Recovery and Fatigue Behavior of Forearm Muscles during a Repetitive Power Grip Gesture in Racing Motorcycle Riders

Despite a reduction in the maximal voluntary isometric contraction (MVCisom) observed systematically in intermittent fatigue protocols (IFP), decrements of the median frequency, assessed by surface electromyography (sEMG), has not been consistently verified. This study aimed to determine whether recovery periods of 60 s were too long to induce a reduction in the normalized median frequency (MFEMG) of the flexor digitorum superficialis and carpi radialis muscles. Twenty-one road racing motorcycle riders performed an IFP that simulated the posture and braking gesture on a motorcycle. The MVCisom was reduced by 53% (p < 0.001). A positive and significant relationship (p < 0.005) was found between MFEMG and duration of the fatiguing task when 5 s contractions at 30% MVCisom were interspersed by 5 s recovery in both muscles. In contrast, no relationship was found (p > 0.133) when 10 s contractions at 50% MVC were interspersed by 1 min recovery. Comparative analysis of variance (ANOVA) confirmed a decrement of MFEMG in the IFP at 30% MVCisom including short recovery periods with a duty cycle of 100% (5 s/5 s = 1), whereas no differences were observed in the IFP at 50% MVCisom and longer recovery periods, with a duty cycle of 16%. These findings show that recovery periods during IFP are more relevant than the intensity of MVCisom. Thus, we recommend the use of short recovery periods between 5 and 10 s after submaximal muscle contractions for specific forearm muscle training and testing purposes in motorcycle riders.


Introduction
Simulation of highly repetitive intermittent muscle contractions present during motorcycle competitions is currently under investigation because of their relationship with the development of clinically significant conditions, especially in the hand/forearm. These conditions, characterized by pain and loss of the hand or forearm function, are defined as exertional compartment syndrome [1][2][3][4]. They frequently lead to long periods of illness in motorcycle riders, especially in those participating in endurance competitions such as 24 h races, where they must brake more than 4000 times and make 10,000 gear changes [5]. Similar pathological patterns can also occur among workers in the manufacturing industry [6]. The fact that many athletes, manual workers, and musicians must endure their mechanical work over long periods of time, muscle contraction intensities that characterize each activity explains the large number of studies focused on neurophysiological fatigue of the forearm muscles [7][8][9][10][11]. These muscles are involved in a great variety of repetitive grip tasks that can lead to neuromuscular fatigue and functional impairment when these tasks become chronic. Thus, it is important to obtain better knowledge and understanding of the mechanisms involved in these physiological situations to prevent forearm syndrome.
When assessing human muscle fatigue with superficial electromyography (sEMG), the power spectrum displacement towards lower frequencies has been extensively documented in continuous fatiguing protocols (CFP), in which submaximal voluntary contractions are maintained until exhaustion [9,[12][13][14][15][16]. Intermittent fatiguing protocols (IFP) have also been extensively studied because intermittent contractions at different intensities are very common in the everyday life of the majority of workers and athletes [17,18]. Consequently, when comparing both types of fatiguing tasks (CFP versus IFP) specifically adapted to motorcycle riders [9], IFP showed a stronger relationship with the level of motorcyclist forearm discomfort compared to CFP [9].
The relative intensity of the contraction with respect to maximal voluntary isometric contraction (MVC isom ) registered in a non-fatigued condition (%MVC isom ) is a key factor that modulates muscle fatigue. Studies looking at CFP confirmed that the higher the intensity of the effort, the shorter the time to task failure [14,16,19], obviously because of the lack of recovery periods. Moreover, it has been generally observed that %MVC isom and the time to task failure (also called time limit) have a significant effect on the decrement of sEMG frequency (MF EMG ) and increment of the sEMG amplitude (RMS EMG ) [14,15,20]. A reduction in MF EMG was observed during CFPs at 10% MVC [21,22], 25% MVC [22][23][24], 30% [21], 40% [9,21,24], 60% [25], 55,70,80, and 90% of MVC [24]. Nevertheless, some caution is recommended in regard to MF EMG because %MVC isom should not be considered as a definitive factor explaining the absence of a reduction in MF EMG during fatiguing protocols [24,25].
A second factor that is necessary to consider when measuring fatigue is the duration of the effort or exertion time. It is known that the duration of fatiguing tasks at a constant relative submaximal %MVC isom is negatively associated with MVC isom decrements, reaching the maximal point at time to task failure [26]. Duration of the effort induces a linear decrement of MF EMG [14,24,27] whose slope may differ slightly depending on the muscle group and type of movement [15,16,20,21]. With %MVC isom and duration of the effort as the main triggers of fatigue in CFPs, the greatest MF EMG decrements were observed at longer durations due to lower %MVCs [28].
A third factor must be taken into account in IFP: the duration of the recovery interspersed between muscle contractions. Controversial MF EMG results have been observed when applying IFPs, despite the lower MVC isom recorded at the end of such fatigue protocols. For example, some authors, but not all [29,30], reported a reduction in MF EMG during an IFP [31,32]. MF EMG was similar to pre-fatigue values with different work-rest cycles, whatever the intensity used in the IFP, [22]. These results are consistent with the findings of Mundale [28], who also studied the factors that lengthen the endurance time of an IFP. It seems that the duration of the recovery period could be one of the key factors explaining the disparity in MF EMG results, particularly among IFPs. Looking at motorcycle riders, we [5] observed no significant MF EMG decrement throughout a 24-h motorcycle endurance race despite the significant decrease in MVC isom . Following the recommendation of previous studies [33][34][35], we took care not to exceed an interval of 4-5 min between the end of each relay and the handgrip assessment. The lack of MF EMG decrement led to conclude that this interval was too long. According to these findings, we decided to compare an IFP and CFP specifically adapted to motorcycle riders [9]. The lack of a reduction in MF EMG in the IFP suggested that rest cycles were too long, achieving basal values of MF EMG between the work cycles. These findings are in agreement with another study by Krogh-Lund and Jorgensen [23] that compared two pairs of fatiguing sustained isometric contractions at 40% MVC isom separated by different rest intervals. They found that the MF EMG at the start of the second contraction did not recover to pre-fatigued values when the rest interval was less than 1 min, [23]. Other studies reached similar conclusions when they used intermittent contractions [24,25,36], suggesting that a MF EMG shift toward the pre-fatigue state occurs independently of the contraction intensity (25-50%) [36].
Some authors [37] suggest that the validity of the spectral shift of the sEMG signal in assessments of fatigue must be taken with caution because a clear MVC decrement is sometimes weakly reflected in the sEMG signal [38]. This is supported by studies that used IFP to assess muscle fatigue [29,30,39,40]. In contrast, the usefulness of the sEMG signal for studying muscle fatigue in occupational field studies [41] is supported by other studies that reported a reduction in MF EMG with IFP [31,32]. These overall discrepancies between studies suggest that the combination of different, contraction-relaxation periods, effort intensities (%MVC isom ), muscle groups, and other non-controlled or non-reported factors, are critical to understanding muscle fatigue in IFPs [18,22,42].
Therefore, this study aimed to verify in road racing motorcycle riders whether the recovery period performing an IFP matching the braking movement was more relevant than the contraction intensity and effort duration in two forearm muscles (flexor digitorum superficialis and carpi radialis). We hypothesized that MF EMG will not decrease during the contractions performed at 50% MVC isom because they are preceded by long recovery periods. On the contrary, MF EMG recorded at 30% MVC isom and during a shorter exertion time (5 s) may decrease due to short recovery periods (5 s).

Subjects
Twenty-one road racing motorcycle riders aged 29.1 ± 8.0 years (body mass: 72.1 ± 5.5 kg; height: 176.2 ± 4.9 cm) participated in this study. Of these riders, 48% were winners within the Spanish and/or World Championships and 24% were on the podium of the Championship at the end of the season over the previous 6 years. The remaining 28% participated in races at the regional level with at least 5 years of racing experience. The study was approved by the Clinical Research of the Ethics Committee for Clinical Sport Research of Catalonia (Ref. number 15/2018/CEICEGC) and written consent was given by all the participants. The data were analyzed anonymously, and the clinical investigation followed the principles of the Declaration of Helsinki.

Procedures
Before the assessment, the brake lever to handgrip distance was adjusted to the participant's hand size to ensure that hand placement in relation to the brake was similar across all subjects. Afterwards, during the familiarization period, the subject practiced six to ten submaximal non-stationary contractions while watching the dynamometric feedback displayed on the PC screen, while the researcher provided feedback about how to interpret the auditory and visual information. A continuous linear feedback and a columnar and numerical display showed the subject the magnitude of the force they exerted against the brake lever. In addition, a different tone was provided depending on the force level. Dynamometric and sEMG signals were recorded and these signals were synchronized with an external trigger. Five minutes before the beginning of the intermittent fatigue protocol (IFP), two MVC isom trials separated by a 1-min rest were performed to provide a baseline value of MVC isom . The 1-min resting period between the two MVC isom s was considered sufficient to avoid fatigue from the previous contraction [43,44]. The higher MVC isom was recorded as the basal value of that day and used to calculate the submaximal efforts (50% and 30% of the maximum). During the IFP, the subject adopted the "rider position" with both hands on the handlebar.

Sequence and Structure of the IFP
The intermittent protocol comprised a succession of a maximum of 25 rounds. Each round comprised two sections ( Figure 1A). Section one consisted of six 5-s voluntary contractions of 30% MVC isom , with a resting period of 5 s between each contraction. Section two comprised a 3-s MVC isom followed by a 1-min resting period and a 50% MVC isom maintained for 10 s. During the 1-min resting period subjects were in the seated position with their hands resting on their thighs. the national and world level) who agreed about the perception of applying approximately this percentage of force during very strong braking in real situations.
Section two was designed to replicate an experimental protocol from one of our previous studies of motorcycle riders [5]. The test stopped when the subject was unable to maintain the established 50% of MVCisom for 10 s, or the concurrent MVCisom was 10% lower than 50% of the MVCisom value. The number of rounds achieved by each subject was used as a performance measure.

Dynamometric Assessment
To simulate the overall position of a rider on a 600-1000-cc racing motorcycle, a static structure was built to preserve the distances between the seat, stirrups, and particularly the combined system of shanks, forks, handlebar, brake and clutch levers, and gas ( Figure  2). As it happens in a road race motorcycle, levers tilt, distances between levers and handle gas, and distance between the handlebar and seat were modified according to the ergonomic requirements of the rider ( Figure 2).
The subjects were asked to exert a force against the brake lever (always the right hand) using the second and third finger to hold the lever half way, and the thumb and other fingers grasping the handgrip at the same time, which is the most common way of braking of road racing motorcycle riders ( Figure 2). Both arms had a slight elbow flexion (angle 150-160°), forearms half-pronated, wrist in neutral abduction/adduction position and alienated with respect to the forearm, dorsal flexion of the wrist no bigger than 10°, Intensities ranging from 10% to 40% of MVC have previously been used to carry out a continuous or intermittent fatigue protocol [18,22,45]. A sequence of 30% of MVC isom was finally adopted after consulting with expert riders (exclusively, winners of races at the national and world level) who agreed about the perception of applying approximately this percentage of force during very strong braking in real situations.
Section two was designed to replicate an experimental protocol from one of our previous studies of motorcycle riders [5]. The test stopped when the subject was unable to maintain the established 50% of MVC isom for 10 s, or the concurrent MVC isom was 10% lower than 50% of the MVC isom value. The number of rounds achieved by each subject was used as a performance measure.

Dynamometric Assessment
To simulate the overall position of a rider on a 600-1000-cc racing motorcycle, a static structure was built to preserve the distances between the seat, stirrups, and particularly the combined system of shanks, forks, handlebar, brake and clutch levers, and gas ( Figure 2). As it happens in a road race motorcycle, levers tilt, distances between levers and handle gas, and distance between the handlebar and seat were modified according to the ergonomic requirements of the rider (Figure 2).
The subjects were asked to exert a force against the brake lever (always the right hand) using the second and third finger to hold the lever half way, and the thumb and other fingers grasping the handgrip at the same time, which is the most common way of braking of road racing motorcycle riders ( Figure 2). Both arms had a slight elbow flexion (angle 150-160 • ), forearms half-pronated, wrist in neutral abduction/adduction position and alienated with respect to the forearm, dorsal flexion of the wrist no bigger than 10 • , and legs flexed with feet above the footrests; in short, the typical overall position of a rider piloting a motorcycle in a straight line. muscles during handgrip force generation.
To measure the force exerted against the brake lever we used a unidirectional gauge connected to the MuscleLab TM system 4000e (Ergotest Innovation AS, Stathelle, Norway). The frequency of measurement was 400 Hz, and the loading range was from 0 to 4000 N. The gauge (Ergotest Innovation AS, Norway), with a linearity and hysteresis of 0.2%, and 0.1 N sensibility, was attached to the free end of the brake lever in such a way that the brake lever system and the gauge system laid over the same plane and formed a 90° angle approximately when the subject was exerting force. The MVC at the end of the IFP was compared to the MVC in the pre-fatigued state. The 30% and 50% MVC contractions were used for sEMG analysis. Figure 2. Simulation of the overall position of a rider above a motorcycle race from 600 cc to 1000 cc. A static structure was built to preserve the distances between seat, stirrups, and particularly the combined system of shanks, forks, handlebar, brake and clutch levers, and gas.

Electromyography
A ME6000 electromyography system (Mega Electronics, Kuopio, Finland) was used to register flexor digitorum superficialis (FS) and carpi radialis (CR) EMG signals. Adhesive surface electrodes (Ambu Blue Sensor, M-00-S, Denmark) were placed 2 cm apart (from center to center) according to the anatomical recommendations of the SENIAM Project [48,49]. The raw signal was recorded at a sampling frequency of 1000 Hz. Data were amplified with a gain of 1000 using an analog differential amplifier and a common-mode rejection ratio of 110 dB. The input impedance was 10 GΩ. A Butterworth bandpass filter Figure 2. Simulation of the overall position of a rider above a motorcycle race from 600 cc to 1000 cc. A static structure was built to preserve the distances between seat, stirrups, and particularly the combined system of shanks, forks, handlebar, brake and clutch levers, and gas.
Special attention was given to controlling the handgrip position, and the wrist, elbow, and trunk angles to avoid any modification of the initial overall body position during the test. One experimenter supervised the recording of force and sEMG signals, and another continuously checked the maintenance of body position. It has been reported that variations in body posture [46] and wrist angles [47] alter the behavior of the forearm muscles during handgrip force generation.
To measure the force exerted against the brake lever we used a unidirectional gauge connected to the MuscleLab TM system 4000e (Ergotest Innovation AS, Stathelle, Norway). The frequency of measurement was 400 Hz, and the loading range was from 0 to 4000 N. The gauge (Ergotest Innovation AS, Norway), with a linearity and hysteresis of 0.2%, and 0.1 N sensibility, was attached to the free end of the brake lever in such a way that the brake lever system and the gauge system laid over the same plane and formed a 90 • angle approximately when the subject was exerting force. The MVC at the end of the IFP was compared to the MVC in the pre-fatigued state. The 30% and 50% MVC contractions were used for sEMG analysis.

Electromyography
A ME6000 electromyography system (Mega Electronics, Kuopio, Finland) was used to register flexor digitorum superficialis (FS) and carpi radialis (CR) EMG signals. Adhesive surface electrodes (Ambu Blue Sensor, M-00-S, Ballerup, Denmark) were placed 2 cm apart (from center to center) according to the anatomical recommendations of the SENIAM Project [48,49]. The raw signal was recorded at a sampling frequency of 1000 Hz. Data were amplified with a gain of 1000 using an analog differential amplifier and a common-mode rejection ratio of 110 dB. The input impedance was 10 GΩ. A Butterworth bandpass filter of 8-500 Hz (-3 dB points) was used. To compute the median frequency (MF EMG , Hz), Fast Fourier Transform was used with a frame width at 1024, a shift method of 30% of the frame width, and the "flat-topped" windowing function. The power spectrum densities were computed and averaged afterwards to obtain one mean or median for each submaximal contraction of 30% MVC isom (5 s duration) and 50% MVC isom (10 s duration). Afterwards, the median frequency (MF EMG ) was normalized with respect to the basal condition during the MVC isom .
In order to obtain the same number of MF EMG values from the IFP of each individual, and for each round and MVC isom intensity, the six 30% MVC isom s of the first section ( Figure 1A) were averaged to obtain one MF EMG (MF EMG30 ). Each MF EMG30 was paired with the only MF EMG of the second section ( Figure 1A) obtained from the 50% MVC isom (MF EMG50 ).

Statistics
Parametric statistics were used after confirming the normal distribution of the normalized parameters used in this study (MVC isom , MF EMG30 , and MF EMG50 ) with the Shapiro-Wilk test. Descriptive results were reported as the mean and standard deviation. A paired sample t-test was used to compare the MVC isom in the pre-fatigued state and at the end of the IFP. Two methodological approaches were used to verify the study's hypothesis. First, we used regression analysis for each individual, to study the strength of the relation and detect possible trends between the number of rounds accomplished (independent variable) and the MF EMG30 (dependent variable). Second, we used a 2 (time points: T 1 and T 2 ) × 2 (muscles: FS and CR) × 2 (%MVC isom : 30 and 50) ANOVA of repeated measures to compare all MF EMG values at the beginning and the end of the IFP, and to study potential interactions with the two muscle groups analyzed (CR and FS) and the two intensities that were preceded by distinct recovery periods (5 s for 30% MVC isom and 1 min for 50% MVC isom ). When necessary, the Greenhouse-Geisser's correction was used if the sphericity test to study matrix proportionality of the dependent variable was significant (p < 0.05). Then, when a significant effect was found, a post-hoc analysis was carried out conducting multiple comparisons between the normalized rounds with Sidak's adjustment. Partial Eta squared (η 2 p) was used to report effect sizes (0.01 ≈ small, 0.06 ≈ medium, >0.14 ≈ large). Statistical analysis was performed using the PASW Statistics for Windows, Version 18.0 (SPSS, Inc., Chicago, IL, USA). The level of significance was set at 0.05.
Individual regression analysis (Table 1, Figure 3) was conducted to verify possible trends between the NMF of the CR and FS and the number of rounds accomplished by the motorcycle riders during an intermittent fatigue protocol (IFP) at two different intensities (30% and 50% of MVC isom ). The overall individual regression analysis showed a significant linear relationship (p < 0.005) between the MF EMG and the number of rounds accomplished by both muscles when they were exercised at 30% MVC isom (CR 30 and FS 30 ), with pauses of 5 s between each contraction. In contrast, when both muscles were exerted at 50% MVC isom (CR 50 and FS 50 ), after 1 min of recovery, no significant relationship was observed (p > 0.133). The higher correlation observed in CR 30 and FS 30 (r ≥ −0.71) in comparison to CR 50 and FS 50 (r ≤ 0.59) supports the hypothesis of a weaker relationship between the MF EMG50 and the number of rounds when both muscles had the opportunity to recover for longer (1 min for CR 50 and FS 50 ). Similarly, the overall individual regression analysis showed that the fraction of MF EMG variance, explained by the number of rounds attained during the intermittent protocol, was bigger with CR 30 and FS 30 (r 2 ≥ 0.50) in comparison to CR 50 and FS 50 (r 2 ≤ 0.40) ( Table 1).  The second methodological approach was used to determine whether less intense and shorter muscle contractions (30% MVCisom instead of 50%; 5 s instead of 10 s) could induce bigger MFEMG decrements in the CR and FS. The second objective was to determine whether the two muscles (CR and FS) had a similar MFEMG decrement due to fatigue. Thus, we compared two times of measurement (T1 and T2), two muscles (CR and FS) and two contraction intensities (30% and 50% of MVCisom) ( Table 3).
A significant three-way interaction was found (p < 0.001) with a large effect size (η 2 p = 0.5) ( Table 3). Paired comparisons found lower values for the FS than the CR at both times and both intensities. Moreover, we observed a higher MFEMG in the CR muscle at 30% MVCisom (CR30) than at 50% MVCisom (CR50) at the beginning of the IFP, but the opposite response was observed at the end. Finally, regarding the CR, while MFEMG was lower at the end than at the beginning of the IFP at the 30% MVCisom (CR30), the opposite was observed at the 50% MVCisom exertion (CR50) ( Table 3). In addition to the regression analysis performed for each individual, Table 2 reveals that a greater number of riders satisfied better levels of statistical condition in CR 30 and FS 30 in comparison to CR 50 and FS 50 . Moreover, the higher correlation values (r > 0.70) and higher levels of significance (p < 0.001) were associated with higher frequency values in CR 30 and FS 30 , while lower correlation values (r < 0.39) and lower levels of significance (p > 0.05) were associated with a higher number of riders in CR 50 and FS 50 .  Figure 3 is an example of the regression analysis carried out in one subject showing higher MF EMG values for the CR in comparison to the FS. Moreover, at 50% MVC, the MF EMG of the CR never dropped below the MF EMG level established during the basal assessment ( Figure 3B), which is consistent with the comparative results (Table 3).  The second methodological approach was used to determine whether less intense and shorter muscle contractions (30% MVC isom instead of 50%; 5 s instead of 10 s) could induce bigger MF EMG decrements in the CR and FS. The second objective was to determine whether the two muscles (CR and FS) had a similar MF EMG decrement due to fatigue. Thus, we compared two times of measurement (T 1 and T 2 ), two muscles (CR and FS) and two contraction intensities (30% and 50% of MVC isom ) ( Table 3).
A significant three-way interaction was found (p < 0.001) with a large effect size (η 2 p = 0.5) ( Table 3). Paired comparisons found lower values for the FS than the CR at both times and both intensities. Moreover, we observed a higher MF EMG in the CR muscle at 30% MVC isom (CR 30 ) than at 50% MVC isom (CR 50 ) at the beginning of the IFP, but the opposite response was observed at the end. Finally, regarding the CR, while MF EMG was lower at the end than at the beginning of the IFP at the 30% MVC isom (CR 30 ), the opposite was observed at the 50% MVC isom exertion (CR 50 ) ( Table 3).
In addition, a significant two-way interaction was found between the time and MVC isom intensity (time per intensity) with a large effect size (η 2 p = 0.63), but not for the other interactions (time per muscle, and intensity per muscle) with a small and medium effect size, respectively ( Table 3). The MF EMG was higher at the beginning than at the end of the IFP when both muscles were exerted at 30% MVC isom , but no significant differences were observed when they were exerted at 50% MVC isom . Finally, we observed a significant main effect for intensity and muscle factor (Table 3).

Discussion
The MVC isom decrement observed in our IFP confirmed the occurrence of muscle fatigue as this physiological phenomenon is commonly defined as the "loss of the maximal force-generating capacity" [37,50]. From a functional and neurophysiological point of view, and according to the literature, the decrement of the sEMG power spectrum is related, among other factors, to: (1) a reduction in the conduction velocity of the active fibers [35]; (2) impairment of the excitation-contraction coupling [27] related to metabolic changes that occur during fatigue [51]; (3) the recruitment of new units [52], based on the knowledge that subjects with a high relative number of fast twitch fibers may have higher sEMG frequency values [53], and that during fatigue, they show a greater shift towards lower MF EMG compared to subjects with a low relative number of fast twitch fibers [54]; (4) structural damage to muscle cells when muscle soreness is reported by the subjects [18]; (5) other reactions taking place beyond the muscle cell membrane [55], based on observations that short resting periods between each muscle activation are sufficient to maintain the neuromuscular excitability at normal levels during IFP. It must be highlighted that this study did not intend to explain the changes in MF EMG induced by fatigue from a physiological perspective, we were focused on the relationship between the MF EMG and the two factors controlled in our IFP: the load intensity and the work-rest cycle.
High variability of MF EMG values at low loads has been attributed to the influence of the number of recruited muscle fibers and the synchronism and firing rate [56]. According to this, it could be more difficult to find a significant pattern at 30% MVC isom rather than 50% MVC isom , but we found that the MF EMG of the CR and FS decreased more consistently throughout the IFP when the muscles were exerted at 30% MVC isom in comparison to 50% MVC isom . The regression analysis of each individual revealed systematically stronger correlations, coefficients of determination, and statistical significance with CR 30 and FS 30 in comparison with CR 50 and FS 50 . Moreover, participants reported a stronger relationship between the number of rounds accomplished and the MF EMG at 30% MVC isom , rather than 50% MVC isom , in both muscles that were assessed. In agreement with this, we found a higher and more significant MF EMG decrement when the participants performed the IFP at 30% MVC isom , which may suggest different neuromuscular fatigue patterns between the CR 50 and FS 50 during the IFP [9]. If force intensity was the only one factor explaining these differences, it would be difficult to argue that time to exhaustion of any fatigue protocol would be longer when muscles work at higher intensities. As expected, other studies proved the opposite [22,23,57]. Moreover, when studying the magnitude of fatigue in two different IFPs at two different intensities (25 and 50% MVC isom ), Seghers and Spaepen [42] observed very similar relative MF EMG decrements in the two muscles analyzed (IFP at 25% MVC isom : 29%, and 30%; IFP at 50% MVC isom : 29%, and 28%), when sustaining an isometric contraction at 75% of prefatigued MVC isom at the end of both protocols [42]. On the other hand, whereas the same authors observed a significant negative slope of the MF EMG during the IFP at 25% MVC isom , during the IFP at 50% MVC isom the slope did not differ significantly from zero. It is possible that the differences in MF EMG changes during the two IFPs could be more related to differences in their work-rest cycles (10 + 10 s in 25% MVC isom and 5 + 15 s in 50% MVC isom ) than in the contraction intensity. In rock climbers, the significant reduction in the MF EMG observed during an intense IFP (80% MVC isom ) [58], with a work-rest cycle of 5 + 5 s (same cycle as in our IFP for the 30% MVC isom ), indicates that the majority of the frequency components of the MF EMG are unaffected by tension [24]. Thus, we believe that the key point for understanding the different MF EMG patterns during our IFP must be the resting period before the two intensities. Only 5 s of recovery were interspersed between braking muscle contractions of the forearm at 30% MVC isom compared to the 60 s (1 min) at 50% MVC isom. This clearly indicates that MF EMG can be explained to a greater extent when the riders have a very short recovery time despite a smaller contraction intensity (30% MVC isom instead of 50% MVC isom ) and a shorter contraction time (5 s for 30% MVC isom instead of 10 s for 50% MVC isom ). Similar results were reported by Nagata et al. [25].
Nevertheless, it is important to highlight that these authors used a continuous fatigue protocol in which the force was maintained at an intensity of 60% MVC isom until exhaustion, which substantially differs to the IFP in our study.
Before undertaking this study, it was not evident that 1 min of recovery before the 50% MVC isom could be long enough to allow a systematic recovery of the MF EMG towards baseline levels (pre-fatigued). The MF EMG recovery curve towards pre-fatigued values can be characterized by an exponential function [59][60][61], as well as a logarithmic course characterized by large inter-individual variations [61,62]. Therefore, a large proportion of the MF EMG spectrum recovery corresponds to the first 1 min of the exponential recovery curve [21,23,43,44,[59][60][61][62][63]. However, depending on the fatigue protocol, this does not mean full restoration comparable to pre-fatigued or basal MF EMG values. Following the completion of ten cycles of work/rest (10 s/10 s) at MVC isom , Mills [59] observed that the mean power frequency of a compound muscle action potential evoked by supramaximal nerve stimulation required 3 min to recover 50% of its initial values. Three to six minutes, depending on age, are sometimes necessary to recover the pre-fatigued MF EMG values of the abductor digiti-minimi muscle after a MVC isom exertion maintained until 50% MVC isom [64]. Other studies [62,65,66] have confirmed that the majority of the MF EMG spectrum is re-established after 1 and 3 min of recovery, but full recovery it may take until the fifth minute [23,62]. Interestingly, Krogh-Lund and Jorgensen [23] observed that the restoration of MF EMG paralleled that of conduction velocity for the last 4 min of recovery. Regarding the first part of the exponential recovery curve, 35 s were sufficient to allow restoration of 50% of the decline in MF EMG during the previous fatigue protocol [61], but a longer interval (1.4 min) was required to reach 50% of pre-fatigued values for the biceps brachii [67]. Faster MF EMG recovery (up to 85% of the pre-fatigued state during the first minute) was found by Krogh-Lund [21] in the brachioradialis and biceps brachii muscles. Nevertheless, the standard error of the measurement (about 60 s) reported by Elfving et al. [61], which was much larger than the average recovery, reflects the large between-subject variability of the MF EMG parameter when studying the recovery phase. The inconclusive results reported in the literature combined with the accepted large variability that characterizes this type of analysis, support the idea that different combinations of IFP (contraction intensities and durations of contraction and relaxation) to assess muscle fatigue can provide different results [42]. Thus, although it is difficult to compare sEMG data from different studies it is even more complicated when the protocol involves voluntary exercise [37]. The fact that the physiological mechanisms causing muscle fatigue are specific to the task [68], should encourage future studies looking at road racing motorcycle riders to focus on the specific conditions of the forearm muscles, in order to understand better pathologies such as exercise-induced compartment syndrome.
The main limitations of this study were that effort duration, contraction intensity, and recovery time were not separated in different IFPs. Ideally, swapping these three factors would mean that riders had to attend the laboratory on at least six occasions to undertake different IFPs and following a randomized protocol. However, this approach was not feasible in the current study due to the busy racing and training schedules and other commitments of the population of this study.

Conclusions
This study reproduced, in the most accurate way and under laboratory conditions, the braking action in road racing motorcycle riders to investigate different work-rest cycles during an IFP. For training purposes, we recommend using short recovery periods between 5 and 10 s after submaximal muscle contractions as the most effective way to induce muscle fatigue than intermittent tasks performed at higher intensities and with longer recovery periods. That is, much less than 1 min for the resting time (no more than 30 s) according to the results of previous studies [21,23,43,44,60,61,63]. Furthermore, contraction intensities above 50% MVC isom may not be useful for road racing motorcycle riders since only around 30% MVC isom is required to break in real conditions when they have to slow down at high speed (more than 270 km/h) to connect a straight line with a slow curve [9]. Muscle contraction times longer than 10 s are not useful either to match road racing requirements, so protocols involving this type of contraction are not recommended for these individuals. Finally, accelerations with the right hand promote hand dorsal flexion and the assessment of both movements (braking and acceleration) have not been combined in a single IFP. This must be taken into account in future studies to match the real conditions of road motorcycle racing in laboratory settings. This knowledge is needed to enhance our understanding of the most appropriate stimulus (muscle contraction intensities and recovery periods) to be applied within the training programs of road racing motorcycle riders in order to mimic racing conditions and to reduce the risk of muscle pathologies such as the forearm chronic exertional compartmental syndrome.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study is not available.