Static stretching is generally considered an integral component of sport conditioning programs and preexercise warm-up routines (13). This type of stretching is often practiced in the belief that it assists athletic performance, reduces the risk of injury, and decreases muscle soreness resulting from strenuous activity (13,23). However, recent research has challenged some of these beliefs regarding the benefits of static stretching. In particular, studies have shown static stretching to induce significant acute deficits in muscular power (14,24), torque (6,14), force (19,26), and maximal strength (11) as well as jump (4,7,25), sprint (8,17), and agility performance (15).
Of importance, these studies have typically involved only a single sprint or power effort. Yet, field-based team sports such as rugby, field hockey, soccer, and Australian football require players to perform multiple maximal sprints interspersed with brief, mostly active, recovery periods in between efforts (5,22). In addition, team-sport performance requires many changes of direction while sprinting (5,10,22). Although an investigation into the effect of static stretching on this performance parameter has shown impaired performance compared with a dynamic warm-up (15), others have observed no effect of static stretching on agility when incorporated into a warm-up routine (12). Furthermore, no studies have examined the effect of conducting static stretching within the performance activity (i.e., during recovery periods between efforts such as during interchange or breaks in play). Therefore, it is still unknown whether static stretching at these times is detrimental to the performance of repeated maximal efforts.
Other limitations of previous research include the use of protocols that are not representative of typical warm-up methods used by athletes in performance settings. For example, the total stretch durations implemented in previous studies range from 90 s (11) to 30 min per muscle group (9), when in practice, mean stretch time is 10-20 s per muscle group, repeated two to three times (1). Also, some studies have applied static stretching in isolation of other warm-up components that are normally performed such as aerobic activity and dynamic drills (3).
The objective of this study was to determine the effect of static stretching during recovery periods between efforts on repeated sprint ability (RSA) and also repeated change of direction speed (CODS) performance while closely emulating typical stretching practices. It was hypothesized that RSA and CODS performance would be significantly slower after static stretching in comparison to passive rest. In addition, it was hypothesized that any deficit in RSA performance would be most evident in the first 5 m of the sprint given that muscular force production requirements are greatest during the initial takeoff phase of a sprint (16).
Twelve healthy, uninjured males (mean ± SD: age = 23 ± 4 yr; height = 178.7 ± 5.8 cm; body mass = 72.8 ± 5.9 kg) were recruited as participants. All were regular team-sport players (soccer, n = 6; Australian football, n = 4; rugby, n = 2; and field hockey, n = 1) and were tested toward the end of their respective competitive seasons. Institutional ethics approval and individual written informed consent were obtained before the commencement of testing.
Participants attended the outdoor laboratory on five occasions, each separated by approximately 1 wk (7 ± 1 d). The first visit involved familiarization with the RSA and the CODS tests as well as the warm-up and stretching protocols to be used in the experimental trials. On the remaining four occasions, participants were required to perform a standardized warm-up followed by either the RSA or the CODS performance test on two occasions each. Both tests involved three sets (of six repetitions each) with a 4-min recovery between sets. During the break between sets, one of the following conditions was applied: control (CON) in which participants were required to stand and rest or static stretch (SS) in which participants were required to complete a 4-min static stretching protocol. The four experimental trials (RSA-CON, RSA-SS, CODS-CON, and CODS-SS) were administered randomly using a Latin square design. Each participant was tested at the same time of day across all trials to control for circadian variability. Participants were asked to abstain from intense physical activity for 48 h before testing.
Each trial was commenced with a standardized warm-up consisting of general aerobic activity, dynamic activities, and run-throughs. The aerobic activity consisted of 5 min of submaximal jogging performed outdoors on a 40-m marked grass track (20 laps maintaining a pace of 15 s per lap or ∼9.6 km·h−1). The dynamic activities involved sports-specific movements, including 80 m each of buttock kicks, high knee lifts, and straight leg skipping, followed by four 40-m laps of carioca, alternating direction each lap. Run-throughs consisted of three repetitions of the applicable test (RSA or CODS) conducted at progressively increasing speeds (60%, 80%, and 100% of perceived maximum). Participants then walked five laps of a 40-m track in 3 min (∼4 km·h−1) followed by 1 min of passive rest.
Static stretching protocol.
The static stretching protocol implemented during the RSA-SS and the CODS-SS trials consisted of a total of six static stretches, each targeting the prime movers of the lower extremities (hamstrings, quadriceps, gastrocnemius, soleus, hip flexors, and adductors, respectively). All stretches were held to the point of slight discomfort (not pain) for a period of 20 s per muscle group per limb (with the exception of the adductor stretch which targeted both limbs simultaneously).
Repeated sprint ability (RSA) test.
The RSA test consisted of three sets of 6 × 20-m maximal sprints, going every 25 s, with an active recovery between sprints (i.e., jog back to start position). The test was performed on an outdoor grass surface to replicate typical team-sport conditions. A 20-m sprint distance was chosen as it approximates the mean sprint distance in common field-based team sports (22). Participants were given strong verbal encouragement throughout all trials to ensure maximal effort for each sprint. Each set of the RSA test was separated by a 4-min recovery period during which the experimental stretching or control intervention was administered. Sprint times were recorded using electronic timing gates (Fusion Sport Smart Speed, Wales, UK) located at the start and finish lines. In addition, 5- and 10-m split times were recorded. For the six sprints completed in each set, both mean sprint time (MST) and total sprint time (TST; sum of six sprints) were subsequently determined. Other performance measures included the first (FST) and the best sprint time (BST) of each set. The typical error and the coefficient of variance for one set of 6 × 20 m were 0.060 s and 1.8%, respectively, whereas for best 20-m time were 0.19 s and 1.1%, respectively.
Change of direction speed (CODS) performance test.
The CODS performance test consisted of three sets of 6 × 20 m maximal sprints, going every 25 s with an active jog back recovery between efforts. The test was performed on an outdoor grass surface and included four changes of direction, each 100°, with 4 m of straight-line sprinting before and after each turn (Fig. 1). These relatively tight turns were chosen to maximize participant reliance on acceleration and deceleration while maintaining the angle of the change in direction similar to that commonly observed in the field (5,10). Participants were again given strong verbal encouragement to ensure maximal effort, and all sets were separated by a 4-min recovery period during which the experimental stretching or control intervention was administered. Similar to the RSA test, sprint times for the CODS test were recorded using electronic timing gates located at the start and finish lines. For the six sprints completed in each set, both MST and TST were determined as well as FST and BST. After completing all four experimental trials, participants were asked to nominate whether they preferred the stretch or the no-stretch protocol during the recovery periods between sets.
Repeated-measures ANOVA was used to compare RSA and CODS performance both within (set 1 vs set 2 vs set 3) and between the two experimental conditions (SS vs CON), with post hoc analyses (Fisher's least significant difference) used where appropriate. The data from set 1 were excluded from the assessment of between treatment effects because the intervention (SS or CON) was not administered until after set 1 of the sprints. These analyses were carried out using the Statistical Package for the Social Sciences for Windows (version 13; SPSS, Inc., Chicago, IL), and statistical significance was set at P < 0.05. In addition, trends in performance were interpreted using Cohen's d effect sizes (pooled SD) and thresholds (<0.4, small; 0.4-0.7, moderate; >0.7, strong). Smallest worthwhile effects were also calculated for all variables to determine the likelihood that the true effect was substantially beneficial, trivial, or harmful (detrimental). The threshold value for smallest worthwhile change in 20-m sprint time was set at 0.8% (18), whereas CODS times were set at 0.2 (Cohen's units) (21). If both benefit and harm were calculated to be >5%, the true effect was assessed as unclear. Where clear interpretation could be made, chances of benefit or harm (called detriment in this manuscript) were assessed as follows: <1%, almost certainly not; 1-5%, very unlikely; 5-25%, unlikely; 25-75%, possibly; 75-95%, likely; 95-99%, very likely; and >99%, almost certainly (2).
Repeated Sprint Ability
Mean sprint time (MST).
The MST for the RSA test are presented in Table 1. Within trials, MST increased (i.e., became slower) in set 2 compared with set 1 for the RSA-SS trial (P = 0.006). In addition, set 3 was significantly slower than set 1 for both RSA-CON (P = 0.026) and RSA-SS trials (P = 0.033). Between trials, MST for each set was consistently slower in the RSA-SS trial after the stretching protocol. Although this was not a statistically significant main effect, post hoc analysis revealed that MST was significantly slower in set 2 over the 0- to 5-m and the 0- to 20-m sprint splits in the RSA-SS trial compared with the RSA-CON (P = 0.044 and 0.031, respectively). The qualitative analysis supported this finding with a moderate effect size (d = 0.40) and a "likely" (91%) detrimental effect for RSA-SS compared with RSA-CON over the first 5 m of the sprints in set 2. Qualitative analysis also revealed "likely" detrimental effects associated with RSA-SS for 20-m sprint times (0-20 m) in set 2 (83%) and 0- to 5-m split times in set 3 (80%) compared with RSA-CON.
Total sprint time (TST).
Similar to MST, TST increased (became slower) in set 2 compared with set 1 in the RSA-SS trial, whereas set 3 was slower than set 1 for both RSA-SS and RSA-CON (Table 1). Between trials, TST for set 2 was found to be significantly slower in RSA-SS compared with RSA-CON (P = 0.031). Further, qualitative analysis supported this finding with a "likely" (83%) detrimental effect associated with RSA-SS compared with RSA-CON. No other differences were noted between trials for TST.
First and best sprint times (FST and BST).
The split times for FST and BST of the RSA test are presented in Table 2. There was no significant within-trial effect for either FST or BST. Likewise, there was no statistical difference in these variables between treatments. However, the qualitative analysis suggested that there was a tendency for FST to be slower in RSA-SS with moderate effect sizes and "likely" detrimental effects for FST in set 2 (0-5 m; d = 0.49; 80%) and set 3 (5-10 m; d = 0.50; 93%). The results for BST were less conclusive with qualitative indications (moderate effect sizes and both "likely" detrimental and beneficial effects) for slower BST with RSA-SS over the first 5 m of the sprint in set 2 but faster BST over 5-10 m in the same set.
Change of Direction Speed
Mean and total sprint times (MST and TST).
The results for CODS are presented in Table 3. There was no significant difference in MST or TST within or between treatments. Although there was a consistent pattern for slower sprint times in sets 2 and 3 for CODS-SS compared with CODS-CON, this was not supported by statistical or qualitative analysis.
First and best sprint times (FST and BST).
There was no difference between sets 1, 2, and 3 for either FST or BST during the CODS-CON trial. In contrast, both FST and BST were significantly slower in sets 2 and 3 compared with set 1 in CODS-SS (FST, P = 0.002 and 0.007, respectively; BST, P = 0.026 and 0.015, respectively). Between trials, FST was consistently slower for CODS-SS compared with CODS-CON in sets 2 and 3; however, this was not supported by statistical or qualitative analysis. For BST, CODS time for set 3 in the SS condition was significantly slower in comparison to CON (P = 0.035), but no qualitative analyses supported this result.
Participant responses recorded after the final testing session revealed that 50% of participants preferred to stretch during recovery periods, whereas the remaining 50% preferred not to stretch. The effect of static stretching on the performance (TST of sets 2 and 3 combined) of each individual participant depending on their preference is shown in Figure 2. Of the six participants preferring to stretch during recovery periods, four performed better in the RSA test (faster TST of sets 2 and 3) when they completed the control (no-stretching) protocol. With respect to CODS performance, of the six participants preferring to stretch, three performed better in the stretching condition whereas the other 3 individuals performed worse.
The purpose of this study was to examine the effects of static stretching during the recovery periods in field-based team sports (i.e., during interchange or breaks in play) on subsequent RSA and CODS performance, two fundamental determinants of team-sport success. In doing so, ecological validity was emphasized by using stretching and testing protocols that reflect actual practice. It was hypothesized that both RSA and CODS performance would be impaired (particularly in the first 5-m acceleration phase for RSA) when static stretching was conducted between testing sets. This hypothesis was based on existing literature which suggests that the performance of tasks that are highly reliant on muscular force and power is attenuated after acute bouts of static stretching (14,19,24,26).
With respect to RSA, there was a consistent tendency for sprint times to be slower after the static stretching intervention. Although this was not supported by statistical significance across all measures (only MST for 0-5 m set 2; MST for 0-20 m set 2; TST set 2 and BST set 3; and RSA-SS significantly slower than RSA-CON), it is interesting to note that in all the performance data collected, only the 5- to 10-m split times for FST and BST in set 2 were faster after the static stretching intervention. In all other measures, static stretching resulted in either equal or, more commonly, slower sprint times. This tendency for slower sprint times after static stretching was also supported by moderate to large effect sizes and qualitative analyses that indicated "likely" detrimental effects for several measures (see Tables 1 and 2). Taken together, these results suggest that it may be preferable to avoid static stretching during recovery periods for optimal RSA performance.
Likewise, there was a consistent tendency for sprint times to be slower in the CODS-SS trial compared with the CODS-CON, with all performance measures being slower in CODS-SS, except for BST set 2, which resulted in equal times. Further, results for BST (set 3) were significantly slower after static stretching. However, unlike the results for RSA, there were no moderate or strong effect sizes or "likely" detrimental effects associated with CODS-SS.
The lack of statistically significant differences between trials across all measures may have been limited by the testing protocols used in the present study. In both tests (RSA and CODS), participants were required to perform six repeated maximal efforts going every 25 s with active (jog back) recovery in between. Rosenbaum and Henning (20) demonstrated that the addition of further dynamic activities, poststatic stretching, may reduce any stretching-induced impairment in performance. This premise is supported by Little and Williams (12), who suggested that extra muscle activity after stretching may reverse any decrease in muscle compliance and neural drive associated with static stretching. Based on this, it is possible that the testing protocols themselves may have been sufficient dynamic activity to dampen any significant detrimental effect of stretching on both RSA and CODS performance. In addition, it is possible that as fatigue developed during the repeated-sprint protocols used in this study (as shown by the significantly slower sprint times in set 3 compared with set 1 within trials), any effects of static stretching may have been attenuated. This may assist in explaining the significantly slower MST noted during the RSA-SS trial for set 2 but not for set 3.
The hypothesis that any deficits in RSA performance would be more evident in the first 5 m of the sprints is supported to some extent by the present results. MST for the first 5 m in set 2 was significantly slower in RSA-SS compared with RSA-CON, whereas no difference was noted for the 5- to 10-m and the 10- to 20-m splits. Although no statistical difference was noted over the first 5 m for set 3, the qualitative analysis did reveal a moderate effect size and a "likely" detrimental effect associated with RSA-SS. Likewise, moderate effect sizes and "likely" detrimental effects were found over the first 5 m for FST and BST in set 2 when stretching was included during the recovery period. These results support the contention that any detrimental effects of static stretching on sprint ability may be largely due to an impairment of the force producing capacity of the muscles of the lower limb during the initial takeoff. However, the reduction in TST as a result of reduced acceleration during the first 5 m may be insignificant when measuring sprint times over distances of 20 m or more. This may be partly responsible for the discrepancy between results from previous investigations that have examined the effect of static stretching on performance tests involving brief, isolated muscular efforts, such as vertical jumps, compared with tests involving multiple efforts over a longer time span such as a sprint.
This is the first study to examine the effects of static stretching during recovery periods on subsequent sprint performance. Previous research has focused on the effects of static stretching as part of the warm-up rather than within the performance. Also, previous research has predominantly examined the effect of static stretching on a single sprint or CODS movement rather than repeated sprint activities. For instance, Fletcher and Jones (8) observed significantly slower 20-m sprint times after passive stretching. Similarly, Nelson et al. (17) reported significantly slower 20-m sprint times after four sets of three passive stretches each lasting 30 s compared with a no-stretching condition. However, such past studies have typically implemented longer stretch durations than commonly prescribed (17), along with static stretching performed in isolation of other warm-up activities and performance tests being conducted immediately after stretching (8).
In contrast to the abovementioned studies suggesting a negative effect of static stretching on single sprint ability, the same is not so for CODS (or agility). The lack of statistical evidence to support an effect of static stretching on CODS performance in the present study is in agreement with the findings of Little and Williams (12), who observed similar agility performance when a warm-up involving static stretching was compared with warm-up excluding stretching. Likewise, Faigenbaum et al. (7) found no significant difference in proagility run times after static stretching. These authors suggested that change of direction performance is less likely to be affected by the design of the warm-up given the high involvement of complex motor-performance skills. Therefore, it is possible that any effects of static stretching on the CODS task in the present study may have been overshadowed by the complexity of the task.
Of interest, half of the participants preferred to stretch during recovery periods between efforts, whereas the remaining half did not. It is possible that there is a psychological influence on performance when individuals are forced to act in opposition to their preference. For instance, those that prefer to stretch may tend to perform worse when they are forced to omit stretching from their normal routine. To explore this issue, each individual participant's TST (sum of sets 2 and 3) was plotted along with their preference; however, no obvious patterns were evident. That is, whether participants preferred to stretch or not did not seem to impact their individual performance.
In conclusion, after a precompetition warm-up similar to that used in field-based team sports, an acute bout (4 min) of static stretching of the lower limbs during recovery periods between efforts may compromise RSA performance but may have less effect on CODS performance. Given the consistent tendency for slower sprint times for both RSA and CODS after static stretching during recovery periods, it may be preferable to avoid static stretching during recovery periods for optimal performance in these aspects of a team game. However, it is important to note that team sports involve several complex tasks and other athletic abilities, and consequently static stretching is unlikely to have a significant negative impact on the final outcome of a game.
The results of this study do not constitute endorsement by ACSM.
1. American College of Sports Medicine. ACSM's Resource Manual for Guidelines for Exercise Testing and Prescription
. 5th ed. Baltimore (MD): Lippincott Williams and Wilkins; 2006. p. 363.
2. Batterham AM, Hopkins WG. Making meaningful inferences about magnitudes. Sportscience
3. Bishop D. Warm up II: performance changes following active warm up and how to structure the warm up. Sports Med
4. Bradley PS, Olsen PD, Portas MD. The effect of static, ballistic, and proprioceptive neuromuscular facilitation stretching on vertical jump performance. J Strength Cond Res
5. Dawson B, Hopkinson R, Appleby B, Stewart G, Roberts C. Player movement patterns and game activities in the Australian Football League. J Sci Med Sport
6. Evetovich TK, Nauman NJ, Conley DS, Todd JB. Effect of static stretching of the biceps brachii on torque, electromyography, and mechanomyography during concentric isokinetic muscle actions. J Strength Cond Res
7. Faigenbaum AD, Bellucci M, Bernieri A, Bakker B, Hoorens K. Acute effects of different warm-up protocols on fitness performance in children. J Strength Cond Res
8. Fletcher IM, Jones B. The effect of different warm-up stretch protocols on 20 meter sprint performance in trained rugby union players. J Strength Cond Res
9. Fowles JR, Sale DG, MacDougall JD. Reduced strength after passive stretch of the human plantar flexors. J Appl Physiol
10. Keogh JW, Weber CL, Dalton CT. Evaluation of anthropometric, physiological, and skill-related tests for talent identification in female field hockey. Can J Appl Physiol
11. Kokkonen J, Nelson AG, Cornwell A. Acute muscle stretching inhibits maximal strength performance. Res Q Exerc Sport
12. Little T, Williams AG. Effects of differential stretching protocols during warm-ups on high-speed motor capacities in professional soccer players. J Strength Cond Res
13. Magnusson SP, Renstrom P. The European college of sports science position statement: the role of stretching in sports. Eur J Sports Sci
14. Marek SM, Cramer JT, Fincher AL, et al. Acute effects of static and proprioceptive neuromuscular facilitation stretching on muscle strength and power output. J Athl Train
15. McMillian DJ, Moore JH, Hatler BS, Taylor DC. Dynamic vs. static-stretching warm up: the effect on power and agility performance. J Strength Cond Res
16. Mero A, Komi PV, Gregor RJ. Biomechanics of sprint running. A review. Sports Med
17. Nelson AG, Driscoll NM, Landin DK, Young MA, Schexnayder IC. Acute effects of passive muscle stretching on sprint performance. J Sports Sci
18. Paton CD, Hopkins WG, Vollebregt L. Little effect of caffeine ingestion on repeated sprints in team-sport athletes. Med Sci Sports Exerc
19. Power K, Behm D, Cahill F, Carroll M, Young W. An acute bout of static stretching: effects on force and jumping performance. Med Sci Sports Exerc
20. Rosenbaum D, Hennig EM. The influence of stretching and warm-up exercises on Achilles tendon reflex activity. J Sports Sci
21. Snowling NJ, Hopkins WG. Effects of different modes of exercise training on glucose control and risk factors for complications in type 2 diabetic patients: a meta-analysis. Diabetes Care
22. 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
23. Witvrouw E, Mahieu N, Danneels L, McNair P. Stretching and injury prevention: an obscure relationship. Sports Med
24. Yamaguchi T, Ishii K, Yamanaka M, Yasuda K. Acute effect of static stretching on power output during concentric dynamic constant external resistance leg extension. J Strength Cond Res
25. Young W, Elliott S. Acute effects of static stretching, proprioceptive neuromuscular facilitation stretching, and maximum voluntary contractions on explosive force production and jumping performance. Res Q Exerc Sport
26. Young WB, Behm DG. Effects of running, static stretching and practice jumps on explosive force production and jumping performance. J Sports Med Phys Fit