The demands of a tennis tournament are high both mentally and physiologically. Many studies have examined the effects of one tennis match (Bergeron et al., 1991; Fernandez-Fernandez, 2007; Filaire et al., 2009; Girard et al., 2006; 2008; Hornery et al., 2007). However, tennis is a tournament sport, and it is important to study what changes take place in player’s performance and physiological variables in a tournament environment when several games are played.

Tennis is an anaerobic sport with aerobic breaks between the rallies (Fernandez et al., 2006). A tennis match lasts normally 1.5-2 hours but it can last even over 4 hours. Typically one point lasts 5-10 seconds and between the points there is a 20s break and during the changeovers the break varies from 60 to 120s. During the match players run an average of 3 meters per shot and during the best of three set match players perform 300-500 high intensity efforts. The exercise intensity during the game is about 50-60% of maximal oxygen uptake or 60-80% of maximal heart rate (Fernandez et al., 2006.) Average blood lactate concentration level ranges from 1.7 to 3.8 mmol·l^-1 (Mendez-Villanueva, 2007; Murias et al., 2007), but during long and intense rallies lactate levels can go up to 8.6 mmol·l^-1. The rate of perceived exertion (RPE, range 6-20) is around 10-17 during a tennis match (Mendez-Villanueva et al., 2007)

Tennis movement is a combination of eccentric and concentric muscle action, called as stretch-shortening cycle. All the movements are rapid and explosive including acceleration, deceleration, stretches, jumps and stroke making from various body positions. It has been previously shown that stroke velocity, running speed and maximum force (MVC) decreases during a tennis match (Ferrauti et al., 2001b; Girard et al., 2006; 2008; 2011). Muscle damage has been suggested one factor behind fatigue (Girard et al., 2006; Mendez-Villanueva et al., 2007). Eccentric muscle action is related to exercise induced muscle damage (EIMD) (Piitulainen et al., 2011). EIMD causes muscle soreness, muscle swelling and inflammation, increased plasma levels of muscle protein and serum creatinekinase (CK) and impairment of muscle function reducing MVC and RFD (Byrne et al., 2004; Morgan and Allen, 1999). Muscle soreness (Girard et al., 2006) and serum CK activity (Hornery et al., 2007), which is a marker for muscle damage, has been reported to elevate during a tennis match. Previously, there is no published data of muscle soreness and CK from a tennis tournament environment.

A tennis match like any other high intensity sport competition causes acute hormonal responses (Filaire, 2009; Salvador et al., 2003). Serum hormones of cortisol and testosterone concentration during a tennis match have been studied before (Bergeron et al., 1991; Booth et al., 1989; Ferrauti et al., 2001a; Filaire et al., 2009). Serum cortisol elevates at rest before the match in tennis (Booth et al., 1989; Filaire et al., 2009) and in other sports (Filaire et al., 2001; Kivlighan et al., 2005; Salvador et al., 2003; Suay et al., 1999). During the tennis match serum cortisol concentration elevates (Booth et al., 1989; Filaire et al., 2009), if the intensity is above 60-70% of maximum VO₂ (Kraemer et al., 1989; Wittert et al., 1991). Serum testosterone concentration elevates significantly before the exercise (Bergeron et al., 1991; Booth et al., 1989; Salvador et al., 2003; Suay et al., 1999) or remains at a control level (Filaire et al., 2001; Gonzalez-Bono et al., 1999. Testosterone concentration elevates during the tennis match (Bergeron et al., 1991; Booth et al., 1989) and in other sports (Kivlighann et al., 2005; Suay et al., 1999) or remains at a control level (Filaire et al., 2001; Gonzalez-Bono et al., 1999; Salvador et al., 2003). Serum testosterone concentration will not elevate if the player does not feel the match is important, if he knows he is going to win or lose, if the opponent is at a different level (Booth et al., 1989: Gonzales-Bono et al., 1999).

The relationship between the outcome of the match and serum hormone levels have been studied, but the results are somewhat conflicting. Winner’s serum cortisol concentration has been significantly lower compared to the losers in tennis (Filaire et al., 2009), but in wrestling (Elias, 1981) and in judo (Suay et al., 1999) winners have showed higher cortisol concentration. Serum testosterone concentration does not seem to relate to outcome of the match (Booth et al., 1989; Filaire et al., 2001; Suay et al., 1999).

The purpose of this study was to examine how player’s physical performance is compromised during a 3-day tournament and what kind of effects the matches have on player’s hormonal responses. In addition this study attempted to examine the level of muscle damage that takes place during a match and during the whole tournament. Based on previous studies, we hypothesized that during a 3-day tennis tournament player´s maximal leg strength and rate of force development would be impaired (Girard et al., 2006; 2008; 2011), muscle soreness increased throughout the tournament (Girard et al., 2006; Hornery et al., 2007) and that winners might have lower serum cortisol concentrations than losers (Filaire et al., 2009).

Eight male tennis players of age 23.0 ± 3.8 years, height 1. 85 ± 0.07m, weight 77.4 ± 8.7kg volunteered to participate in this study. They were informed in detail about the nature of the experiment and its possible risks involved. All subjects provided written informed consent and they had a change to withdraw from the study at any stage of the study. All players had over 10 years of playing background, and they were all national level top tennis players. Six out of eight players had or had had professional ranking during their career. During a normal training week players practiced 8.5 ± 2.5 hours for tennis and 4.0 ± 1.5 times for physical fitness. The study was performed after 6-week training period and it was the first tournament of the season for all players. All players were informed to lighten their training 3 days before the tournament.

All measurements of the present study were performed during five days (Figure 1). The first three days players played an invitational singles tennis tournament and the last two days was the recovery phase. During the tournament the measurements were performed before, during (after 40 and 80 minutes) and after the match (T0, T40, T80, T120). The measurements were performed every day at the same time of the day and in the same order. The players were divided into two four-player groups using their national ranking. One group played every day in the morning and the other in the afternoon. The players played one match per day. After the tournament players had recovery tests at 24 and 48 hours after the last match. Matches were played at the indoor hard court (Greenset^® comfort) using the normal ITF (International Tennis Federation) rules except the playing time was standardized to 120 minutes. The matches were played using 4 new Wilson Tour balls. The air temperature was 22 °C. Before, during and after the tournament players were informed to use their normal food and fluid intake. In addition, they were instructed to use their normal social and sleep rhythm that they use during a normal tennis tournament.

Maximum voluntary force (MVC) of the lower extremities was measured using isometric bilateral leg press (Häkkinen et al. 1998). Isometric leg press was performed using 107° knee angle. In addition to maximal peak force (N) the maximal rate of force development was analysed (N·s^-1). Players had 3 trials before and after the match, and during the match only two. The best and technically acceptable trial was taken into the final analysis.

Speed strength was measured using a standing 5- jump, countermovement jump (CMJ), 5m run (5m) and serve velocity test. The standing 5-jump was performed at the tennis court. The maximal standing 5-jump was used to measure explosiveness of leg extensor muscles in to the horizontal direction (Mero et al., 1981). Maximal vertical jumping ability was measured using a counter movement jump (CMJ) on a contact mat with a clock (Komi & Bosco 1978). Acceleration running speed was measured with a standing start over 5 m. The subject was standing 0.5 m from the first photocell gate and then accelerated maximally over 5 m to the second photocell gate (accuracy of 0.01s). In all tests the best performance of three trials (recovery from 2 to 5 minutes between the trials) was selected for the final analysis. The service velocity test was performed to measure explosiveness of the whole body in the actual tennis performance. In the service test each player served 10 serves to deuce court; only the serves that went in to the service box were accepted. Serve velocity was measured using radar gun (Stalker Pro, Stalker Radar, USA).

Blood samples were taken from an antecubital vein in a sitting position. During the match day samples were taken before (T0), during (T40 and T80) and after the match (T120). Recovery samples on day 4 and 5 were taken at the same time of day as before the match sample. The control sample was taken within two weeks from the tournament on a normal rest day.

Four milliliters blood from a vein was senrifugated (10min, 3500rpm). Serum samples were frozen until the analyzing process. The following analyses were performed: hormonal and creatinekinase concentrations were analysed with a kemiluminometric measurement using Immunilite 1000 machine (Diagnostic Products Co, LA, USA). Cortisol intra-assay coeffecient of variation was 4.6%, testosterone 5.7% and creatinekinase 0.8%. The sensivity for cortisol was 5.5 nmol·l^-1, testosterone 0.5 nmol·l^-1 and creatinekinase 7 U·l^-1. Blood lactate and glucose was measured before (T0), during (T40 and T80) and after the match (T120) from fingertip blood sample (EKF Biosen C_Line Sport, Magdeburg, Germany). The measurement range for B-lact was 0.5-40 mmol·l^-1 and the intra-assay CV was 1.5% and B-gluc range 0.5-50 mmol·l^-1 and CV 1.5%.

Before every test day players filled the DOMS questionnaire in which they gave their subjective feeling (1-10, 10 is a lot of pain) how much DOMS they felt in the upper and lower body and core. They also marked the feeling they had for playing tennis (1-10, 10 is the best feeling).

During every changeover players marked their RPE feeling using the Borg scale 6-20. Heart rate was measured during the matches (Suunto smartbelt, Finland).

The matches were videotaped for the match. From the video the following variables were measured: break between points, point length, stroke count/point, work/rest -ratio, winners, unforced errors, aces, double faults and first serve percentage.

Conventional statistical methods were used to obtain means, standard deviations and correlation coefficients. Holm-Bonferron post hoc test was used to assess statistical differences between the matches and different time points during the tournament. Winners and losers were compared using the relative values. Correlations between the matches and recovery days were analyzed using Pearson correlation matrix. The morning and afternoon groups’ changes were compared using covariance analysis. P ≤ 0.05 was regarded as statistically significant. The statistical analyses were carried out using the software program Microsoft Excel 2008 for mac 12.2.1 -version and spss 18.0 -program.

MVC of leg extensors was significantly lower after the third match compared to the start of the tournament (p < 0.05) (Figure 2). MVC was significantly lower before the second and the third match compared to MVC before first match. During the first match MVC reduced significantly (p < 0.05). RFD reduced significantly after the second and the third match compared to RFD before the tournament (Figure 2). RFD was also reduced significantly before the second and the third match (p < 0.05) and remained reduced on the second day (p < 0.05) compared to RFD before the first match. RFD reduced significantly during the first match (p < 0.001). No significant differences were observed between the winners and losers in MVC or RFD.

There were no significant changes between the matches in CMJ, 5m and 5-jump (Table 1). After the second match 5m was significantly lower than before the match. 5-jump reduced significantly during the first match. Serve velocity was significantly slower before the third match compared to the first match. During the first match serve velocity reduced significantly (T0 vs T40 and T80).

No significant differences between the winners and losers were observed in CMJ, 5m, 5-jump or serve velocity. The losers reduced significantly their 5m during the match but no changes occurred among the winners.

Serum cortisol concentration was significantly higher before every match (p < 0.05) and T40 (p < 0.05) compared to the control (Figure 3). Cortisol concentration of the winners raised significantly compared to the control value in all time points T0 (p < 0.01), T40 (p < 0.001), T80 (p < 0.05), T120 (p < 0.05). The losers had the same response only at T0 (p < 0.05) and T40 (p < 0.05). The covariance between the morning and afternoon group was significant (p < 0.05).

Serum testosterone concentration increased significantly during every match at T40 (p = 0.02, p = 0.02, p = 0.015) compared to the control value and to the first match at T80 (p = 0.027) (Figure 4). There were no significant differences between the winners and losers between or during the matches. The covariance between the morning and afternoon group was significant (p = 0.026).

CK increased significantly throughout the whole tournament (Figure 5). Between the matches CK was significantly different between the first and second and the first and third match (p < 0.005). Between the second and third match the difference was significant at T80 and T120 (p = 0.003). There were no differences between the winners and losers.

DOMS of the legs and upper body increased significantly during the second and third match compared to the start of the tournament as well as for the legs between the second and third match (p < 0.02) (Figure 6). Legs DOMS remained significantly higher during both rest days compared to the start of the tournament (p = 0.002; p = 0.038). Upper body DOMS remained significantly higher during a first rest day compared to the start of the tournament (p = 0.012). There were no differences between the winners and losers in DOMS.

Mean blood lactate during the match was 3.4; 4.1; 3.9 nmol·l^-1 (Table 2). During the matches mean heart rate was 149 ± 12; 142 ± 16; 150 ± 11. Mean duration of the points was 6.0 ± 1.0; 5.8 ± 0.8; 5.6 ± 0.4, mean work/rest -ratio 0.25 ± 0.05; 0.24 ± 0.05; 0.23 ± 0.03 and mean RPE was 13.2; 12.8; and 14.4, respectively.

The present tennis tournament led to significant changes in player’s physical performance, considerable hormonal responses and increases in variables indicating muscle soreness. Maximal and explosive isometric force of lower extremities reduced significantly during the tournament. Serum cortisol increased significantly already before the match and testosterone during the first 40-80 minutes of the match and players CK and DOMS increased significantly throughout the whole tournament. One day of rest day after the tournament was not sufficient to reduce CK concentration and DOMS and to recover force production the leg extensor muscles to the level observed before the tournament.

The present tournament led to the reductions in players MVC by 26% and RFD by 38% from the start of the tournament. During the matches MVC decreased constantly in all matches, but the decrease was significant only in the first match at T80 and T120 compared to T0. Girard et al. (2006; 2008) reported that the decrease in the lower extremities MVC became significant 150 minutes after the beginning of the tennis match in a prolonged 3-hour tennis match. After the first match MVC in the present study did not recover between the matches remaining significantly lower before the 2^nd (8%) and before the 3^rd (16%) match than that recorded before the tournament.

Neuromuscular impairment during a prolonged tennis match may be due to the failure of central activation and alterations in excitation- contraction coupling process (Girard et al., 2008), which is linked to EIMD (Warren et al., 1999). An increase in serum CK has been previously reported during a tennis match (Hornery et al., 2007) and accompanied by a progressive reduction in maximal force and an increase in muscle soreness (Girard et al., 2006; 2008). Type II muscle fibers are fast and strong muscle fibers that are recruited during explosive and forceful actions (Karatzaferi et al., 2001). It has been suggested that the damaged type II muscle fibers might impair the tennis performance by reducing player’s abilities to move between the points and perform the strokes properly (Mendez-Villanueva, 2007). Singh et al. (2011) reported that elevated muscle soreness was related to impaired 10m running speed and vertical jump 48-hours after the rugby performance. In the present study player’s physical performance 48-hours after the tournament was still reduced from the start of the tournament. It seems that muscle damage is one of the important factors behind the impairment of the physical performance during a tennis match and during a 3-day tennis tournament.

The impairment of legs RFD has been previously reported in a sport competition (Thornlund et al., 2008; 2009). Interestingly, in this present study the countermovement jump (CMJ) did not reduce during the matches although MVC and RFD reduced. Actually during the 2^nd match CMJ increased during a match. Girard et al., 2006 reported similar kinds of results where CMJ increased since the beginning of the match and MVC reduced. This kind of improving has been explained by insufficient warm-up before the match and by an increase in muscle temperature through various mechanisms such as increased transmission rate of nerve impulses, decreased viscous resistance, and/or increased glycogenolysis, glycolysis, and high energy phosphate degradation (Bishop et al., 2003; Girard et al., 2006). Thornlund et al., 2009 observed similar results after the football match where leg extensors RFD reduced but CMJ remained at the same level before and after match. It could be that the muscles elastic component of muscles can function throughout the tennis tournament and maintain the performance in CMJ, 5m run and in 5-jump. RFD was measured in an isometric leg press with no stretch-shortening cycle action. Previously, it has been shown that muscle fascia and tendons may allow tendons to absorb energy rapidly even if the contractile elements´ maximum capacity for energy absorption is limited (Roberts and Azizi, 2010). It has been suggested that the use of elastic elements may be effective although the force generating elements are damaged (Byrne and Eston, 2002).

The serve speed decreased before the 2^nd and 3^rd match compared to the beginning of the tournament. Reid et al. (2008) showed that service velocity correlated with greater muscle force during the loading stage of serve. After the last match players were able to increase service speed, although the increase was not significant. That result was probably followed by the information that the test was last of the tournament and they did not have any match left so the players might felt they can give everything they had left. It must be taken into consideration that during this tournament the physical load during a match was little bit heavier than during a normal match because of the multiple measurements.

The present match analysis indicated that high in-tensity on-court movement has led to the reduced physical performance, hormonal responses and muscle soreness. The average match lactate concentration (3.4-4.1 mmol·l^-1) was a little above reported in previous studies (Fernandez-Fernandez et al., 2007; Mendez-Villanueva et al., 2007). The work/rest -ratio and the significant increase in RPE throughout every match were in line with previous studies (Fernandez-Fernandez et al., 2007; Girard et al., 2006; Kovacs, 2006; Mendez-Villanueva et al., 2007). The duration of the point and the work/rest -ratio increased during a match, because the balls became softer and maybe because players tried to avoid unforced errors hitting easier shots when they got tired (Ferrauti et al., 2001b).

Serum cortisol concentrations elevated significantly at rest already before and serum testosterone during the match. Increased cortisol (Booth et al., 1989; Filaire et al., 2009) and testosterone (Bergeron et al., 1991; Booth et al., 1989) concentration before, and during and after the match has been reported in earlier tennis studies. The elevated cortisol concentration before the match is related to preparation for the upcoming bout. The tennis match like any other intensive physical and mental exercise causes the stress reaction in the player’s body, which brings on the release of cortisol. The increase in serum cortisol may influence in multiple ways causing actions to the immune system, stimulating glucose metabolism, and altering mood, memory, and behavioral response to threatening circumstances (Ericson et al., 2003). The high levels of serum cortisol and testosterone concentration during the matches are related to high intensity of the action (Ahtiainen et al., 2003; Häkkinen and Pakarinen, 1995; Kraemer et al., 1989; Wittert et al., 1991). The relationship between serum hormone concentrations and outcome of the match could not be observed in the present study. Traditionally low serum cortisol concentration and high testosterone concentration has been related to positive outcome of the bout (Mazur, 1985; Mazur and Booth, 1998). Comparing the winners and losers in a game like tennis the background of the players and previous encounters should be taken into consideration. According to the cognitive activation theory of stress, the previous experiences from the match situations have effect on the player’s reactions and hormonal responses. Previous success promotes changes to succeed and a lack of success is related to elevated change to failure (Ursin and Eriksen, 2004. ) Also the individuals who do not try to be dominant can experience the winning more stressful than the losing. When taking into consideration previous studies and the present study, it may be reasonable to suggest that winning or losing the match does not have impact on hormonal responses, but the player reacts to match by using his own natural coping mechanisms and the way that suits best to himself (Salvador et al., 2005).

During a tennis tournament player´s physical performance is impaired and the recovery is compromised. Impaired strength levels need to be taken into consideration while planning match preparation and recovery before, during and after the tournament. It is recommended that a after physically tough tournament players need over 24 h rest for the full physical recovery. Player’s physical training must be designed to prepare player’s whole body to cope under heavy speed strength load throughout the match and the tournament. Interval training using normal tennis match work/rest -ratio 1/4, 1/3, 1/2 is recommended.

The tennis tournament impairs player’s MVC and RFD, increases muscle damage and soreness and reduces recovery. It seems that muscle damage is one factor behind the impairment of the physical performance during a tennis match and tournament. Serum cortisol and testosterone concentration elevated before and during the tennis match, but it seems that the result of the tennis match is difficult to predict using cortisol or/and testosterone levels before, during or after the match. After the tournament the levels of serum hormones and physical performance recovered except the RFD. It seems that a tennis tournament causes such a heavy speed strength load for the legs in addition to muscle damage that one day of rest after the tournament is insufficient to recover explosive attributes of leg extensor muscles.