Journal Logo


Effect of Playing Surface Properties on Neuromuscular Fatigue in Tennis


Author Information
Medicine & Science in Sports & Exercise: November 2012 - Volume 44 - Issue 11 - p 2182-2189
doi: 10.1249/MSS.0b013e3182618cf9
  • Free


Fatigue is usually described as a time-dependent exercise-induced reduction in the maximal force generating capacity of the muscle. This complex phenomenon is the result of the combination of many factors from the central (nervous) to the peripheral (muscular) level (11). It is well known that the extent and origins of neuromuscular fatigue differ according to the type of muscle contraction, the muscular group involved, the exercise duration/intensity, and the environmental conditions (7,30,36). Among the environmental conditions, the effect of the playing surface has received little attention in the literature.

The unique characteristic of the tennis game is to propose confrontations on surfaces with different frictional and cushioning properties, which directly influence the ball rebound and the game speed (8). Two playing surfaces are recognized as radically different: The clay court (CLAY) is characterized by a slow game, a sleepy surface, and high bouncing balls. Conversely, the hard court (HARD) (GreenSet®, Barcelona, Spain) is a fast but rough surface with high frictional properties, producing normal ball rebounds. As a consequence, tennis players adjust their game strategies to the surface they play on, so that both the duration of the rallies and the effective playing time are longer on CLAY than on HARD (32). The aerobic metabolism is therefore more solicited on CLAY than on HARD (26,32). In addition, the frictional and cushioning properties of the surface also influence the intensity of ground impacts and thus the lower limb muscle contractions (18). Although CLAY surface limits ground impacts and diminishes the intensity of the braking (i.e., eccentric) phases because it allows sliding, HARD may expose the player to higher impact forces and intense eccentric contractions during the braking phases (12). Impacts and eccentric contractions have been shown to be responsible for the development of peripheral and central fatigue (30). Specifically, the repetition of ground impacts and eccentric contractions favors the occurrence of muscle damage and peripheral fatigue (6) and the reduction of the descending motor drive (23,28). Finally, the foot plant is loaded specifically as a function of the playing surface (12). Because the calf muscles appear to play an important role for the support of the longitudinal arch of the foot during side-to-side movements, which are common in tennis (25), differential loading, and thus fatigue, could be expected on the calf muscles as a function of the playing surface.

Surprisingly, only two studies have described the neuromuscular fatigue induced by a tennis match, but none compared the neuromuscular fatigue induced by different playing surfaces. Indeed, tennis players compete over different surfaces (mainly CLAY and HARD at the amateur level) during the season, so it is fundamental to know their neuromuscular system fatigue as a function of the surface properties to adapt their training programs and their equipment (i.e., shoes). Girard et al. demonstrated on both the knee extensors (13) and the plantar flexors (16) the occurrence of peripheral and central fatigue after a 3-h tennis match on HARD. Because the intensity of impacts and braking phases are different on CLAY and HARD, we might hypothesize that the playing surface properties could strongly influence the extent and the origin of neuromuscular fatigue during a tennis game, if the effective playing time is matched between surfaces. Specifically, both central and peripheral fatigue could be more pronounced on HARD than on CLAY. Therefore, the purpose of the present set of experiments was to characterize the potential effect of the ground surface properties on the development of neuromuscular fatigue after a prolonged standardized tennis match. To avoid the confounding effect of the playing duration on the development of neuromuscular fatigue, tennis matches on HARD and CLAY were matched for effective playing time.



Ten healthy male subjects (age: 26.2 ± 4.3 yr; height: 181 ± 6 cm; mass: 72.8 ± 8.4 kg) were enrolled in this study. All the participants were experienced tennis players ranked at the regional or national level and were involved in a training program about 10 h·wk−1. The subjects were asked to refrain from strenuous exercise during the week preceding the protocol. The experiment was conducted according to the Declaration of Helsinki. The participants were fully informed of the procedure and the risks involved and gave their written consent. They were also allowed to withdraw from the study at will. Approval for the project was obtained from the human ethics committee of the university.

Experimental Design

The participants were involved in two testing sessions, organized at least 3 d apart. During each session, neuromuscular function was evaluated before (PRE) and immediately after (POST) a tennis match played on either HARD or CLAY. The plantar flexor muscles were evaluated, rather than the knee extensors, (i) to allow the investigation of the reflex responses (see the latter part of this article), which are still challenging to record on the knee extensors (10), (ii) because the plantar flexors seem more fatigable than the knee extensors after prolonged tennis playing (13,16), and (iii) because they are involved in the support of the foot, which is loaded differentially as a function of the playing surface in tennis (12). Neuromuscular evaluation consisted in determining the isometric maximal voluntary contraction (MVC) force of plantar flexors to provide a global index of fatigue. Maximal voluntary activation level (%VA), as well as maximal soleus (SOL) and lateralis gastrocnemius (LG) EMG activities, was evaluated to evidence central fatigue. To gain insight into the implication of spinal and supraspinal mechanisms in the generation of central fatigue, the maximal amplitudes of the H-reflex and V-wave were also measured. Finally, single and multiple electrical stimulations were delivered at rest to determine the extent and origin of peripheral fatigue. At the beginning of each testing session, the subjects followed a standardized warm-up consisting of 5 min of running and 5 min of playing tennis. In addition, subjects were asked to perform approximately four MVCs and were familiarized with several submaximal electrical stimuli.

Exercise Characteristics

Tennis matches were randomly organized on indoor HARD (GreenSet®) and CLAY between the same opponents. The players were matched for playing level. The effective playing time was set to 45 min. This duration was chosen to induce a tennis game with a minimal duration of 3 h (given an effective to total playing time ratio of 20%–25% [26]). This 3-h duration has been previously associated with the development of both central and peripheral fatigue (13,16). To calculate the effective playing time, each rally duration was measured from the start (i.e., ball leaving the hand of the serving player) to the end of the rally and summed until the total duration reached 45 min. The duration of the rest periods between rallies, change ends, and sets were set according to the rules defined by the International Tennis Federation, that is, 20 s between rallies and 90 s after each odd game. To avoid any potential influence of hypoglycemia and/or hyperthermia on the development of central fatigue (35), the players were fed ad libitum with energy bars, gels, and drinks (Active Bar, X-Treme Power Drink, and X-Treme Energy Gel; Inko S.A., France). Finally, each tennis match was preceded by a standard warm-up, as practiced during tennis tournaments.

Experimental Setting

Force measurements.

Isometric contractions of the plantar flexors performed during the experiment included maximal voluntary and electrically evoked contractions. During all the MVCs, the subjects were strongly encouraged. To ensure reproducible trials, subjects were also provided with visual feedback of the plantar flexor muscles force on a monitor that was placed in front of them. During the testing, the subjects were seated on an isometric dynamometer equipped with a pedal and strain gauges (Good Strength; Metitur, Finland). Safety belts were strapped across the chest, thigh, and hips to avoid lateral, vertical, and frontal displacements. Velcro straps were also used to limit heel lift. All measurements were taken from the subject’s dominant leg, with the knee and ankle angles set at 90° from full extension. The ergometer was located near the tennis court to reduce the delay between the end of the match and the start of the measurement (approximately 3 min with this setting).

Electrical stimulation.

After posterior tibial nerve detection with a ball probe cathode pressed into the popliteal fossa, electrical stimulation was applied percutaneously to the motor nerve via a self-adhesive electrode (10-mm diameter, Ag–AgCl, type 0601000402; Contrôle Graphique Medical, Brie-Comte-Robert, France). An elastic band, strapped around the knee joint, applied a constant pressure on the cathode electrode. The anode, a 10 × 5-cm self-adhesive stimulation electrode (Medicompex SA, Ecublens, Switzerland), was located on the patella. A constant current stimulator (Digitimer DS7A, Hertfordshire, United Kingdom) was used to deliver a square-wave stimulus of 1-ms duration with maximal voltage of 400 V. The optimal stimulation intensity was determined from M-wave measurement (see the latter part of this article) at PRE and POST.

Percutaneous muscular stimulations were also given via self-adhesive electrodes (Medicompex SA) to deliver submaximal stimulations at high (80 Hz) and low (20 Hz) frequencies to assess the occurrence of high- or low-frequency fatigue. This method was preferred to nerve stimulation because it is less painful during tetanic stimulation, and its validity for the evaluation of low-frequency fatigue has been established (29). The subjects were instructed to relax while seated and strapped. The anode (10 × 5 cm) was placed on the motor points of the medial and lateral gastrocnemii muscles. The cathode (10 × 5 cm) was placed over the lower part of the calf. The stimulating electrodes were removed between each test sessions, but their exact positions were marked on the skin. Two 0.75-s stimulation trains separated by a 30-s rest interval were applied at 80 and 20 Hz (61 and 16 stimuli, respectively; Fig. 1). The intensity of stimulation was set to match the optimal intensity determined during nerve stimulation.

Overview of the stimulating procedures. Dashed arrows indicate stimulation at submaximal intensity. Full arrows indicate stimulation at supramaximal intensity. The series of MVCs needed for the determination of Hsup and V were performed randomly. The stimulations necessary to evoke the maximal H-reflex (Hmax) were preceded by a recruitment curve (not shown) to determine Hmax stimulation intensity.

EMG recordings.

The EMG signals of the LG and SOL muscles were recorded using bipolar silver chloride surface electrodes (Blue Sensor N-00-S; Ambu, Denmark) during MVCs and electrical stimulations. The recording electrodes were taped lengthwise on the skin over the muscle belly according to surface electronmyography for noninclusive assessment of muscles (SENIAM) recommendations (19), with an interelectrode distance of 25 mm. The position of the electrodes was marked on the skin so that they could be fixed in the same place during the following test session or should electrode replacement be required during the experiment. The reference electrode was attached to the malleolus. Low impedance (Z < 5 kΩ) at the skin–electrode surface was obtained by shaving, abrading the skin with thin sand paper, and cleaning with alcohol. EMG signals were amplified (Octal Bio Amp ML 138; ADInstruments, Australia) with a bandwidth frequency ranging from 10 to 1 kHz (common mode rejection ratio >96 dB, gain = 1000) and simultaneously digitized together with force signals using an acquisition card (PowerLab 16SP, ADInstruments) and the Chart 5.0 software (ADInstruments). The sampling frequency was 2 kHz.

Experimental Variables and Data Analysis

Evoked reflexes and M-waves.

The passive H–M recruitment curve was first carefully examined to determine the stimulus intensity necessary to obtain the maximal SOL and LG H-reflexes (Hmax) and M-wave peak-to-peak amplitude (Mmax). This recruitment curve was determined both before and after the tennis match to account for potential postexercise facilitation of the reflex responses (40). Two stimuli were delivered at each intensity level, interspaced by 5 s between each pulse. Intensity was increased by 2 mA from 1 mA until there was no further increase neither in peak twitch force (i.e., the highest value of the plantar flexor twitch force was reached) nor in concomitant SOL and LG peak-to-peak M-wave amplitudes. Hmax, Mmax, and the corresponding stimulation intensities were then determined from the recordings. To check the consistency of the stimulation across testing sessions, the amplitude of submaximal M-wave accompanying Hmax (MHmax) was also measured and normalized to Mmax (MHmax·Mmax−1).

After determination of the optimal intensity of stimulation (i.e., the intensity corresponding to Mmax) from the H–M recruitment curve, three single pulses at supramaximal intensity (150% of the optimal intensity; range, HARD PRE: 22–66 mA, HARD POST: 22–60 mA, CLAY PRE: 18–62 mA, CLAY POST: 18–60 mA), each separated by 5 s, were then delivered, and the mean value of the three maximal M-waves amplitudes was taken as the Mmax value (Fig. 1). The same procedure was applied for the data processing of the M-wave peak-to-peak duration (MPTP) and M-wave root mean square value (MRMS). To avoid the confounding influence of changes in sarcolemmal excitability on the EMG signal, the Hmax response on LG and SOL was normalized to the M-wave amplitude (Hmax·Mmax−1).

H-reflexes were also evoked during MVCs (Fig. 1). Single pulses were delivered at Hmax intensity (range, HARD PRE: 3–19 mA, HARD POST: 4–23 mA, CLAY PRE: 3–21 mA, CLAY POST: 5–19 mA) during three 4-s MVCs, each separated by a 30-s rest period. The amplitudes of the superimposed H-reflex (Hsup) were calculated. Particular attention was paid to obtain reproducible responses (i.e., coefficient of variation <5%). When these conditions were not filled, one or two additional MVCs were performed. To check the consistency of the stimulation across testing sessions, the amplitude of submaximal M-wave accompanying Hsup (MHsup) was also measured and normalized to the superimposed maximal M-wave (Msup, see the latter part of this article) to calculate the MHsup·Msup−1 ratio.

The same procedure was used to evoke V-waves during MVCs, except that the stimulation intensity was supramaximal (i.e., 150% of optimal intensity; [17]). The amplitudes of the V-wave (V) and Msup were calculated. The Hsup and V responses were normalized to Msup (Hsup·Msup−1 and V·Msup−1, respectively) and averaged over three reproducible trials. The series of MVCs needed for the determination of Hsup and V were performed randomly.

Maximal voluntary contractions and maximal activation level.

Subjects performed two 4-s MVCs of the plantar flexor muscles with supramaximal paired stimuli (120% of optimal intensity) delivered over the isometric plateau (superimposed doublet) and 3 s after the contraction (potentiated doublet) to assess %VA according to the twitch interpolation technique (2). The ratio of the amplitude of the superimposed doublet over the size of the potentiated doublet was then calculated to obtain %VA as follows:

An intensity equal to 120% rather than 150% of optimal intensity was chosen for paired stimulation (i) to reduce discomfort during repetitive testing (27) and (ii) to minimize the crosstalk contamination of the superimposed doublet from the stimulation of the antagonist muscles, which may affect the outcome of the twitch interpolation technique (3,11).

Peak forces reached during the two MVCs were averaged, and the resulting MVC value was retained for further data analysis. The root mean square (RMS) values of the LG and SOL EMG activity were calculated during the MVC trials over a 0.5-s period after the force had reached a plateau and before the superimposed stimulation was evoked. This RMS value was then normalized to RMS value of the M-wave (RMS·MRMS−1). A total of 8 MVCs was performed for the determination of MVC value, %VA, and Hsup·Msup−1 and V·Msup−1 ratios. Both before and after the tennis match, the first MVC was not different from the eighth, suggesting that the testing procedures did not induce additional fatigue.

Contractile properties.

The characteristics of the three twitches evoked by single pulses at rest (Fig. 1) were averaged, and the following parameters were considered during data analysis: amplitude of the twitch peak force (Pf), contraction time (CT), and half relaxation time (HRT). The amplitude of the potentiated doublet (Db) was also measured, because it has been presented as a better indicator of peripheral fatigue (39). Twitch characteristics were nevertheless measured to compare our data with the literature (16). Finally, the ratio of the forces induced by tetani delivered at low (20 Hz) and high (80 Hz) frequencies was calculated to investigate the occurrence of high- or low-frequency fatigue.


Normal distribution was checked using a Shapiro–Wilk test of normality. Each studied variable was then compared between the different instances using a two-way (time × surface) ANOVA with repeated measures. If the ANOVA revealed a significant main effect for any factor or interaction, Newman–Keuls post hoc tests were applied to determine between-means differences. For all statistical analyses, significance was accepted when P < 0.05. All data were expressed as means ± SD within the text and tables and means ± SE in figures for the sake of clarity.


The total playing duration on HARD and CLAY was similar (203 ± 22 vs. 197 ± 23 min, respectively). No significant effect of the playing surface was observed on any experimental variable.

MVCs, EMG activity, and voluntary activation level.

MVC was significantly reduced at POST (HARD: 987.1 ± 242.9 vs. 889.8 ± 210.7 N at PRE and POST, respectively; CLAY: 1001.1 ± 210.5 vs. 960.7 ± 264.9 N at PRE and POST, respectively; P < 0.05; Fig. 2). Playing surface did not affect the MVC reduction.

Average reduction of the MVC after the tennis match on HARD and CLAY playing surfaces. Data are presented as mean ± SE. *Significant reduction from baseline values, P < 0.05.

The variables reflecting the voluntary activation capacities during MVC showed no significant effect of time or surface, but interesting tendencies were observed. RMS·MRMS−1 strongly tended to be reduced after the game on SOL (P = 0.06) and LG (P = 0.08) (Fig. 3). Conversely, %VA was not significantly affected by the exercise (Fig. 4).

Average reduction of the RMS·MRMS −1 ratio measured on the SOL (upper panel) and LG (lower panel) muscles after the tennis match on HARD and CLAY playing surfaces. Data are presented as mean ± SE.
Voluntary activation levels before (PRE) and after (POST) the tennis match played on HARD and CLAY surfaces. Data are presented as mean ± SE.

Reflex responses.

No main effect or significant interaction was found for MHmax·Mmax−1 and MHsup·Msup−1 ratios (Table 1), suggesting that stimulus conditions were stable. The reflex responses evoked during voluntary contractions (Hsup·Msup−1 and V·Msup−1) were not significantly modified after the game (Table 1). Conversely, the Hmax·Mmax−1 ratio was significantly reduced after the exercise on both LG and SOL muscles (P < 0.05, Table 1). Playing surface did not affect the Hmax·Mmax−1 reduction after the game (SOL: HARD, −16% ± 49%, vs. CLAY, 35% ± 36%; LG: HARD, −16% ± 33%, vs. CLAY, 24% ± 54%, respectively).

Characteristics of the reflex responses and accompanying M-waves evoked on the SOL and LG muscles before (PRE) and after (POST) the tennis game on hard surface or clay.

Contractile properties and M-waves.

The ratio of the forces evoked by tetani delivered at low (20 Hz) and high (80 Hz) frequencies was not significantly affected by the tennis game (HARD: 79% ± 7% vs. 76% ± 16% at PRE and POST, respectively; CLAY: 82% ± 9% vs. 79% ± 11%, respectively). Conversely, some characteristics of the evoked twitch were modified after the exercise (Table 2): Pf and CT were reduced at POST (P < 0.05). Db was also significantly reduced after the game (−8.9% ± 11.5% and −6.8% ± 20.7% for HARD and CLAY, respectively; P < 0.05). Although Pf reduction tended (P = 0.07) to be more pronounced on HARD as compared with CLAY (−10.5% ± 11.9% vs. −1.3% ± 13.0%, respectively), this tendency was not confirmed for Db (P = 0.4).

Characteristics of the evoked twitch before (PRE) and after (POST) the tennis game on hard surface or clay: Peak force (Pf), CT, and HRT values are presented as mean ± SD (n = 10).

M-wave characteristics were only minimally affected by the tennis game. Anecdotal reduction of the MPTP was observed on the SOL muscle at POST (HARD: 2.5 ± 0.6 vs. 2.0 ± 0.4 ms at PRE and POST, respectively; CLAY: 2.7 ± 1.2 vs. 2.5 ± 1.1 ms; P < 0.01). The other characteristics of the M-wave were not significantly affected by the exercise.


The aim of this experiment was to characterize the effect of the ground surface properties on the development of neuromuscular fatigue after a prolonged tennis match. The results demonstrated that when tennis matches were played on HARD and CLAY but matched for effective playing time (i.e., 45 min), no significant effect of the playing surface was observed on the respective levels of neuromuscular fatigue. The reduction of MVC occurred concomitantly with significant alterations of the contractile properties of the plantar flexor muscles. The implication of central factors was less clear, as evidenced by the significant reduction of the Hmax·Mmax−1 ratio, the strong tendency of the normalized EMG to be reduced but the nonsignificant reduction of the activation level. In addition, the reflex responses evoked during voluntary contractions were not significantly modified by the exercise.


The force decrement reported in the present study on HARD (−9%) is smaller as compared with the data of Girard et al. (16), who reported a 15% reduction of the plantar flexor muscles MVC after a 3-h tennis match on HARD. However, in the present study, 90-s rest periods were organized every odd game, according to the rules defined by the International Tennis Federation, which was not the case in the protocol of Girard et al. (16). This recovery delay may have limited the development of fatigue during the game. Furthermore, the redetermination of the Hmax stimulation intensity after the game, before starting the MVCs, may have constituted an additional recovery period (approximately 2 min). Finally, our subjects reported a weekly training duration that was twice that of the participants in the study of Girard et al. (16). This difference in training status may have also contributed to reduce the extent of fatigue. Interestingly, it can be suggested that the differences in MVC reduction between the study of Girard et al. (16) and the present experiment are likely not related to the differences in mechanisms beyond the neuromuscular junction because twitch amplitude changes similarly for the two studies (see the latter part of this article). Nevertheless, our results show a moderate MVC reduction, which is consistent with the fact that high-intensity intermittent activities are responsible for moderate fatigue development (13,14,16,37) as compared with endurance sports of similar duration (31). In the present experiment, MVC reduction was not significantly different on HARD and CLAY. Because tennis matches were matched for effective playing time, these results suggest that the playing surface properties do not significantly influence the extent of neuromuscular fatigue. Specifically, the differences in eccentric contractions and impact forces during the game on different surfaces (12) are maybe too small to significantly influence the development of neuromuscular fatigue. Alternatively, one may also suggest that tennis players adapt their playing (32) and activation strategies (33,34) to cope with the ground surface properties to limit neuromuscular fatigue. The latter hypothesis seems more likely but warrants further investigations using for instance EMG and motion analysis procedures.

Peripheral fatigue.

As observed for MVC data, the extent of peripheral fatigue was limited. Pf reduction was moderate (−10.5%) and comparable with the data reported by Girard et al. (16) on the plantar flexors after a 3-h tennis game on HARD surface (−10.6%). Db force decreased to the same extent (−9%) after the game on HARD. Conversely, M-wave characteristics and the low-to-high frequency ratio were not significantly affected by the exercise. These results suggest that peripheral fatigue was limited to the alteration of contractile properties. Such alteration may be ascribed to the development of muscle damage during the eccentric actions (i.e., braking phases, turns, and flexions) or to the cross-bridges sensitivity to Ca2+ (9). The influence of metabolic by-products, such as Pi, on cross-bridge activity cannot be ruled out either (1). However, the lack of modification of the low-to-high frequency ratio is somehow surprising because the performance of high-intensity exercise and/or eccentric actions is associated with the preferential reduction of low-frequency evoked force, that is, low-frequency fatigue (20,24,28,29). This type of alteration has been evidenced after tennis matches on the knee extensor muscles (13). This was not the case in the present experiment: Both low and high frequencies evoked forces declined after the game, but to the similar extent, leading to an unchanged ratio. Other authors reported a similar absence of low-frequency fatigue after strenuous ultraendurance runs involving a high repetition of eccentric actions (27,38), although ultraendurance runs place a higher mechanical stress on the neuromuscular system as compared with racket sports (15).

Central fatigue.

We used a set of classical parameters to quantify the occurrence of central fatigue. Voluntary activation level and voluntary EMG were used to identify any reduction of the central drive, as previously reported on the plantar flexor muscles after a 3-h tennis game (16). Among these variables, no significant reduction was observed but a strong tendency (i.e., close to the significant level) toward the reduction of voluntary EMG, which just failed to reach the significance level. These results are similar to those obtained by Girard et al. (13) on the knee extensor muscles, which reported that voluntary EMG was more sensitive than voluntary activation level for the detection of central fatigue after a tennis game. They partly attributed these results to the intersubject variability during the voluntary activation measurements. A high variability was also observed in the current study, especially after the game on CLAY.

We also analyzed the evoked spinal reflex responses to identify any modulation of the central drive at the spinal level. Responses were evoked both on the relaxed muscle and during voluntary contractions to provide a more functional evaluation of spinal modulation during voluntary effort. Interestingly, the H-reflex was significantly depressed when evoked on the relaxed muscle but was not significantly modified when evoked during voluntary contractions. Together with the absence of significant modification of the voluntary activation level, these results suggest that neural compensation mechanisms were present during voluntary efforts to maintain the motor neuron pool excitability to preserve the force output in the fatigued state (4,21). Nevertheless, our results are contradictory with the findings of Girard et al. (16), who demonstrated reduced H and V responses on the relaxed and contracted muscle after a 3-h tennis game. As mentioned earlier, this discrepancy might result from a lesser metabolic demand in our protocol than that imposed by Girard et al. (16) and/or to the differences in training status. Alternatively, it is also worthy to note that these authors used a predetermined stimulation intensity to evoke the reflex responses. Keeping the stimulation intensity constant during the whole experiment may underestimate the postexercise facilitation of the H-reflex and may therefore overestimate the depressing effects of exercise-induced fatigue (40). However, among these two hypotheses, the former certainly accounts for the major part of the discrepancies between our findings and those of Girard et al. (16).

Methodological considerations.

A few limitations of our study must be noted. First, it is possible that the reduction of MVC force may not accurately reflect the fatigability of the tennis players. They felt tired when they had to repeat several intense muscle actions near the end of the match. However, highly motivated subjects were able to repeat few MVCs at a high force level, even after a 3-h match. Results could have been different if the evaluation of neuromuscular function would have consisted in a repetition of several MVCs (i.e., 10–25), with very short recovery in between (such as the protocol proposed by Kroll et al. [22]). Future studies should consider this possibility. Second, we cannot exclude the possibility that postactivation depression may have affected our H-reflex data in the relaxed condition, because H-reflex data were evoked in the relaxed muscle every 5 s, which might not be sufficient to fully avoid homonymous postactivation depression (5,41). However, it is likely that this phenomenon may have similarly affected H-reflex responses measured before and after the game, so the comparison between these instants should remain valid. Moreover, because the recovery period between MVCs was greater than 10 s, this phenomenon may not have influenced the H-reflex responses evoked during MVCs. Third, as mentioned earlier, some parameters were highly variable among subjects, especially after the game on CLAY. Such variability may be ascribed to the training status of the subjects. Indeed, French players train most of the time on HARDs because tournaments played on CLAY are less frequent. Such variability and the limited sample size certainly affected the outcome of the statistical analysis. Finally, our players were not accustomed to play for a 45-min effective duration, that is, four to five sets. Consequently, we noted that when players felt fatigue increasing, some of them tended to reduce the game intensity to reduce the number of faults, increase the effective playing duration, and complete the match as quickly as possible. Conversely, others kept playing at the same intensity, but fatigue development increased the number of faults. This dramatically decreased the effective playing duration of each rally and thus increased the total playing duration. Although these two opposite strategies may have affected the extent and the variability of fatigue responses among players, this reflects the reality of the tennis game. Consequently, future experiments could compare matches played on the two surfaces, but standardized for perceived exertion: The matches would be stopped when a given level of perceived exertion is reached, so that the comparison of the potential fatiguing effect of the surfaces would be assessed solely by the effective time played.


In our protocol, characterized by a standardized and long effective playing time, as well as rest periods consistent with International Tennis Federation rules, the playing surface properties did not influence the development of neuromuscular fatigue in expert tennis players. Precisely, playing on HARD or CLAY surface lead to a moderate reduction of the maximal voluntary force, which was mainly associated with the alteration of the contractile properties of the plantar flexor muscles. The implication of central factors in the voluntary force decrement was less clear. The ability of the tennis players to readjust their neuromuscular activation strategies as a function of the ground surface properties warrants further investigation. These results could provide additional insights into the understanding of neuromuscular fatigue development in tennis.

No funding was received for this research project.

The authors are grateful to the Ligue de Provence de Tennis for giving access to their tennis courts. The authors also thank Inko for providing energy bars and drinks.

The authors of this article (JBF, VM, JG, FC, and LG) do not have any conflict of interest.

They also state that the results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Allen DG, Westerblad H. Role of phosphate and calcium stores in muscle fatigue. J Physiol. 2001; 536 (Pt 3): 657–65.
2. Allen GM, Gandevia SC, McKenzie DK. Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle Nerve. 1995; 18 (6): 593–600.
3. Awiszus F, Wahl B, Meinecke I. Influence of stimulus cross talk on results of the twitch-interpolation technique at the biceps brachii muscle. Muscle Nerve. 1997; 20 (9): 1187–90.
4. Biro A, Griffin L, Cafarelli E. Reflex gain of muscle spindle pathways during fatigue. Exp Brain Res. 2007; 177 (2): 157–66.
5. Crone C, Nielsen J. Methodological implications of the post activation depression of the soleus H-reflex in man. Exp Brain Res. 1989; 78 (1): 28–32.
6. Dousset E, Avela J, Ishikawa M, et al.. Bimodal recovery pattern in human skeletal muscle induced by exhaustive stretch-shortening cycle exercise. Med Sci Sports Exerc. 2007; 39 (3): 453–60.
7. Enoka RM, Stuart DG. Neurobiology of muscle fatigue. J Appl Physiol. 1992; 72 (5): 1631–48.
8. Fernandez J, Mendez-Villanueva A, Pluim BM. Intensity of tennis match play. Br J Sports Med. 2006; 40 (5): 387–91; discussion 91.
9. Fitts RH. The cross-bridge cycle and skeletal muscle fatigue. J Appl Physiol. 2008; 104 (2): 551–8.
10. Folland JP, Wakamatsu T, Fimland MS. The influence of maximal isometric activity on twitch and H-reflex potentiation, and quadriceps femoris performance. Eur J Appl Physiol. 2008; 104 (4): 739–48.
11. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue. Physiol Rev. 2001; 81 (4): 1725–89.
12. Girard O, Eicher F, Fourchet F, Micallef JP, Millet GP. Effects of the playing surface on plantar pressures and potential injuries in tennis. Br J Sports Med. 2007; 41 (11): 733–8.
13. Girard O, Lattier G, Maffiuletti NA, Micallef JP, Millet GP. Neuromuscular fatigue during a prolonged intermittent exercise: application to tennis. J Electromyogr Kinesiol. 2008; 18 (6): 1038–46.
14. Girard O, Lattier G, Micallef JP, Millet GP. Changes in exercise characteristics, maximal voluntary contraction, and explosive strength during prolonged tennis playing. Br J Sports Med. 2006; 40 (6): 521–6.
15. Girard O, Millet GP. Neuromuscular fatigue in racquet sports. Neurol Clin. 2008; 26 (1): 181–94; x.
16. Girard O, Racinais S, Micallef JP, Millet GP. Spinal modulations accompany peripheral fatigue during prolonged tennis playing. Scand J Med Sci Sports. 2011; 21: 455–64.
17. Gondin J, Duclay J, Martin A. Soleus- and gastrocnemii-evoked V-wave responses increase after neuromuscular electrical stimulation training. J Neurophysiol. 2006; 95 (6): 3328–35.
18. Hardin EC, van den Bogert AJ, Hamill J. Kinematic adaptations during running: effects of footwear, surface, and duration. Med Sci Sports Exerc. 2004; 36 (5): 838–44.
19. Hermens HJ, Freriks B, Merletti R, et al.. SENIAM 8: European Recommendations for Surface Electromyography. Roessingh Research and Development, Enschede, The Netherlands; 1999.
20. Jones DA. High- and low-frequency fatigue revisited. Acta Physiol Scand. 1996; 156 (3): 265–70.
21. Kirsch RF, Rymer WZ. Neural compensation for muscular fatigue: evidence for significant force regulation in man. J Neurophysiol. 1987; 57 (6): 1893–910.
22. Kroll W, Clarkson PM, Kamen G, Lambert J. Muscle fiber type composition and knee extension isometric strength fatigue patterns in power- and endurance-trained males. Res Q Exerc Sport. 1980; 51 (2): 323–33.
23. Kuitunen S, Avela J, Kyrolainen H, Komi PV. Voluntary activation and mechanical performance of human triceps surae muscle after exhaustive stretch-shortening cycle jumping exercise. Eur J Appl Physiol. 2004; 91 (5–6): 538–44.
24. Lattier G, Millet GY, Martin A, Martin V. Fatigue and recovery after high-intensity exercise part I: neuromuscular fatigue. Int J Sports Med. 2004; 25 (6): 450–6.
25. Lynch SA, Renström P. Foot problems in tennis. In: Renström P, editor. Tennis. Oxford (UK): Blackwell; 2002. pp. 155–64.
26. Martin C, Thevenet D, Zouhal H, et al.. Effects of playing surface (hard and clay courts) on heart rate and blood lactate during tennis matches played by high-level players. J Strength Conditioning Res. 2011; 25 (1): 163–70.
27. Martin V, Kerherve H, Messonnier LA, et al.. Central and peripheral contributions to neuromuscular fatigue induced by a 24-hour treadmill run. J Appl Physiol. 2010; 108: 1224–33.
28. Martin V, Millet GY, Lattier G, Perrod L. Why does knee extensor muscles torque decrease after eccentric-type exercise? J Sport Med Phys Fitness. 2005; 45 (2): 143–51.
29. Martin V, Millet GY, Martin A, Deley G, Lattier G. Assessment of low-frequency fatigue with two methods of electrical stimulation. J Appl Physiol. 2004; 97 (5): 1923–9.
30. Millet GY, Lepers R. Alterations of neuromuscular function after prolonged running, cycling and skiing exercises. Sports Med. 2004; 34 (2): 105–16.
31. Millet GY, Martin V, Lattier G, Ballay Y. Mechanisms contributing to knee extensor strength loss after prolonged running exercise. J Appl Physiol. 2003; 94 (1): 193–8.
32. Murias JM, Lanatta D, Arcuri CR, Laino FA. Metabolic and functional responses playing tennis on different surfaces. J Strength Conditioning Res. 2007; 21 (1): 112–7.
33. Nigg BM. Impact forces and muscle tuning: a new paradigm. Exerc Sport Sci Rev. 2001; 29 (1): 37–41.
34. Nigg BM, Segesser B. The influence of playing surfaces on the load on the locomotor system and on football and tennis injuries. Sports Med. 1988; 5 (6): 375–85.
35. Nybo L. CNS fatigue and prolonged exercise: effect of glucose supplementation. Med Sci Sports Exerc. 2003; 35 (4): 589–94.
36. Nybo L. Hyperthermia and fatigue. J Appl Physiol. 2008; 104 (3): 871–8.
37. Perrey S, Racinais S, Saimouaa K, Girard O. Neural and muscular adjustments following repeated running sprints. Eur J Appl Physiol. 2010; 109 (6): 1027–36.
38. Place N, Lepers R, Deley G, Millet GY. Time course of neuromuscular alterations during a prolonged running exercise. Med Sci Sports Exerc. 2004; 36 (8): 1347–56.
39. Place N, Maffiuletti NA, Martin A, Lepers R. Assessment of the reliability of central and peripheral fatigue after sustained maximal voluntary contraction of the quadriceps muscle. Muscle Nerve. 2007; 35 (4): 486–95.
40. Rupp T, Girard O, Perrey S. Redetermination of the optimal stimulation intensity modifies resting H-reflex recovery after a sustained moderate-intensity muscle contraction. Muscle Nerve. 2010; 41 (5): 642–50.
41. Zehr EP. Considerations for use of the Hoffmann reflex in exercise studies. Eur J Appl Physiol. 2002; 86 (6): 455–68.


©2012The American College of Sports Medicine