The ability of athletes to perform repeated sprints and changes in direction is regarded by coaches and researchers as a predictor of superior performance in many intermittent and team sports. As well, they are seen as being important indicators of players' fitness in these sports (3,12,26,27,29,32). Studies also show that repeated-sprint ability (RSA) is positively correlated with running distance in top-level soccer matches (22) and that more highly skilled soccer players have better RSA performance compared with lesser skilled players (8,23). Change-of-direction (COD) ability has been used to predict performance and player selection in professional American Football League players (3), and better rugby tacklers tend to have better COD ability (7).
In this study, RSA was defined as “the ability to perform repeated straight sprints or shuttle sprints with minimal recovery between sprint bouts.” Our definition was adapted from a definition used by Spencer et al. (27), which did not specify directional changes during sprints. It is important to specify the type of sprint because different types of maneuvers (e.g., cutting and pivoting) could also be included in RSA tests. Previous RSA studies used protocols such as 15-m shuttle sprints, that is, 30 m per repetition (4,5), 40-m straight sprints (17), and 20-m straight sprints (1).
Our study used the definition of COD suggested by Sheppard and Young (26): “a pre-planned rapid whole-body movement with change of velocity or direction.” Previous studies have investigated various aspects of COD, including the reliability and validity of COD tests (24), determinants of COD (16), and the correlation between power and strength qualities and COD (18). Recently, Haj-Sassi et al. (14) conducted time-motion analyses of a number of field and court sports, suggesting that players perform many intermittent forward, backward, and lateral high speed movements during competitions. The COD is thus an important skill in many sports, and given that it occurs repeatedly throughout competition, repeated COD (RCOD) may be an important fitness component for field and court sports participants. To our knowledge, only 3 RCOD studies have been published, with each employing a different terminology (i.e., “change-of-direction speed,” “repeated-agility sprint ability,” and “multiple-sprint ability,” respectively) (1,30,31), which implies that this ability does not yet have a single, widely accepted name. For the purposes of this study, RCOD was defined as “repeated COD with minimal recovery between bouts.”
During RSA, energy is supplied initially by anaerobic metabolism (e.g., adenosine triphosphate, phosphocreatine and glycolysis), which reduces gradually during subsequent sprints as aerobic metabolism increases (27). The relative contribution of different energy systems in RSA tests can be affected by variables such as sprint type (straight or shuttle sprints), sprint distance, recovery duration between sprints (4,5,17), and recovery mode (1,4). No evidence of the contribution of energy metabolism in RCOD exists. Even though RSA and RCOD are similar, both require repeated high-intensity efforts in a short duration with brief rests, we speculate that these 2 abilities are independent. This was supported by the study of Dellal et al. (6), which found that the physiological impact of intermittent shuttle sprint was substantially higher than that of intermittent in-line running. Also, Lakomy and Haydon (17) found that, in multiple (>10) sprint trials, increased fatigue and slower sprint times occurred for enforced deceleration within 6 m after the finish line as compared with sprints without forced deceleration.
To our knowledge, studies concerning RSA, COD, and RCOD have reported only absolute values. Although informative, it is difficult to compare RSA and RCOD results and to prioritize whether players need to focus on RSA or RCOD training solely by looking at absolute values. Thus, this study proposed a relative, norm-referenced RSA/RCOD index, the goal of which was to provide coaches with an objective tool to assist in prioritizing and individualizing athlete training for improving RSA and RCOD performance.
The purposes of this study were to (a) examine the relationship between RSA and RCOD matched on intervals and distances, (b) determine the discrimination abilities of RSA and RCOD between players with different physical skills, and (c) develop an RSA/RCOD index to assess the relative weightings of these abilities in individual players. Construct validity can be demonstrated when these tests can discriminate between physically active individuals (ACT), college soccer team players (COL), and professional soccer players (PRO). It was hypothesized that PRO would perform better than COL would, and COL better than ACT. It was expected that the RSA/RCOD Index, together with the absolute values of RSA and RCOD performance, would assist soccer coaches and sport scientists in prioritizing the individual training needs of players in RSA and/or RCOD.
Experimental Approach to the Problem
A single visit was made by each subject to the sports hall to perform the RSA and RCOD tests in this within-subject repeated measures study. The subjects warmed up with 5 minutes of jogging at self-paced speed, followed by 1.5- to 2-minute dynamic stretching and several RSA and RCOD familiarity trials at 75, 85, and 100% of maximal speed. The RCOD test began 2 minutes after the warm-up, and the RSA test was conducted after the RCOD test with a 15-minute passive rest between the 2 tests. Extra rest time was allowed upon the request of the subjects to ensure full recovery. The experiment was conducted in an indoor sport court with a wooden sport surface; temperature (25.5 ± 5.01° C) and humidity (59.3 ± 9.14%) were measured hourly throughout the study. To examine the reliability of the measurements, 25 subjects repeated the same RSA and RCOD tests 1 week later.
In total, 59 male subjects participated in this study including 25 ACT, 16 COL, and 18 PRO subjects. The subjects' age, height, body mass, and body mass index (BMI) are shown in Table 1. The study was conducted according to the Declaration of Helsinki and the protocol fully approved by the Human Research Ethics Committee before the commencement of the assessments. Written informed consent was received from all the subjects after giving a brief but detailed explanation about the aims, benefits, and risks of this investigation. The subjects were told that they were free to withdraw from the study at any time without penalty. During the study, all the subjects were instructed to maintain normal daily food and water intake, and no dietary interventions were undertaken. They were also instructed not to participate in vigorous exercises 48 hours before the test.
Repeated-Sprint Ability and Repeated Change-of-Direction Tests
The RSA test comprised straight-line sprints (6 × 20 m with 25-second active recovery) although the RCOD (6 × 20 m with 25-second active recovery) test consisted of four 100° COD at every 4 m as shown in Figure 1 (1). During the active recovery, the subjects jogged slowly back to the starting line and waited for the next sprint. The sprint time for 20 m was measured using an infrared timing system (Brower Timing Systems, Salt Lake City, UT, USA) located at the starting and finishing lines, 1 m above the ground. The investigators used a hand-held stopwatch to monitor recovery time. The players commenced each sprint, starting from a standing position 0.5 m behind the sensor. Each player was encouraged verbally to give a maximal effort during all the RSA and RCOD tests.
The following data were recorded during RSA and RCOD performances: the fastest time (FT), the average time (AT), and the total time (TT) of all sprints and the percentage decrement score (%Dec) as reported by Glaister et al. (13). The use of TT was recommended by previous studies of RSA and COD (1,21,28). The %Dec has been reported recently as the most valid and reliable method for quantifying fatigue in the RSA test (13). The RSA/RCOD index is a ratio of RSA to RCOD measurements. For example, the RSA/RCOD index for TT (Index-TT) for PRO was calculated from the mean of their RSA-TT divided by the mean of their RCOD-TT.
A 1-way analysis of variance (ANOVA) was used to examine the differences in age, body mass, height, BMI, and the RSA/RCOD index between the 3 groups. Pearson product-moment correlation coefficients were used to assess the relationship between RSA and RCOD variables. The magnitude of each correlation was categorized using the modified scale by Hopkins (15): trivial: r < 0.1; low: 0.1–0.3; moderate: 0.3–0.5; high: 0.5–0.7; very high: 0.7–0.9; nearly perfect > 0.9; and perfect: 1. Two-way ANOVA with repeated measures (3 groups × 4 parameters in each test) was used to examine differences in RSA and RCOD performances among the 3 groups. When a significant difference was found in the above analysis, pairwise comparisons were made using Bonferroni's adjustment to control the type-1 error rate. Intraclass correlation coefficient (ICC) and coefficient of variation (CV) of all RSA and RCOD measures among the 25 subjects are shown in Table 2. The ICC of RSA-FT, RSA-AT, RSA-TT, RCOD-FT, RCOD-AT, and RCOD-TT were very high (range: 0.79–0.90). All CV values were ≤10% except RSA-%Dec (46%) and RCOD-%Dec (51%). The significant level was set a priori as p ≤ 0.05.
Significant differences were found in age, height, body mass, and BMI among the 3 groups (p < 0.05). Significant differences between the 3 groups were also found in RSA (p < 0.05) and RCOD (p < 0.05), respectively. Pairwise comparisons revealed that both COL and PRO performed better than ACT did in all RSA and RCOD measures (p < 0.05, Table 1) except for RSA-%Dec and RCOD-%Dec (p > 0.05) where no significant differences were found between COL and PRO. Also, no significant differences were observed between COL and PRO in all RSA and RCOD measures. Significant differences in Index-FT, Index-AT, and Index-TT were found in all comparisons (p < 0.01, Table 3) between ACT and soccer players (COL and PRO), but no significant differences were found between COL and PRO. Table 4 indicates how the RSA/RCOD Index could be applied to prioritizing training.
The RSA-FT, RSA-AT, and RSA-TT show high to very high correlations with RCOD-FT, RCOD-AT, and RCOD-TT (r = 0.69–0.71, p < 0.05, Table 5), although body mass and BMI show low to high correlations with all RSA and RCOD measures (r = 0.28–0.58, p < 0.05) except for the %Dec data for both RSA and RCOD.
The first purpose of this study was to examine the relationship between RSA and RCOD matched on intervals and distances. Our results indicated that RSA-FT was highly correlated with RCOD-FT, RCOD-AT, and RCOD-TT. At the same time, RSA-AT and RSA-TT were very highly correlated to RCOD-FT, RCOD-AT, and RCOD-TT. As a result, we believe that RSA and RCOD have similar metabolic demands. However, our results also indicated that the shared variance (i.e., R2) between RSA and RCOD was between 48 and 50%. This was in agreement with the findings of Brughelli et al. (3), who pointed out that straight sprints and COD were mostly separate motor qualities because the shared variance was ≤53% based on reviewed literature. Therefore, we concluded that RSA and RCOD were different motor abilities requiring specific training.
Previously, it has been found that lateral, horizontal, and vertical jump abilities were limited in predicting COD performance (R2 < 43%) (19). In addition, the training effects of straight sprint and COD were found to be nontransferable (3). Weaker relationships were found when the COD tasks are more complex, such as more changes in direction and increased angles of COD (32), which were attributed to different running techniques used by subjects during COD as compared with straight sprint (25). Sheppard and Young (26) noted that the training effects transfer might be even weaker if technical sport skills were added to the COD task. Specifically, the movement paths of straight sprint and COD are different, requiring specific running techniques and neuromuscular adaptations for each.
The RSA-FT was almost perfectly correlated with RSA-AT and RSA-TT (r = 0.96, p < 0.01), as was RCOD-FT with RCOD-AT and RCOD-TT (r = 0.99, p < 0.01). This demonstrated that the fastest trial in straight sprint and COD had the greatest influence on the ultimate RSA and RCOD performance, respectively. One possible reason for this finding is that the RSA and RCOD protocols used in this study induced less fatigue in the participants than did the protocols used in previous studies. Pyne et al. (21) employed 6 × 30-m trials with an about 15-second active recovery test protocol and found that 20-m straight sprint speed had a higher correlation with RSA than did a 20-m multistage shuttle endurance test (r = 0.66 vs. −0.20). The RSA protocol used by Pyne et al. (21) induced a modest degree of fatigue, and therefore with shorter distance and longer recovery duration, it is reasonable to believe that the protocol used in this study would induce less fatigue compared with that of Pyne et al. (21).
Nimphius et al. (20) found that both sprint speed and COD were very highly correlated with body mass in female athletes over the course of a season (r = 0.70–0.93, p < 0.05). This was also supported by a cross-sectional study by Gabbett et al. (11). This study supported this, in part, because body mass and BMI were significantly correlated to RSA and RCOD (r = 0.28–0.58, p < 0.05). In view of the above discussion and the varying relationships reported between body mass, physical fitness, and sport skill (9–11), it seems that a more holistic approach should be taken in future studies. For instance, body composition including fat mass, muscle mass, and bone density should be measured and correlated with RSA and RCOD performance. In this way, further clarification of the relationship between RSA, RCOD, and body composition will become clearer.
The second purpose of the study related to the discrimination abilities of RSA and RCOD. We hypothesized that PRO would perform better than COL and that COL would perform better than ACT. Based on the results of this study, RSA and RCOD measures distinguished between ACT and soccer players (COL and PRO) in FT, AT, and TT (p < 0.05), and therefore, construct validity is established. The RCOD test protocol used in this study employed forward diagonal COD instead of lateral COD (as in agility t-test), which may have favored the soccer players (COL and PRO). This is supported by the study of Bloomfield et al. (2) who found that professional soccer players spent a greater percentage of time in forward diagonal (right: 5.0%; left: 4.6%) rather than lateral (right: 3.9%; left: 4.5%) directions during purposeful movements. In this study, PRO had better but statistically insignificant performances than COL in RSA and RCOD measured as FT, AT, and TT. Previous studies found that RSA not only differentiated professional soccer players from amateur soccer players (23) but also between national- and state-level female soccer players (8). On the other hand, reports on comparison of RCOD results between different levels of players do not exist. However, COD performance of elite and subelite players have been shown to be similar (9,11).
The third purpose of the study was to develop an RSA/RCOD Index to assess the relative RSA and RCOD abilities of individual subjects. This index would assist soccer coaches and sport scientists in prioritizing the training needs of players, focusing training on RSA, RCOD, or both. To our knowledge, this study was the first to introduce the use of an RSA/RCOD Index. Significant differences in Index-FT, Index-AT, and Index-TT were found in ACT compared with both COL and PRO (p < 0.05). By comparing a player's RSA/RCOD Index with the norm-referenced ratio reported in this study, coaches could determine whether RSA or RCOD would be a priority for improvement (an illustrative example is given in the Practical Applications). However, it is important to note that the current index was generated from a small sample size, and caution should be exercised in generalization to other populations. Typically, a representative index norm requires a larger number of samples chosen specifically for the sport, and therefore, different sports and skill levels may have different index values.
This study demonstrated that FT, AT, and TT of both RSA and RCOD were reliable measures (CV ≤ 10% for all; ICC ≥ 0.80, except the reliability of RCOD-FT = 0.79). The RSA-TT and RCOD-TT were also shown to be consistent with that of other studies (1,21,28). In contrast, the %Dec showed great variation (ICC: 0.19 and 0.11; CV: 51 and 46% for RSA and RCOD, respectively) and failed to distinguish subjects in any groups. Moreover, only RCOD-% Dec had a significant but low correlation with RCOD-AT and RCOD-TT (r = 0.29, p < 0.05). Glaister et al. (13) concluded that %Dec was the most reliable and valid method to quantify fatigue in RSA among 8 proposed methods. Yet they also found comparable ICC and CV values as found in this study. Therefore, we recommend evaluating FT, AT, and TT in RSA and RCOD tests, even though they do not provide any direct evidence of the effect on players of fatigue.
The FT, AT, and TT of both RSA and RCOD test protocols were shown to be reliable and valid, discriminating physically active individuals from soccer players (COL and PRO). The RSA and RCOD are separate motor abilities (R2 ≤ 50%) that require separate, specific training. The proposed RSA/RCOD Index is a norm-referenced ratio that can help coaches prioritize training needs in RSA and RCOD abilities in soccer players. For example, a COL player with an index of 0.55 (20.59 seconds/37.44 seconds) can achieve the standard index of 0.59 by either improving his RCOD or slowing down his RSA. Obviously, this player should train to improve RCOD. However, the average performances of COL in this study were 19.8 and 33.7 seconds for RSA and RCOD, respectively, which means that this player would also need to improve his RSA because it is below the group average. In the final analysis, the more urgent need for this player is to focus on improving RCOD. As this illustrative example indicates, coaches may use the index and the norm established in this study to assess whether an improvement in RSA or RCOD is needed more urgently. At the same time, they will still need to look at the absolute performance of players to fine tune the effectiveness of training and optimize an individual player's performance.
The results of this study do not constitute endorsement of the product by the authors or the National Strength and Conditioning Association. There was no conflict of interest declared. This study was not supported by any sources of funding.
1. Beckett JR, Schneiker KT, Wallman KE, Dawson BT, Guelfi KJ. Effects of static stretching on repeated sprint and change of direction performance. Med Sci Sports Exer 41: 444–450, 2009.
2. Bloomfield J, Polman R, O'Donoghue P. Physical demands of different positions in FA premier league soccer. J Sports Sci Med 6: 63–70, 2007.
3. Brughelli M, Cronin J, Levin G, Chaouachi A. Understanding change of direction ability in sport: A review of resistance training studies. Sports Med 38: 1045–1063, 2008.
4. Buchheit M, Mendez-Villanueva A, Delhomel G, Brughelli M, Ahmaidi S. Improving repeated sprint ability in young elite soccer players: Repeated shuttle sprints vs. explosive strength training. J Strength Cond Res 24: 2715–2722, 2010.
5. Castagna C, Manzi V, D'Ottavio S, Annino G, Padua E, Bishop D. Relation between maximal aerobic power and the ability to repeat sprints in young basketball players. J Strength Cond Res 21: 1172–1176, 2007.
6. Dellal A, Keller D, Carling C, Chaouachi A, Wong DP, Chamari K. Physiologic effects of directional changes in intermittent exercise
in soccer players. J Strength Cond Res 24: 3219–3226, 2010.
7. Gabbett TJ. Physiological and anthropometric correlates of tackling ability in rugby league players. J Strength Cond Res 23: 540–548, 2009.
8. Gabbett TJ. The development of a test of repeated-sprint ability for elite women's soccer players. J Strength Cond Res 24: 1191–1194, 2010.
9. Gabbett TJ, Jenkins DG, Abernethy B. Physiological and anthropometric correlates of tackling ability in junior elite and subelite rugby league players. J Strength Cond Res 24: 2989–2995, 2010.
10. Gabbett TJ, Jenkins DG, Abernethy B. Correlates of tackling ability in high-performance rugby league players. J Strength Cond Res 25: 72–79, 2011.
11. Gabbett TJ, Kelly JN, Sheppard JM. Speed, change of direction speed, and reactive agility
of rugby league players. J Strength Cond Res 22: 174–181, 2008.
12. Glaister M. Multiple sprint work: Physiological responses, mechanisms of fatigue and the influence of aerobic fitness. Sports Med 35: 757–777, 2005.
13. Glaister M, Howatson G, Pattison JR, McInnes G. The reliability and validity of fatigue measures during multiple-sprint
work: An issue revisited. J Strength Cond Res 22: 1597–1601, 2008.
14. Haj-Sassi R, Dardouri W, Gharbi Z, Chaouachi A, Mansour H, Rabhi A, Mahfoudhi ME. Reliability and validity of a new repeated agility
test as a measure of anaerobic and explosive power. J Strength Cond Res 25: 472–480, 2011.
15. Hopkins WG. Measures of reliability in sports medicine and science. Sports Med 30: 1–15, 2000.
16. Jones P, Bampouras TM, Marrin K. An investigation into the physical determinants of change of direction speed. J Sports Med Phys Fitness 49: 97–104, 2009.
17. Lakomy J, Haydon DT. The effects of enforced, rapid deceleration on performance in a multiple sprint test. J Strength Cond Res 18: 579–583, 2004.
18. Marcovic G. Poor relationship between strength and power qualities and agility
performance. J Sports Med Phys Fitness 47: 276–283, 2007.
19. Meylan C, McMaster T, Cronin J, Mohammad NI, Rogers C, Deklerk M. Single-leg lateral, horizontal, and vertical jump assessment: Reliability, interrelationships, and ability to predict sprint and change-of-direction performance. J Strength Cond Res 23: 1140–1147, 2009.
20. Nimphius S, McGuigan M, Newton RU. Relationship between strength, power, speed, and change of direction performance of female softball players. J Strength Cond Res 24: 885–895, 2010.
21. Pyne DB, Saunders PU, Montgomery PG, Hewitt AJ, Sheehan K. Relationships between repeated sprint testing, speed, and endurance. J Strength Cond Res 22: 1633–1637, 2008.
22. Rampinini E, Bishop D, Marcora SM, Ferrari Bravo D, Sassi R, Impellizzeri FM. Validity of simple field tests as indicators of match-related physical performance in top-level professional soccer players. Int J Sports Med 28: 228–235, 2007.
23. Rampinini E, Sassi A, Morelli A, Mazzoni S, Fanchini M, Coutts AJ. Repeated-sprint ability in professional and amateur soccer players. Appl Physiol Nutr Metab 34: 1048–1054, 2009.
24. Sassi RH, Dardouri W, Yahmed MH, Gmada N, Mahfoudhi ME, Gharbi Z. Relative and absolute reliability of a modified agility
T-test and its relationship with vertical jump and straight sprint. J Strength Cond Res 23: 1644–1651, 2009.
25. Sayers M. Running techniques for field sport players. Sports Coach 23: 26–27, 2000.
26. Sheppard JM, Young WB. Agility
literature review: Classifications, training and testing. J Sports Sci 24: 919–932, 2006.
27. Spencer M, Bishop D, Dawson B, Goodman C. Physiological and metabolic responses of repeated-sprint activities: Specific to field-based team sports. Sports Med 35: 1025–1044, 2005.
28. Spencer M, Fitzsimons M, Dawson B, Bishop D, Goodman C. Reliability of a repeated-sprint test for field-hockey. J Sci Med Sport 9: 181–184, 2006.
29. Spencer M, Lawrence S, Rechichi C, Bishop D, Dawson B, Goodman C. Time-motion analysis of elite field hockey, with special reference to repeated-sprint activity. J Sports Sci 22: 843–850, 2004.
30. Walklate BM, O'Brien BJ, Paton CD, Young W. Supplementing regular training with short-duration sprint-agility
training leads to a substantial increase in repeated sprint-agility
performance with national level badminton players. J Strength Cond Res 23: 1477–1481, 2009.
31. Wilkinson M, McCord A, Winter EM. Validity of a squash-specific test of multiple sprint ability. J Strength Cond Res 24: 3381–3386, 2010.
32. Young WB, McDowell MH, Scarlett BJ. Specificity of sprint and agility
training methods. J Strength Cond Res 15: 315–319, 2001.
Keywords:© 2012 National Strength and Conditioning Association
multiple-sprint; intermittent exercise; football; agility