Exponential versus linear tapering in junior elite soccer players : effects on physical match performance according to playing positions

Th e purpose of the present study was to investigate the eff ects of exponential and linear tapering protocols on the distance covered according to playing positions among junior elite soccer players. One-hundred and fi ft y-eight junior elite soccer players (mean age: 17.07±0.79 years; mean height: 177.85±6.64 cm; mean weight: 71.27±7.96 kg; mean body-mass index: 22.50±1.66 kg/m2) were randomly selected into two groups: (1) exponential and (2) linear tapering group. Training sessions were conducted three times a week for eight weeks. Aft er four and eight weeks of the tapering period, participants were measured in the distance covered according to four basic playing positions on the fi eld: (1) goalkeepers, (2) defenders, (3) midfi elders, and (4) forwards. Th e results showed that the largest eff ects between the preand post-measurement had defenders in the exponential and goalkeepers in the linear group, goalkeepers in the distance covered by slow and medium running in both groups, forwards in fast running and sprinting in the exponential group, and midfi elders in the linear group. In conclusion, our results suggest that the exponential tapering protocol brought somewhat larger eff ects, especially in fast running and sprinting in forwards and slow and medium running in goalkeepers. Future studies need to implement both the exponential and the linear tapering protocols in order to enhance performance but favouring the exponential to produce larger eff ects.


Introduction
Th e game of soccer is oft en described as a team and contact sport characterized by diff erent running intensities, jumps, acceleration and deceleration, and tackling (Valter Di Salvo et al., 2007;Valter Di Salvo, Gregson, Atkinson, Tordoff , & Drust, 2009).Previous studies have shown that senior elite players cover between 10,500 and 12,000 metres per soccer match, with around 10% of high-intensity activities (Dellal et al., 2011;Dellal et al., 2010).Junior soccer players, however, cover between 9,000 and 10,000 meters per soccer match, have lower heart rate responses (82% vs 93% in senior players) and blood lactate concentrations (Wong et al., 2008).Additionally, a few studies have found positional diff erences between players, especially in the anthropometric characteristics and diff erent physiological demands of each player (Bloomfi eld, Polman, & O'Donoghue, 2007;Rampinini, Coutts, Castagna, Sassi, & Impellizzeri, 2007).In general, midfi elders cover a greater distance than other players on diff erent playing positions, whereas forwards perform more sprinting activities and defenders dribble a shorter distance (Rampinini et al., 2007).
Over the past two decades, numerous studies have recorded activities performed by the players in diff erent playing positions, aimed to discover which physical match activities are most important for soccer requirements (Buchheit et al., 2014;Valter Di Salvo et al., 2009;Fessi et al., 2016;Sarmento et al., 2014).Th e physiological and psychological requirements for soccer can be achieved through training load reduction (Inigo Mujika & Padilla, 2003).Training load reduction, i.e. reduction of the physiological and psychological load before the competition period, is known as "tapering" (Inigo Mujika & Padilla, 2003).Th e tapering protocol may improve and enhance performance in both athletes from individual and team sports (Bosquet, Montpetit, Arvisais, & Mujika, 2007;Elloumi et al., 2012).Previous fi ndings have shown that a high training load during the training period could impair performance (Slattery, Wallace, Bentley, & Coutts, 2012).In general, it has been proposed that training load should decrease in order to reduce the frequency and volume of training but not the intensity (Bosquet et al., 2007).
To the best of our knowledge and aft er an extensive literature review, there has only been one study investigating the eff ects of the tapering protocol on physical match performance in soccer players (Fessi et al., 2016).Th eir results showed large eff ect changes in intensity running, high-intensity running, high-speed running activities, and the number of sprints between the standard and the taper week in favour of the taper week.However, there has been no study investigating the eff ects of two diff erent tapering protocols on physical match performance in soccer players.Neither has there been a study investigating those eff ects according to diff erent playing positions.Th us, the main purpose of the present study was to investigate the eff ects of the exponential and linear tapering protocol on distance covered according to playing positions in junior elite soccer players.

Participants
One-hundred and fi ft y-eight (N=158) elite Croatian junior soccer players (mean age: 17.07±0.79yrs.; mean height: 177.85±6.64cm; mean weight: 71.27±7.96kg; mean body-mass index: 22.50±1.66kg/m², mean training experience: 9.42±1.54years) were randomly selected into the exponential tapering group and linear tapering group.Randomization was done with replacement, in which each participant had an equal chance of being selected.Th ere were 14 forwards, 18 defenders, 40 midfi elders, and 7 goalkeepers in the exponential group and 15 forwards, 13 defenders, 45 midfi elders, and 6 goalkeepers in the linear taper group.Basic descriptive statistics of the study participants in the distance-covered variables are presented in Table 2. Before the study began, each participant gave assent, and their parents/guardians had given informed consent to participate in the study.All participants were told about potential risks during the study.During the study, participants were not allowed to be in another training program that could potentially bias the results.All the procedures performed in this study were in accordance to the Declaration of Helsinki and were approved by the Institutional Review Board of the Faculty of Kinesiology, University of Split, Croatia.

Match activities
With the Focus 3 analyser system, we also included the distance covered by players during the soccer match and categorized this into fi ve classes: (1) walking/jogging (0.4-3.0 km/h), slow running (3.0-8.0 km/h), medium running (8.0-13.0km/h), fast running (13-18 km/h) and sprint (>18 km/h).We also included total distance covered, as a sum of all categories combined.All the results were expressed in metres.

Testing protocol
In the fi rst phase of the study, in agreement with soccer clubs, all the measurements were done in the morning period between 9.00-12.00h.Two days before the testing, the participants did not have any type of training with signifi cant load, which could potentially aff ect the results.All variables within the study were measured three times: the initial, transitive and fi nal periods.Th e transitive measurement was done four weeks aft er the initial measurement.Aft er the transitive measurement, both experimental groups underwent diff erent tapering protocols: linear or exponential.Training for both groups was held three sessions per week and consisted of 4×4 min running exercises with the intensity of 90-95% heart rate max separated with a four-minute jogging period of 40% heart rate max .Th e tapering protocol lasted for four weeks, followed by fi nal measurement.In total, the whole protocol lasted for eight weeks.It is noteworthy that coaches used the same training methods of teaching.Furthermore, all participants had similar levels of physical activity outside the testing period and had similar diet protocols.Detailed protocols of both tapers are presented in Table 1.

Statistical analysis
Basic descriptive parameters are presented as means ± standard deviations.To assess whether the data were normally distributed, a Kolmogorov-Smirnov test was used.To test if the main eff ect of the factor Group (exponential vs linear), the factor Position (goalkeepers vs defenders vs midfi elders vs forwards), the factor Time (pre-test, mid-test, and post-test) and the factorial interactions Group×Position, Group×Time, Position×Time and Group×Position×Time were signifi cant, a 3-factorial between-between-within 2×4×3 ANOVA was used.Homogeneity of variance was tested by using Leven's test, and diff erences between groups and trials were determined by using the Bonferroni correction.Data have been identifi ed as outliers if was out of M ± 2σ interval.Partial-eta squared (partial η 2 ) was used for eff ect size assessment.Statistical analyses were performed by using the Statistical Package for Social Sciences soft ware (SPSS ver.23).Type one error was set at α=5%.

Results
First, we calculated the coeffi cient of variation (CV) for every variable measured three times.Results showed small variation between three measurements in height (CV=0.001),sitting height (CV=0.002),weight (CV=0.002),body-mass index (CV=0.004),% of fat-mass (CV=0.006),% water (CV=0.001)and % of muscle mass (CV=0.002).In motor abilities, the results in CV showed somewhat larger variations in 5 m sprint (CV=0.083), 10 m sprint (CV=0.059),30 m sprint (CV=0.the eff ect size between pre-and post-measurement ranged within 0.43 in forwards to 0.72 in defenders in the exponential group, and between 0.60 in midfi elders and 1.07 in goalkeepers.In the slow and medium running variable, the biggest changes occurred in goalkeepers in both the exponential and linear groups between preand post-measurement.However, forwards in the exponential group enhanced their results between pre-and post-measurement (ES=0.95) in the distance covered by fast running than forwards in the linear group did (ES=0.41).Th e largest eff ect size occurred in forwards and midfi elders in the exponential group between the pre-and post-measurement and in midfi elders in the linear tapering group.Defenders and midfi elders in both the exponential and linear tapering groups had the largest distance covered during the game: between 10,000 and 10,500 meters.

Discussion
Th e purpose of the present study was to investigate the eff ects of exponential and linear tapering protocols on the distance covered according to playing positions in junior elite soccer players.Our results showed signifi cant improvements following both linear and exponential tapering periods in soccer players in all playing positions.As mentioned by some previous studies, it is diffi cult to analyse match performance only through distance covered, since soccer represents a highly complex game (Coutts, Chamari, Rampinini, & Impellizzeri, 2008).However, our results are in accordance with some previous studies, which have reported certain aerobic enhancements in both individual sport athletes (Costill, King, Th omas, & Hargreaves, 1985;Inigo Mujika & Padilla, 2003;Sanchez et al., 2013) and team sport athletes (Coutts et al., 2008;Elloumi et al., 2012).In one of the most recent studies, the results showed that maximal oxygen uptake increased aft er a two-week tapering program in the experimental group of soccer players (Fortes, Vianna, Silva, Gouvêa, & Cyrino, 2016).It has been previously established that maximal oxygen uptake enhancements occur due to activation of the PGC-1α complex related to carbohydrates and fats control and enhancing fat and glucose oxidation, which can potentially improve aerobic endurance (Silva & Araújo, 2015).Similar physiological changes by following a tapering period have been reported in other sports, like kayaking (García-Pallarés, Sanchez-Medina, Pérez, Izquierdo-Gabarren, & Izquierdo, 2010), cycling (Neary, Martin, & Quinney, 2003), swimming (Trappe, Costill, & Th omas, 2001), and running (I Mujika et al., 2002).Moreover, tapering has been shown to increase oxygen extraction important for aerobic activities (Neary et al., 2003), to reduce muscle damage and catabolism, and to enhance anabolism and muscle glycogen stores (Coutts et al., 2008).
Next, our results showed that forwards had the largest changes between pre-and post-measurement in the exponential group, followed by defenders, yet the largest diff erences were observed in defenders in the linear group in the fast running variable.Similar fi ndings were obtained in sprinting activities.Since this is the fi rst study examining the eff ects of two diff erent types of tapering on the distance covered during the soccer match, we concentrated on physiological and playing position demands required by each player.For example, fi ndings have proved that forwards perform the most maximal sprints and for longer durations, along with higher levels of stopping and faster deceleration activities (Bloomfi eld et al., 2007).Furthermore, forwards are more engaged in actions that require specifi c activities, like jumping and heading the ball (Reilly, 2003).It is possible that forwards improved their performance, since studies have shown that velocity, agility, and speed are the most important characteristics during a soccer match (Gil, Gil, Ruiz, Irazusta, & Irazusta, 2007).Moreover, morphologically, they are much leaner and stronger players with somewhat better physiological characteristics than the other players on diff erent playing positions, indicating that the result of the game primarily depends on the forwards group (Gil et al., 2007).
Along with forwards, midfi elders showed the largest changes in sprinting activities.Th is could be explained by the fact that midfi elders are engaged in signifi cantly less walking and low running intensities, but spend the most time in running and sprinting activities (Bloomfi eld et al., 2007).Studies have also shown that midfi elders have higher levels of maximal oxygen uptake (VO2max) and cover greater distances in contrast to the players in other playing positions (Reilly, Bangsbo, & Franks, 2000).
Goalkeepers were shown to produce the largest changes in walking/jogging and slow running activities in our study, which could be related to the fact that they perform the poorest in the endurance tests (Tumilty, 1993).Defenders in our study had the largest changes in slow running in contrast to other playing positions.In gen- eral, time changes in our study according to playing position could be explained by diff erent playing roles on the fi eld during the soccer match.Since diff erent requirements are needed from each player, we speculate that both tapering protocols produce diff erent eff ect changes.For example, goalkeepers mostly cover distance with walking/jogging and low intensity running.Since they use such activities the most, it is possible that tapering infl uenced the most specifi cally on those parameters.Moreover, similar values were obtained among forwards and midfi elders, who had the largest changes in fast running and sprinting activities, especially in the exponential group.One meta-analysis has shown that training load reduction between 40-60% from maximal load in a period of two weeks causes the largest improvements in performance (Bosquet et al., 2007).Th is kind of training load reduction seems to have benefi cial eff ects against muscle fatigue and training stress during the tapering period.Furthermore, we only looked at physical soccer match parameters, but not on psychological parameters.Future studies should take both physiological and psychological parameters into account, since previous fi ndings have shown that certain monotony is related to the onset of overtraining combined with high training loads (Foster, 1998).
Our study has several limitations.First, we did not control for diet and other physical activities of the participants during the tapering period, which may potentially lead to bias.However, all the participants were instructed prior to the study to have somewhat similar diet and not to participate in other sport activities.Second, we did not analyse technical and tactical elements during the soccer match, along with the psychological abilities.It is necessary to establish the relation between technical-tactical elements and distance covered in diff erent running intensities (Fessi et al., 2016).
In conclusion, our results show that both the exponential and linear tapering protocols improved distance covered by diff erent running intensities in soccer players according to their playing positions.However, the exponential tapering protocol brought somewhat larger change eff ects, especially in forwards and midfi elders in fast running and sprinting activities during the soccer match.Furthermore, results suggest that goalkeepers had the biggest improvements in distance covered by walking/jogging and low running, since their playing role requires this kind of activity.As mentioned before, future studies should focus on investigating how diff erent tapering protocols may aff ect physical and psychological abilities and sport-specifi c (technical and tactical) elements during the soccer match.
Avoiding overtraining and optimizing performance could be achieved through the logical variation of training methods and volume loads.As mentioned earlier, a taper involves a reduction in the physiological and psychological stress, which could potentially enhance performance.Our results showed that decreasing the 5% of initial values or 5% of the previous session values in every forthcoming workout had similar eff ects on distance covered by diff erent running intensities in soccer players according to their playing positions.However, forwards and midfi elders in fast running and sprinting activities during the soccer match showed better improvements following the exponential tapering.Th e novel fi nding in the current study is that sprinting during matches was increased aft er exponential tapering in junior soccer players.However, caution is advised when interpreting the results of this study, bearing in mind the complexity and the various factors that could infl uence physical match activities.Our results confi rmed the reports of others, which suggest that volume is the optimal variable to manipulate exponentially, i.e., reducing the volume of training while maintaining both the intensity and the frequency of sessions.D ellal, A., Chamari, K., Wong, d. P., Ahmaidi, S., Keller, D., Barros, R., . . . Carling, C. (2011).Comparison of physical and technical performance in European soccer match-play: FA Premier League and La Liga.

TABLE 1
Training protocol for both tapers (3×/week, 4×4 min of running to from 90-95% heart ratemax) EFFECTS OF TAPERING ON PERFORMANCE IN SOCCER | V. CIGROVSKI ET AL.

Fast running (13.0-18.0 km/h)
025), 96369 agility test (CV=0.023),repeatedsprintability (CV=0.018),squatjump(CV=0.033)andsquatjump with arm swing (CV=0.029).Last, in functional abilities, results in VO2max showed small CV (0.012).Basic descriptive statistics of the study participants are presented in Table2.As shown, defenders and midfi elders had the biggest distance covered, opposed to forwards and goalkeepers (p<0.05).Midfi elders covered around 2000 metres in medium running, yet midfi elders and forwards covered approximately 400 m by sprinting.In the distance covered by walking/jogging, TABLE 2 Basic descriptive statistics of the study participants (N=158) *-signifi cant diff erence between mid and post-test; †-Signifi cant diff erence between groups EFFECTS OF TAPERING ON PERFORMANCE IN SOCCER | V. CIGROVSKI ET AL.