50, p = 0.029), and 80%MVC (F_1,60 = 11.12, p = 0.009), as well as a main effect of time point in recovery following exercise at 40% (F_6,60 = 36.45, p < 0.001), 50% (F_6,60 = 40.81, p < 0.001), 70% (F_6,60 = 28.26, p < 0.001), and 80%MVC (F_6,60 = 17.48, p < 0.001). MVC was significantly reduced in both men and women following each work rate (p < 0.05) and remained significantly lower than baseline values throughout 150 s of recovery (p < 0.05). Baseline force, force immediately following task failure, and force at 150 s of recovery in men and women are shown in Figure 2. There was a significant interaction between intensity and time point in recovery within men (F_18,90 = 1.99, p = 0.041) and within women (F_18,90 = 2.28, p = 0.006). Within men, no differences were found in MVC among intensities at task failure or after 150 s of recovery. Conversely, within women, exercise at 70 and 80% MVC did not decrease maximal force production to the same degree as exercise 40% MVC at task failure (p < 0.003) or after 150 s of recovery (p < 0.03).

Figure 1:

Absolute maximal voluntary contraction force (MVC) with Pre (baseline) and each Post-exercise measurement beginning immediately after task failure within each intensity. Men (●), Women (○). * Significantly lower than Pre-exercise value. α significantly different from task failure (0 s) within men, β Significantly different from task failure (0 s) within women, Φ Significantly different from men, p < 0.05.

Figure 2:

Absolute maximal voluntary contraction force (MVC) with Pre-exercise (baseline), immediately following task failure (0 s), and at 150 s of recovery in men and women. † Significantly greater than 40% MVC, p < 0.05.

Potentiated twitch force (Q_pot):

When analyzed as a single group (n = 12) absolute Q_pot was significantly decreased at task failure compared to baseline following all exercise intensities (p < 0.05). There was a significant interaction between intensity and time point in recovery (F_18,180 = 3.95, p <0.001). At task failure, Q_pot was not reduced to the same degree following exercise at 80% MVC compared to exercise at 50% MVC, with no other differences at task failure detected. Q_pot remained significantly lower than baseline throughout 150 s of recovery following each exercise intensity, although was not statistically different from baseline at 150 s of recovery following exercise at 80% MVC (p = 0.145). Further, Q_pot following extreme-intensity exercise (70 and 80% MVC) were greater after 150 s of recovery compared to following severe-intensity exercise (40 and 50% MVC; p < 0.01).

Absolute Q_pot forces in both men and women at baseline, task failure, and up to 150 s into recovery are shown in Fig 3, while Q_pot relative to the decrease in force caused by exercise are shown in Fig 4. There was a significant interaction between sex and time point in recovery following exercise at 40% (F_6,60 = 3.59, p = 0.004) and 50% (F_6,60 = 6.64, p < 0.001). There was a main effect of sex (F_1,60 = 6.34, p = 0.03) and time point in recovery (F_6,60 = 62.61, p < 0.001) following exercise at 70%MVC and a main effect of time point in recovery (F_6,60 = 40.74, p < 0.001) following 80% MVC. Men showed higher baseline Q_pot compared to women. Q_pot at task failure was significantly reduced from baseline for both men and women at all four work rates (p < 0.05), but no sex differences were detected at task failure following any work rate (p = 0.97, 0.41, 0.15, 0.37 after 40, 50, 70, and 80% MVC, respectively). Baseline force, force immediately following task failure, and force at 150 s of recovery in men and women are shown in Figure 5. There was a significant interaction between intensity and time point in recovery within men (F_18,90 = 2.61, p = 0.007) and within women (F_18,90 = 2.59, p = 0.002). Within men, exercise at 70% MVC did not decrease Q_pot to the same degree as exercise at 50% MVC at task failure (p < 0.044) or at 150 s of recovery following 40 or 50% MVC (p < 0.04). Within women, Q_pot was not different at task failure across work rates. Further, Q_pot following exercise at 80% MVC was not decreased to the same degree as following exercise at 40% or 50% MVC at 150 s of recovery (p < 0.05).

Figure 3:

Absolute potentiated twitch force (Q_pot) with Pre (baseline) and each Post-exercise measurement beginning immediately after task failure within each intensity. Men (●), Women (○). * Significantly lower than Pre-exercise value. α significantly different from task failure (0 s) within men, β Significantly different from task failure (0 s) within women, Φ Significantly different from men, p < 0.05.

Figure 4:

Potentiated twitch force (Q_pot) relative to baseline and end-exercise to examine changes in recovery relative to amount of fatigue accumulated during each intensity beginning immediately after task failure. 100% represents baseline Q_pot forces, whereas 0% represents Q_pot at task failure. Men (●), Women (○). α significantly different from task failure (0 s) within men, β Significantly different from task failure (0 s) within women, Φ Significantly different from men, p < 0.05.

Figure 5:

Absolute potentiated twitch force (Q_pot) with Pre-exercise (baseline), immediately following task failure (0 s), and at 150 s of recovery in men and women. † Significantly greater than 40% MVC, ‡ Significantly greater than 50% MVC, p < 0.05.

Voluntary activation (%VA):

No differences were found at baseline within men (93.5 ± 2.2%, p > 0.94), women (95.3 ± 3.2%, p > 0.99), or between sexes (p > 0.70). The %VA was not different from baseline at task failure following any of the four work rates in either sex, nor was %VA different across all four work rates within men (p > 0.08; 82.8 ± 11.0%, 87.3 ± 6.6%, 90.4 ± 12.1%, 92.71 ± 3.2% following exercise at 40, 50, 70, and 80% MVC, respectively), or within women (p > 0.84; 93.8 ± 5.3%, 89.6 ± 11.6%, 90.9 ± 10.3%, and 91.2 ± 3.6% following exercise at 40, 50, 70, and 80% MVC, respectively). Further, there were no differences in %VA throughout 150 s of recovery between men and women across each of the four work rates (p > 0.05).

Discussion

Major findings:

Our hypotheses were partially confirmed by these data. MVC and Q_pot at task failure were decreased less following extreme-intensity exercise (80% MVC) compared to severe-intensity exercise (40 and 50% MVC), consistent with Hypothesis #1. Further, MVC and Q_pot following extreme-intensity exercise (70 and 80% MVC) were significantly greater (faster partial recovery) after 150 s of recovery compared to following severe-intensity exercise (40 and 50% MVC), consistent with Hypothesis #2. However, in contrast to both Hypotheses 1 and 2, voluntary activation (%VA) did not significantly change from baseline after either severe or extreme exercise, nor was it different at 150 s of recovery.

In men, maximal force produced at task failure fell to a similar value across all intensities. However, women produced greater force (i.e., less decline) at task failure following extreme-intensity exercise (i.e., 70 and 80% MVC) compared to lower-intensity exercise (i.e., 40% MVC). Partial MVC force recovery was apparent in both sexes following severe-intensity exercise. However, no recovery in maximal voluntary force was detected in women following exercise at 70% MVC or in either sex after exercise at 80% MVC (Fig 1). Q_pot (indicator of peripheral fatigue) was reduced less at task failure following extreme-intensity exercise (70 and 80% MVC) compared to severe-intensity exercise (40 and 50% MVC) in men. However, Q_pot was reduced to a similar degree across all four work rates in women. Although peripheral fatigue was still apparent in men following all exercise intensities at 150 s of recovery, there was significant partial recovery within 60 s (Figs 3 & 4). Women showed a similar response in partial recovery following severe-intensity exercise, but reached forces similar to baseline values after extreme-intensity exercise, implying full recovery (Fig 3). Lastly, no statistical differences in %VA (indicator of central fatigue) were detected following any of the four work rates in men or women, nor across recovery following any of the work rates.

Recovery kinetics:

Although neuromuscular recovery following exercise has previously been reported (Alexander et al. 2019; Bigland-Ritchie et al. 1978; Senefeld et al. 2018), few studies have examined changes through the initial seconds/minutes immediately following exercise cessation. Recovery from fatigue occurs faster following task failure at intensities closer to MVC (Alexander et al. 2019) compared to lower intensities (Alexander et al. 2019; Baker et al. 1993; Sahlin and Seger 1995). Therefore, we hypothesized that Q_pot and MVC would recover faster following extreme-intensity exercise compared to severe-intensity exercise. Consistent with this, in women, Q_pot returned to baseline values within 2 minutes of recovery following exercise at 70% MVC and after one minute of recovery following exercise at 80% MVC, while Q_pot remained significantly reduced throughout 150 s of recovery following exercise at 40 and 50% MVC (Fig 3). This suggests that peripheral fatigue development following extreme-intensity isometric exercise is fully recovered in women in < 2 minutes, while still apparent after severe-intensity exercise. The current data also suggest that men do not fully recover to baseline values within 150 s following extreme-intensity exercise (Fig 3); however this is inconsistent with previous data from our lab (Alexander et al. 2019). It is likely that these differences are due to differences in mode of exercise, e.g., one- vs two-legged knee extension (Hureau et al. 2018) or dynamic vs isometric contractions (Babault et al. 2006; Kruger et al. 2019). Indeed, Senefeld et al. (2018) found isometric exercise decreased Q_pot to a greater degree than dynamic exercise (Senefeld et al. 2018). Further, they found that a 60 s sustained, maximal isometric knee-extension contraction reduced Q_pot to ~ 35% of baseline force, and had recovered to ~ 85% of baseline force by 150 s in both men and women. In addition to differences in modality, it is also possible that the task, and therefore task failure, is responsible for the difference in recovery responses. For example, the current study required subjects to continue exercise until the target force could not be reached, thus intermittently utilizing W’_sev or W’_ext. Conversely, sustained isometric contractions do not reach a “task failure” but allow force to decrease throughout the test until a pre-determined force or time end point is reached, while continuously utilizing W’_sev / W’_ext. However, our current data support the finding that, although recovery from isometric exercise may be slower than from dynamic exercise, much of the force loss is recovered in the immediate minutes after exercise cessation. Our current data further characterize the recovery of force throughout the 150 s immediately following exercise (Figs 1 and 4) and is consistent with other data that shows much of the loss of MVC and Q_pot force (Froyd et al. 2013; Gruet et al. 2014) and %VA (Felippe et al. 2020; Gruet et al. 2014) (indicating peripheral and central fatigue, respectively), is recovered in the first 150 s following exercise cessation.

Our data support the finding that recovery from higher intensity (i.e., extreme) exercise occurs faster than from lower intensity (i.e., severe-intensity) exercise. Although similar observations have been made regarding faster recovery from high frequency compared to low-frequency fatigue (Edwards et al. 1977; Jones 1996), the current study is the first to detail this finding in the context of exercise intensity domains. It is important to note in this context that the range of forces associated with severe-intensity exercise (40–50% MVC) and extreme-intensity exercise (70–80% MVC) in the current study and our previous one (severe 26–44% MVC; extreme 70–90% MVC) (Alexander et al. 2019) fall in the range of forces produced by low frequency (20 Hz) and high frequency (50 Hz) electrical stimulation, respectively (Jones 1996). Thus, our paradigm of severe- and extreme-intensity exercise domains, each with different W’, provides a novel approach with which to explore mechanisms of fatigue as related to the force-frequency relationship.

It is well accepted that physiological responses to exercise are intensity-domain dependent. Given the previous finding that the boundary between severe- and extreme-intensity domains occurs ~ 60% MVC (Alexander et al. 2019), previous studies have also showed profound differences in physiological responses during exercise. For example, there is evidence of limitations to blood flow due to intramuscular pressures causing contraction-induced occlusion beginning ~ 50 – 60% MVC (Barnes 1980). During intermitting contractions, muscle reperfusion occurs during the relaxation portion of the duty cycle. The pertinent question is if, under a specific set of circumstances, the duration of relaxation is sufficient to permit adequate blood flow to sustain aerobic metabolic needs. To the point, Broxterman et al. (2014) showed that increasing time of relaxation between contractions, while keeping power output constant, increased CP and T_lim during severe-intensity exercise, supporting the conclusion that increasing perfusion during the relaxation phase leads to increased exercise tolerance, at least for severe-intensity exercise (Broxterman et al. 2014). Further, Wernbom et al., (2006) found external occlusion during dynamic contractions to task failure decreased exercise tolerance at 20 – 40% MVC, but not at 50% MVC (Wernbom et al. 2006). This suggests that for continuous dynamic exercise at 50% MVC, contractions themselves can significantly occlude muscle blood flow (Wernbom et al. 2006). It is likely that the intramuscular pressures exerted during extreme-intensity exercise block blood flow during contractions, but it is currently unclear if the reperfusion occurring between contractions during extreme-intensity exercise increases in proportion to the demand of the muscle. While there are extensive numbers of studies describing metabolic and cardiovascular responses to severe-intensity exercise, little is known regarding these responses during extreme-intensity exercise. It is reasonable to speculate that the tissue demand for blood flow and O₂ delivery during extreme-intensity exercise would be greater than the incremental increase of blood flow and vascular conduction. Therefore, this mismatch of relaxation perfusion to exercising muscle demand may provide a potential explanation as to why V̇O_2max may not be reached before task failure occurs during extreme-intensity exercise (Hill et al. 2002). Further, limitations to O₂ delivery combined with the relatively high force requirements place a greater demand on alternative ATP resynthesis pathways, such as anaerobic glycolysis and greater utilization of phosphocreatine. Indeed, phosphocreatine (PCr) breakdown has been shown to increase nonlinearly at higher intensities (Barstow et al. 1994). In addition, PCr has a fast half-recovery time (~1 min) after maximal exercise (Bogdanis et al. 1995), thus potentially providing an explanation for a faster recovery in force seen following extreme-intensity exercise (Figs 1 and 4).

Indicators of central fatigue development have been found at maximal and near-maximal work rates (Bigland-Ritchie et al. 1978; Yoon et al. 2007). Although %VA was not significantly different in men or women in the current study, individual responses showed a wide range in central fatigue development across exercise intensities. Chartogne et al. (2020) found a negative correlation between central fatigue and peripheral fatigue development during 5-minute repeated MVCs (Chartogne et al. 2020). During severe-intensity exercise, however, Burnley et al. (2012) showed evidence of central fatigue at intensities closer to CP (~40 – 50% MVC), but %VA was not significantly reduced following exercise > 50% MVC (Burnley et al. 2012). Further, contraction-induced metabolite accumulation (e.g., increased blood and intramuscular H⁺, intramuscular inorganic phosphate) have been shown to increase until task failure during severe-intensity exercise (Black et al. 2017; Jones et al. 2008), leading to increased peripheral fatigue. Although no significant evidence of central fatigue was evident in the current study, it could be that a small, but not statistically detectable, contribution of central fatigue may be enough to cause a critical combination of fatiguing factors that limit exercise tolerance at extreme-intensity work rates causing task failure to occur earlier than predicted by the severe-intensity power-duration relationship, but which is also quickly reversible (Felippe et al. 2020; Froyd et al. 2013; Gruet et al. 2014).

Sex differences:

Consistent with previous studies, men produced higher MVC (Fig 1) and Q_pot forces (Fig 2) compared to women (Miller et al. 1993; Senefeld et al. 2018). In contrast, women were able to sustain exercise longer at 40 and 50% MVC; however, sex differences in exercise tolerance were not evident at 70 and 80% MVC, presumably during extreme-intensity exercise. This is consistent with the study by Maughan et al. (1986) who showed no significant differences in exercise endurance at 80 and 90% MVC between men and women, while women exercised significantly longer than men at 50 and 70% MVC (Maughan et al. 1986). In addition, Ansdell et al. (2017) found that women could exercise longer than men at 50% MVC (Ansdell et al. 2017). It has been suggested that women’s ability to exercise longer than men during severe-intensity exercise could be due to greater oxygen delivery (Russ and Kent-Braun 2003) and oxidative utilization (Ansdell et al. 2019; Russ and Kent-Braun 2003), which could in turn be explained by women developing less absolute muscle occlusion force during contractions than men for the same %MVC task (Barnes 1980).

Our data show that the magnitude of peripheral fatigue development measured at task failure was dependent on exercise intensity in men, but not in women. Women have been shown to have faster recovery from exercise compared to men (Ansdell et al. 2019; Senefeld et al. 2018). That these differences are less apparent after maximal exercise lends further support that blood flow delivery and/or intramuscular conditions common to both men and women (e.g., elevated H⁺, inorganic phosphate, and/or decreased phosphocreatine, ATP) contribute to the limited exercise tolerance of the extreme-domain compared to that predicted by the severe-intensity domain power-duration relationship (Ditor and Hicks 2000; Laforest et al. 1990).

Experimental considerations:

Several factors should be considered when interpreting the data from the current study. First, the boundary between the severe- and extreme-intensity domain has not yet been clearly defined. Although it is agreed that extreme-intensity exercise would results in task failure before V̇O_2max is reached (Hill et al. 2002), preliminary data from our lab showed inconsistent V̇O₂ responses to maximal knee-extension exercise, precluding V̇O₂ as a reliable demarcation. Therefore, work rates that resulted in task failure in under 2 minutes were considered extreme-intensity exercise. We previously showed that this occurred at ~ 60% MVC in men (Alexander et al. 2019), and therefore work rates at 70 and 80% MVC were chosen to increase the likelihood of being in the extreme-intensity exercise domain, and 40 and 50% MVC to be most likely in the severe-intensity domain. However, it is possible that 70% MVC may have been severe-intensity exercise for some individuals and 50% MVC may have been extreme-intensity exercise for others.

This study used absolute percentages of MVC for comparisons. It is possible that women at 40 and 50% MVC were on the lower end of their relative power-duration relationship (i.e., closer to critical torque) compared to men, therefore able to exercise longer before W’ was completely utilized (Ansdell et al. 2019). However, although women exercised longer, they also showed faster recovery following exercise. Future studies should match men and women for aerobic exercise capacity as well as MVC to better identify specific sex differences in the contributions of peripheral and central fatigue during and following extreme-intensity exercise.

Finally, the current study examined 12 subjects, consisting of 6 men and 6 women. Initially the primary purposes of this study were not to examine differences due to sex. However, because responses to exercise in women are largely lacking in relation to men, we recruited both sexes. Initial analyses were made with grouped data, but the initial 3-way ANOVA showed a significant main effect for sex. Therefore, we presented data separated by sex. However, the low n for each sex may have hidden other important sex differences. Future studies should be designed so as to appropriately evaluate potential differences between men and women. In addition, we were not able to ascertain intra-muscular metabolic changes and fiber type differences among the subjects, which might have provided further mechanistic insights.

Additionally, due to the multiple visits plus recovery days required for each subject to complete the study (minimum of 8 days), we did not control for menstrual cycle phase or hormonal contraceptive usage in the female participants. As shown by Ansdell et al. (2019), endogenous hormone concentrations can affect fatigability during intermittent isometric exercise in the range of work rates (60% MVC) investigated in the current study (Ansdell et al. 2019). If present, this could have introduced additional variability in the results, potentially masking true underlying differences in the women (type II error; e.g., in T_lim, MVC, Q_pot). Conversely, this increased variability could also be interpreted as strengthening the significant differences that were observed.

Conclusions:

That the physiological responses to exercise are intensity domain-dependent is well known. However, little is known about how the responses to extreme-intensity exercise compare with the other domains, specifically severe-intensity exercise. The current data showed that women were able to sustain severe-intensity exercise longer than men, but there were no differences in extreme-intensity exercise tolerance. It is also apparent that there are differences in the neuromuscular recovery characteristics immediately following exercise in these two intensity domains. Men and women both showed significant partial recovery following extreme-intensity exercise. However, while men showed evidence of residual peripheral fatigue 150 s into recovery, women appeared to be fully recovered from peripheral fatigue within 120 s after exercise cessation. These findings have important implications in the understanding the potential mechanisms contributing to intensity-dependent exercise tolerance and time course of recovery. Future study designs should be designed so as to appropriately evaluate potential differences between men and women. In addition, we were not able to ascertain intra-muscular metabolic changes and fiber type differences among the subjects, which might have provided further mechanistic insights.

NOVELTY:

• Severe- and extreme-intensity exercise cause independent responses in fatigue accumulation and the subsequent recovery time courses.

• Recovery of MVC and Q_pot occurs much faster following extreme-intensity exercise in both men and women.

Data availability statement:

All data supporting the results presented in the manuscript are included in the manuscript figures as values of n were ≤ 30. The data that support the findings of this study are available on request from the corresponding author.