Introduction
Soccer competition is characterized by intermittent physical activities in which sequences of actions requiring a variety of skills of different intensities are strung together. In professional players, explosive-type efforts such as sprints, jumps, duels, and kicking only represent a small percentage of the total time (∼5% for ref. [9]). The other 95%, corresponding to low-intensive efforts, are composed of walking, slow, and moderate running (35, 40, and 20% of total time, respectively). According to field positions, the total distance crossed during games was approximately 10 km for central defenders and approximately 12 km for midfielders and forwards (33). As a result of the succession of these high- and low-intensive efforts, players exhibit heart rates ranging from 80 to 90% of their maximal values (3,13). Therefore, in addition to the essential technical skills and anaerobic power (required for high-intensity activities), soccer players must have sufficient aerobic capacities to maintain high performance levels throughout the game. However, the repetition of these activities during the two 45-minute periods of soccer games can induce a neuromuscular fatigue that could alter the crucial explosive-type efforts (6).
To quantify fatigue, some authors measured sprint alterations during and at the end of soccer games. They observed small but significant decreases at competition end (13,17). For simplicity and to obtain a larger picture of neuromuscular fatigue processes induced by soccer games, several authors applied effort simulations on treadmills. For instance, Rahnama et al. (23) used four 22-minute periods for soccer match modeling. They noticed electromyographic (EMG) activity reductions of the main lower limb muscles immediately after the simulation. Peak torque for eccentric and concentric modalities also decreased at halftime with a more pronounced effect at game end (22). Additionally, Oliver et al. (20) measured vertical jump height decreases after a single 42-minute period. However, these simulations were of different durations than real soccer games and are generally performed on treadmills with limited running speeds and accelerations (e.g., from 6 to 21 km·h−1 in Rahnama et al. [22]). Therefore, this study aimed to quantify neuromuscular fatigue of lower limb muscles induced by a modeled but almost realistic soccer game simulation (also including jumps and ball shoots). To accurately represent a competitive match, modeling was based on the reproduction of real effort sequences and was achieved on a field during two 45-minute periods.
Methods
Experimental Approach to the Problem
The experiment was based on 2 separate sessions. During the first session, the subjects were familiarized with all test procedures. They also conducted a maximal aerobic velocity (MAV) test to (a) calculate the different speeds applied during the simulation and (b) to determine maximal heart rate. During the second session, the subjects performed a soccer game modeling composed of two 45-minute periods interspersed with a 15-minute rest. It consisted of the reproduction and repetition of two 5-minute realistic sequences extracted from a French first League game. To determine fatigue time course, laboratory (voluntary and electrically evoked torque associated with the corresponding EMG activity) and specific tests (vertical jump and 30-m sprint performance) were conducted before, at halftime, and immediately after the simulated game (Figure 1). Statistical analyses allowed us to evaluate fatigue during and after the soccer match modeling. Time (baseline, halftime, and game end) was used as an independent variable. Values obtained for the different tests were used as dependent variables.
;)
Figure 1: Modeling process of the soccer game. Upper part: soccer game modeling. Arrows correspond to baseline, halftime, and game end tests. White and gray squares correspond to 5-minute sequences extracted from an official French first league game. Lower part: example of effort modeling with the first and fifth minutes from a given sequence. Bars correspond to efforts intensity with different durations.
Subjects
Eight volunteers were recruited from the local sport sciences faculty, and they signed an informed consent. Physical characteristics of the participants were as follows (mean ± SD): age 20.4 ± 1.3 years, body mass 70.4 ± 6.9 kg, and height 174.9 ± 5.2 cm. All were amateur soccer players competing at least at a regional level. Their training experience was >10 years and training volume was approximately 3–4 sessions per week. They were asked to restrict fatiguing efforts at least 3 days before each session. The study was in agreement with the Helsinki statement and was approved by the institute's local ethics committee.
Soccer Game Modeling
Modeling was based on a video analysis of professional soccer players' efforts. We considered a game with 2 opposing teams from the first half of the French first League. The selected game was played in October as a part of the 12th championship diary. We chose a midfielder player because efforts can be considered to be representative of soccer players' general activity. The selected player participated in the whole game and was a member of the victorious team (final score: 1–0).
Two 5-minute sequences were randomly extracted in each halftime and video analyzed using Dartfish software (Fribourg, Switzerland). From video analyses, action order and durations were determined. Actions were partitioned into walking, backward, slow and moderate running, and sprints (see Figure 1 for two 1-minute sequences). For simplicity, the average durations of these actions were calculated during each sequence and reproduced instead of real values. However, it appeared that durations for walking and slow running were too wide (range: from 2 to 50 seconds and from 1 to 22 seconds, respectively). These efforts were therefore dichotomized: the first part with shorter durations than the average value and the second part with longer durations (Table 1).
Table 1: Effort characteristics (velocities and durations) used during the soccer game modeling.*
Each sequence (with the same order than that obtained after video analyses but with averaged durations) was repeated 9 times to obtain the 90-minute game modeling (Figure 1). One vertical jump and one explosive ball shoot were included every 5-minutes, to approach, as close as possible, specific soccer situations.
Velocities were determined according to Bangsbo et al. (4). Walking, backward run, and sprint activities were characterized by 1 velocity. Three velocities were established for slow and moderate runs. They were alternated so as to enhance simulation objectivity. The simulation was consequently composed of 9 different velocities and 7 durations for each sequence (Table 1).
The soccer game modeling was conducted on a grass field using soccer cleats. Distances and intensities were individualized and, for simplicity, they were materialized using colored cones (1 color for 1 effort type). Finally, all of these parameters were checked and instructions (i.e., intensities, durations, etc.) were continuously verbally given to the participants by one of the investigators.
Experimental Procedure
The experiment was based on 2 separate sessions, each performed in midafternoon. The first session served as a familiarization session and also consisted of an MAV test conducted 1 week before the soccer game modeling. The simulated game, performed during the second session, was preceded by a standardized warm-up that mimicked those used by soccer players. Pretests were then conducted and included 30-m sprints, vertical jumps, maximal voluntary and electrically evoked contractions of the right quadriceps and hamstring muscle groups. Maximal voluntary contractions were performed on an isokinetic dynamometer under isometric, concentric, and eccentric conditions. During all contractions, EMG data were simultaneously recorded. After pretests, the soccer game modeling was conducted on a grass field using soccer cleats. Tests were repeated on 2 other occasions, that is, at halftime (45 minutes) and immediately after (90 minutes) the soccer game modeling. Tests were randomly presented in each session. The participants were allowed to drink ad libitum before, at halftime, and after the game modeling. The subjects were strongly encouraged by the investigators to perform all actions maximally during the tests and to respect modeling instructions.
Tests
Maximal Aerobic Velocity Test
The participants performed an MAV test, on a motorized treadmill (Technogym, Cesena, Italy), to determine individuals' velocities used during simulation. The test started at 8 km·h−1 and the velocity increased each minute by 0.5 km·h−1. The heart rate was continuously measured using a portable heart rate monitor (Polar Electro Oy, Kempele, Finland). This test was conducted until exhaustion, and MAV was the last velocity maintained 1 minute long.
Neuromuscular Properties
Torque production capacity and the associated EMG activity of the right leg extensor and flexor muscles were quantified on a Biodex isokinetic dynamometer (Biodex Corporation, Shirley, NY, USA) that was previously validated (29). The participants were seated upright on the dynamometer chair at a 95° hip angle. Velcro straps were applied tightly across the thorax and pelvis, the leg being fixed to the dynamometer lever arm. The axis of rotation of the dynamometer was aligned to the lateral femoral condyle, indicating the anatomical joint axis of the knee. The arms were positioned across the chest with each hand clasping the opposite shoulder. Leg extensions and flexions were conducted within an 80° range of motion (from 90° to 10° knee flexion; 0° corresponding to complete leg extension). For all torque measurements, appropriate corrections were made for the gravitational effect of the leg by recording and subtracting the resistive torque of the leg on the relaxed subjects. Whatever the contraction, the subjects were encouraged by investigators to push as hard as possible throughout the whole range of motion for concentric and eccentric contractions.
Quadriceps maximal voluntary torque was measured under isometric, concentric, and eccentric conditions. Hamstring maximal voluntary torque was recorded in isometric and concentric conditions only. Isometric contractions were maintained approximately 3 seconds at a 70° knee flexion angle. Concentric and eccentric contractions were performed at a 60°·s−1 angular velocity. Two attempts were made for each condition. Maximal values were retained for analyses.
Electromyographic activity was concomitantly measured and recorded by using the Biopac system (MP 150, Biopac Systems, Inc., Santa Barbara, CA, USA) with 4 pairs of silver chloride surface electrodes applied over the 3 superficial knee extensor muscles (vastus lateralis, vastus medialis, and rectus femoris) and 1 knee flexor (long head of biceps femoris). The interelectrode distance was 2 cm (center to center). The reference electrode was fixed to the right patella. Low impedance (<2,000 kΩ) of the skin-electrode interface was obtained by shaving, abrading with sandpaper, and cleansing with alcohol. The EMG signals were amplified with a bandwidth frequency ranging from 10 to 500 Hz (common mode rejection ratio = 110 dB, impedance = 1,000 MΩ, gain = 500) and recorded with a sampling frequency of 2,000 Hz. The EMG activity was quantified by means of the root mean square (RMS) amplitude, calculated over a 200-millisecond period. Quadriceps RMS values were subsequently normalized with respect to M-wave peak-to-peak amplitudes. Because no significant difference was obtained between the 3 superficial knee extensors, a mean quadriceps femoris RMS was subsequently calculated by averaging their normalized RMS values. The RMS values corresponding to maximal voluntary torque obtained during each contraction were retained for analyses.
Intrinsic mechanical properties of the knee extensors were quantified by means of electrical stimulations delivered using a Compex-2 stimulator (MediCompex SA, Ecublens, Switzerland) over the femoral nerve on the relaxed subjects. The cathode (ball probe, ∼10-mm diameter) was pressed onto the femoral triangle and moved to the position giving the greatest visible contraction of the whole quadriceps muscle group. The anode (self-adhesive electrode, 10 × 5 cm) was positioned midway between the superior aspect of the greater trochanter and the inferior border of the iliac crest. Single square-wave stimuli were used (1-millisecond duration). The maximal intensity of stimulation was set by progressively increasing the stimulus intensity until the maximal isometric twitch torque was reached. Peak twitch (peak torque induced by stimulation on relaxed quadriceps) and M-wave (EMG peak-to-peak amplitude) were quantified during the maximal electrical stimulations. M-waves were obtained from vastus lateralis, vastus medialis, and rectus femoris muscles. A mean quadriceps M-wave was calculated by averaging M-wave amplitudes of the 3 superficial knee extensors.
Sprint Characteristics
Maximal speed was measured on a track using infrared photoelectric cells (Test Atletici Computerizzati, TEL.SI. s.r.l., Vignola, Italy) positioned at a 0.95-m height. The players started whenever they wanted from a standing position. The forward foot was behind a start line, and the backward foot was positioned on an electronic mat. The participants were asked to run 30 m as fast as possible. Time started when the foot quit the electronic mat. Two runs were performed, and the fastest was used for analysis. Simultaneously, strides contact time, amplitude, and frequency were measured using an Optojump system (Microgate, Bolzano, Italy). Sprint characteristics (i.e., speed, contact time, amplitude, and frequency) were averaged on 3 distance intervals (from 0 to 10 m, 10 to 20 m, and 20 to 30 m).
Vertical Jumps
Jumping ability was evaluated on a force plate (Globus, Codogne, Italy). Squat jump (SJ) was measured starting from a static semisquatting position (knee angle 90°) and without any preliminary movement. Countermovement jump (CMJ) was performed starting from a standing position, then squatting down to a knee angle of 90 ± 5°, and finally extending the knee in 1 continuous movement. The arms were kept on the hips to minimize the upper body contribution. The position of the upper body was standardized to avoid flexion and extension of the trunk. The participants performed 2 trials for each jump. Only the highest jumps were retained for analyses.
Statistical Analyses
Results are expressed as mean values ± SD. Because homoscedasticity and normal distributions of the data were verified by means of Levene's and Kolmogorov-Smirnov tests, parametric statistics were used. Differences between time (baseline, halftime, and game end) were compared by using a 1-way analysis of variance. Time factor was analyzed as repeated measures. F ratios were considered significant at a p level <0.05. A Newman-Keuls post hoc test was subsequently conducted if significant time effects were found. Statistical power values were calculated for various significant differences and ranged between 0.65 and 0.99 with most power values >0.80. Intraclass correlation coefficients were calculated for each variable and ranged between 0.754 and 0.975.
Results
Performing the soccer game modeling solicited 83.9 ± 6.4% of the maximal heart rate. This value corroborated previous results (3,13).
Torque and Electromyographic Activity of Quadriceps Muscles
Compared with baseline, isometric and concentric maximal voluntary torque significantly decreased (p < 0.05) at halftime (−12.7 ± 7.8 and −8.6 ± 14.0%, respectively) (Figure 2A). Significant reductions (p < 0.01) were obtained at game end for isometric, concentric, and eccentric conditions (−18.5 ± 10.7, −12.2 ± 9.3, and −25.4 ± 17.6%, respectively). No difference was measured between halftime and game end. Knee extensors averaged EMG activity revealed significant decreases at halftime in the isometric conditions (−14.1 ± 4.9%, p < 0.01). At game end, significant reductions (p < 0.05) were found compared with initial values in isometric and eccentric conditions (−15.7 ± 12.1 and −15.3 ± 12.3%, respectively) and compared with halftime for eccentric contractions (−13.8 ± 13.2%) (Figure 2B). No time effect was observed for peak twitch and M-wave amplitudes (Table 2).
Table 2: Mechanical properties, M-wave amplitudes of single twitch, and vertical jump height values.*†
Figure 2: Mean (±SD) quadriceps maximal voluntary torque (A) and its corresponding EMG activity (B), hamstring maximal voluntary torque (C), and biceps femoris EMG activity (D). Black, gray, and white bars correspond to the tests performed at baseline, halftime, and game end, respectively. Significant differences (*p < 0.05, **p < 0.01, and ***p < 0.001).
Torque and Electromyographic activity of Hamstring Muscles
Hamstring maximal voluntary torque significantly declined between baseline and halftime for isometric contractions (−8.2 ± 8.7%, p < 0.05). At game end, significant decreases were obtained (p < 0.05) compared with baseline for isometric and concentric conditions (−8.2 ± 2.3 and −12.3 ± 12.7%, respectively) and compared with halftime in concentric conditions (−8.3 ± 11.2%) (Figure 2C). No significant time effect was found for hamstring EMG activity (Figure 2D).
Sprint and Vertical Jump Tests
Compared with baseline, sprint speed significantly declined at halftime for the 20- to 30-m interval (−3.2 ± 2.8%, p < 0.05) (Figure 3A). At game end, sprint speed was further reduced for all distance intervals (−3.6 ± 3.9, −4.3 ± 4.2, and −4.8 ± 2.2% for intervals 0–10, 10–20, and 20–30 m, respectively). Stride frequency also significantly decreased at halftime and at game end (p < 0.05) for all 3 intervals (Figure 3B). No alteration was recorded for stride amplitude and contact time (Figures 3C, D, respectively).
Figure 3: Mean (±SD) speed (A), stride frequency (B), stride amplitude (C), and contact time (D) during 30-m sprints. Sprint characteristics are calculated over 3 distance intervals (0–10, 10–20, and 20–30 m). Black, gray, and white bars correspond to tests performed at baseline, halftime, and game end, respectively. Significant differences (*p < 0.05, **p < 0.01, and ***p < 0.001).
As compared with initial values, SJ height significantly decreased (p < 0.05) at halftime and at game end (−5.2 ± 4.5 and −8.0 ± 6.7%, respectively). No time effect was observed for CMJ (Table 2).
Discussion
The main aim of this study was to quantify the neuromuscular fatigue induced by a soccer game simulation. During and after the game modeling, we observed a fatigue-induced effect on the knee extensors and knee flexors torque production capacity. These alterations were partly associated with EMG activity impairments. Sprint and vertical jump abilities also decreased at halftime and at game end.
This study revealed knee extensors and knee flexors torque production capacity impairments that were primarily obtained at halftime with only small additional reductions during the second half. At game end, strength was altered for all contractile conditions. Similar results have previously been found after a soccer game simulation (23), in elite female soccer players (1) or after a simulated handball game (31). A recent study (10) contradicted these results. Indeed, throughout an intermittent treadmill protocol replicating soccer activity, torque was only altered during eccentric knee flexions without any modification on knee extensors. Quite similarly, but considering fatigue after a marathon, knee flexors eccentric strength decreases were registered without any changes in concentric conditions (12). The decreases in eccentric hamstring strength with fatigue suggested potentially greater injury risks that should be considered during soccer conditioning (11). These previous studies (10,12) differed from ours by the maintained quadriceps torque production capacity. This discrepancy could partly be explained by the lack of intensive efforts such as jumps or direction changes during the soccer simulation or marathon running ([10,12], respectively). Explosive-type activities, regularly performed in the present modeling, are characterized by frequent and important concentric but also eccentric knee extensions. Eccentric solicitations might induce a peripheral fatigue with muscular ultrastructural damage (2) such as disruptions of muscular cells referred to Z-striate deformations or sarcomere overstretching leading to a reduction of functional actin-myosin links (8,14). The greater torque decrease, observed here on knee extensors compared with knee flexors (e.g., −18 and −8% in isometric condition for knee extensors and knee flexors, respectively), could also be explained by the repetition of explosive-type efforts and soccer-specific actions such as ball kicking or passes. Knee extensors are more predisposed to fatigue because of their likely greater contribution compared with flexors during these specific efforts.
Fatigue observed at game end could be attributed to either or both peripheral and central mechanisms. Surprisingly, despite contraction repetition throughout the modeling (particularly under eccentric conditions), peripheral fatigue was not registered here. Contractile properties, as witnessed by evoked contractions, were not altered throughout the modeling. This result is in contradiction with those of previous works such as that of Lepers et al. (15) that obtained combined impairments of peak twitch torque and M-wave amplitude after cycling exercises. Such peripheral fatigue could be attributed to exercise-induced muscle damage, metabolic disturbances, glycogen depletion, calcic processes (1). Although unchanged contractile properties, these peripheral phenomena could not be excluded from potential fatigue origins observed here. Indeed, other mechanisms could simultaneously positively impact peak twitch amplitude. For example, Millet et al. (16) registered increased peak twitch torque after an ultramarathon. This enhancement was attributed to the concomitant but opposed effects of fatigue with mechanisms such as increases in muscle-tendon stiffness, muscle temperature changes and postactivation potentiation (16), all responsible for peak twitch increases. Previous authors have effectively demonstrated the coexistence of potentiation and fatigue in skeletal muscles (25). In addition, quadricep fatigue would appear to be centrally mediated as revealed by RMS decreases. This result is in agreement with those of previous studies considering fatigue after soccer or handball games modeling ([22,31], respectively). This central fatigue could be attributed to alterations of the motor cortex excitability and excitation, descending nervous command toward motoneurons, and motoneuronal excitability (28). Hamstring fatigue would appear different from that of quadriceps muscles. Because no EMG activity alteration was obtained during knee flexion, it could be speculated that fatigue was predominantly peripheral in origin. This peripheral fatigue could likely be attributed to larger muscle damage. As compared with quadriceps, hamstring muscles sustain greater strains, for example, during runs and ball shoots to slow down the leg (i.e., eccentric solicitations).
Knee extensor and knee flexor torque decreases were also associated with alterations in specific performance such as sprints and jumps. Some studies also obtained reduced jumping ability with fatigue (1,20). Others did not register any alterations (13,30). At halftime and game end of this study, SJ height was reduced, whereas CMJ remained unchanged. The dissimilarity between performance reduction while performing SJ and CMJ has previously been registered. For example, Byrne and Easton (7) reported greater decrements during SJ than during CMJ after 100 barbell squats. Precisely, SJ and CMJ do not evaluate the same performance indexes. It would seem that CMJs quantify the force production capacity of lower limbs during a longer active state than the SJs (5). The decline of SJ values, associated with knee extensors torque decreases, translated an explosive strength reduction of the lower limb. Indeed, soccer games induced changes in rapid muscle force production as attested by the reduced quadriceps rate of force development (30). Conversely, the unaffected CMJ performances showed that jumping ability with longer active state was not distorted by soccer game exercises.
Sprint ability was also altered at game end. This finding corroborated those of Andersson et al. (1) who found a 3% time reduction during 20-m sprints after a 90-minute soccer game. Quite similarly, but throughout a marathon, 20-m sprint times regularly decreased from the 20th km (18). Knee extensor and flexor strength decrements seemed to be responsible for this running speed alteration. Indeed, elite soccer players demonstrated a strong correlation between quadriceps muscle strength and sprinting speed (32). Sprint performance reduction, observed here, was primarily attributed to stride frequency reductions rather than to stride amplitude or foot contact time. Although Small et al. (26) registered smaller stride lengths after 90-minute intermittent soccer exercises, Nummela et al. (19) confirmed our results and obtained sprint performance decrease associated with a significant stride frequency reduction after a 5-km running time trial. Such frequency reduction could be attributed to decreased preactivation of extensor muscles just before the foot contact on the ground (21).
Practical Applications
Soccer game induced a fatigue on both knee extensor and knee flexor muscles. This decrease in torque production capacity may likely impair joint stability and therefore increase the potential muscle strain injury. The reduction of strength production capacity also appeared to impair functional and specific efforts. It negatively impacts explosive situations such as vertical jumps and sprints or the ability to perform soccer-specific skills (24,27) and could be of interest during physical and soccer-specific training. For example, the stride frequency decrease simultaneous to sprint slowing can be helpful for speed training orientation: focus on frequency development with ladder or agility drills.
References
1. Andersson, H, Raastad, T, Nilsson, J, Paulsen, G, Garthe, I, and Kadi, F. Neuromuscular fatigue and recovery in elite female soccer: Effects of active recovery.
Med Sci Sports Exerc 40: 372–380, 2008.
2. Armstrong, RB, Ogilvie, RW, and Schwane, JA. Eccentric exercise-induced injury to skeletal muscle.
J Appl Physiol 54: 80–93, 1983.
3. Bangsbo, J. The physiology of soccer in special reference to intense intermittent exercise.
Acta Physiol Scand 15: 1–155, 1994.
4. Bangsbo, J, Norregaard, L, and Thorso, F. Activity profile of competition soccer.
Can J Sports Sci 16: 110–116, 1991.
5. Bobbert, MF and Casius, LJR. Is the effect of a countermovement on jump height due to active state development?
Med Sci Sports Exerc 37: 440–446, 2005.
6. Bradley, PS, Mascio, MD, Peart, D, Olsen, P, and Sheldon, B. High-intensity activity profiles of elite soccer players at different performance levels.
J Strength Cond Res 24: 2343–2451, 2010.
7. Byrne, C and Eston, R. The effect of exercise-induced muscle damage on isometric and dynamic knee extensor strength and
vertical jump performance.
J Sports Sci 20: 417–425, 2002.
8. Clarkson, PM. Eccentric exercise and muscle damage.
Int J Sports Med 18: S314–S317, 1997.
9. Cometti, G. La préparation physique en football. Magny-Les-Hameaux, France: Chiron, 2002.
10. Greig, M. The influence of soccer-specific fatigue on peak isokinetic torque production of the knee flexors and extensors.
Am J Sports Med 36: 1403–1409, 2008.
11. Greig, M and Siegler, JC. Soccer-specific fatigue and eccentric hamstrings muscle strength.
J Athl Train 44: 180–184, 2009.
12. Koller, A, Sumann, G, Schobersberger, W, Hoertnagl, H, and Haid, C. Decrease in eccentric hamstring strength in runners in the Tirol Speed Marathon.
Br J Sports Med 40: 850–852, 2006.
13. Krustrup, P, Zebis, M, Jensen, JM, and Mohr, M. Game-induced fatigue patterns in elite female soccer.
J Strength Cond Res 24: 437–441, 2010.
14. Kuipers, H. Exercise-induced muscle damage.
Int J Sports Med 15: 132–135, 1994.
15. Lepers, R, Theurel, J, Hausswirth, C, and Bernard, T. Neuromuscular fatigue following constant versus variable-intensity endurance cycling in triathletes.
J Sci Med Sport 11: 381–389, 2008.
16. Millet, GY, Lepers, R, Maffiuletti, NA, Babault, N, Martin, V, and Lattier, G. Alterations of neuromuscular performance with ultra-long term fatigue. J Appl Physiol 92: 486–492, 2002.
17. Mohr, M, Krustrup, P, Nybo, L, Nielsen, JJ, and Bangsbo, J. Muscle temperature and
sprint performance during soccer matches-beneficial effect of re-warm-up at half time.
Scand J Med Sci Sports 14: 156–162, 2004.
18. Nicol, C, Komi, PV, and Marconnet, P. Fatigue effects of marathon running on neuromuscular performance. I. Changes in muscle force and stiffness characteristics.
Scand J Med Sci Sports 1: 10–17, 1991.
19. Nummela, AT, Heath, KA, Paavolainen, LM, Lambert, MI, St Clair Gibson, A, Rusko, HK, and Noakes, TD. Fatigue during a 5-km time trial.
Int J Sports Med 29: 738–745, 2008.
20. Oliver, J, Armstrong, N, and Williams, C. Changes in jump performance and muscle activity following soccer-specific-exercise.
J Sports Sci 26: 141–148, 2008.
21. Paavolainen, L, Nummela, A, Rusko, K, and Häkkinen, K. Neuromuscular characteristics and fatigue during 10 km running.
Int J Sports Med 20: 1–6, 1999.
22. Rahnama, N, Lees, A, and Reilly, T.
Electromyography of selected lower-limb muscles fatigued by exercise at the intensity of soccer match-play.
J Electromyogr Kinesiol 16: 257–263, 2006.
23. Rahnama, N, Reilly, T, Lees, A, and Graham-Smith, P. Muscle fatigue induced by exercise simulating the work rate of competitive soccer.
J Sports Sci 21: 933–942, 2003.
24. Rampinini, E, Impellizzeri, FM, Castagna, C, Azzalina, A, Ferrari Bravo, D, and Wisloff, U. Effect of match-related fatigue on short-passing ability in young soccer players.
Med Sci Sports Exerc 40: 934–942, 2008.
25. Rassier, DE and MacIntosh, BR. Coexistence of potentiation and fatigue in skeletal muscle.
Braz J Med Biol Res 33: 499–508, 2000.
26. Small, K, McNaughton, LR, Greig, M, Lohkamp, M, and Lovell, R. Soccer fatigue, sprinting and hamstring injury risk.
Int J Sports Med 30: 573–578, 2009.
27. Stone, KJ and Oliver, JL. The effect of 45 minutes of soccer-specific exercise on the performance of soccer skills.
Int J Sports Physiol Perform 4: 163–175, 2009.
28. Taylor, JL, Todd, G, and Gandevia, SC. Evidence for a supraspinal contribution to human muscle fatigue.
Clin Exp Pharmacol Physiol 33: 400–405, 2006.
29. Taylor, NA, Sanders, RH, Howick, EI, and Stanley, SN. Static and dynamic assessment of the Biodex dynamometer.
Eur J Appl Physiol 62: 180–188, 1991.
30. Thorlund, JB, Aagaard, P, and Madsen, K. Rapid muscle force capacity changes after soccer match play.
Int J Sports Med 30: 273–278, 2009.
31. Thorlund, JB, Michalsik, LB, Madsen, K, and Aagaard, P. Acute fatigue-induced changes in muscle mechanical properties and neuromuscular activity in elite handball players following a handball match.
Scand J Med Sci Sports 18: 462–472, 2008.
32. Wisloff, U, Castagna, C, Helgerud, J, Jones, R, and Hoff, J. Strong correlation of maximal squat strength with
sprint performance and
vertical jump height in elite soccer players.
Br J Sports Med 38: 285–288, 2004.
33. Withers, RT, Maricic, Z, Wasilewski, S, and Kelly, L. Match analysis of Australian professional soccer players.
J Hum Mov Stud 8: 159–176, 1982.
