INTRODUCTION

Common clinical practices suggest that pre-exercisestretching can enhance performance and prevent injuriesby increasing flexibility. However, current scientificresearch does not support this notion.1–3 Rather, it wouldappear that the acute effects of stretching can have detri-mental effects on performance parameters such as musclestrength,2,4,5 and jumping performance.6,7 In this paper, thepossible mechanisms of stretching are reviewed in orderto provide guidelines regarding use of stretching as anappropriate strategy to enhance performance and reducethe risk of injury. The effects of stretching on performanceand injury prevention are presented, and further areas forresearch are also recommended.

Several reviews of stretching have been publishedrecently, so the reader is encouraged to refer to these

for more information in selected areas; e.g. the effectsof stretching on muscle properties,8 the effects ofstretching on injury prevention,9,10 and the effects ofstretching on performance.11 The current reviewfocuses on the possible mechanisms of stretching onbiomechanical and neurological changes of muscles,and consequently how these mechanisms affect theperformance and the risk of injury from exercise. Thecurrent paper also focuses specifically on dynamicstretching techniques and dynamic flexibility, whichmay be more beneficial to athletes than the more tra-ditional stretching in terms of performance andinjury prevention.

Literature for this review was located using threeelectronic databases (PubMed, SPORT Discus, andProQuest 5000 International) in addition to manualjournal searches. The computer databases provided

© W. S. Maney & Son Ltd 2004 DOI 10.1179/108331904225007078

STRETCHING: MECHANISMS AND BENEFITS FOR SPORTPERFORMANCE AND INJURY PREVENTION

PORNRATSHANEE WEERAPONG*, PATRIA A. HUME* AND GREGORY S. KOLT†

*New Zealand Institute of Sport and Recreation Research, Division of Sport and Recreation, Faculty of Health and Environmental Studies, Auckland University of Technology, Auckland, New Zealand

†Faculty of Health and Environmental Studies, Auckland University of Technology, Auckland, New Zealand

ABSTRACT

Stretching is usually performed before exercise in an attempt to enhance performance andreduce the risk of injury. Most stretching techniques (static, ballistic, and proprioceptiveneuromuscular facilitation) are effective in increasing static flexibility as measured by jointrange of motion, but the results for dynamic flexibility as measured by active and passivestiffness, are inconclusive. The mechanisms of various stretching techniques in terms ofbiomechanics and neurology, the effectiveness of the combination of stretching with othertherapies such as heat and cold, and the effectiveness of stretching for performance and injuryprevention are reviewed. The possible mechanisms responsible for the detrimental effects ofstretching on performance and the minimal effects on injury prevention are considered, withthe emphasis on muscle dynamic flexibility. Further research is recommended to explore themechanisms and effects of alternative stretching techniques on dynamic flexibility, musclesoreness, sport performance, and rate of injury.

Keywords: Stretching, sport performance, injury prevention

Physical Therapy Reviews 2004; 9: 189–206

access to biomedical and sport-oriented journals, ser-ial publications, books, theses, conference papers, andrelated research published since 1965. The key searchterms included: sport stretching, static stretching,dynamic stretching, ballistic stretching, propriocep-tive neuromuscular facilitation, performance, sportinjury, delayed onset muscle soreness, injury preven-tion, and muscle stiffness. There were a limited num-ber of published randomised controlled trials;therefore, other types of publication such as literaturereviews were included in this review. Articles not pub-lished in English and/or in scientific journals, as wellas those that focused on the psychological effects ofstretching, or the effects of stretching in special popu-lations were not included in this review. The criteriafor inclusion were that the article must have:1. Focused on normal, healthy participants. Age, gender,

and fitness differences were not excluding factors.2. Investigated the acute effects of stretching.

Immediate and long-term effects of flexibilitytraining were not excluding factors.

3. Discussed the possible mechanisms of stretchingin relation to biomechanical and/orneuromuscular properties of muscle, sportperformance, rate of injury, or muscle soreness.

DEFINITION OF STRETCHING

Several literature reviews have considered flexibility12,13 asthe outcome of stretching exercise. However, stretchingitself has not as yet been adequately defined. Magnussonet al.14 stated that ‘stretching has been characterised inbiomechanical terms in which the muscle–tendon unit isconsidered to respond viscoelasticity during the stretch-ing manoeuvre’. However, this definition of stretching is

more a biomechanical result of stretching rather than adefinition of the action of stretching. In our review,stretching is defined as movement applied by an externaland/or internal force in order to increase muscle flexibilityand/or joint range of motion. The aim of stretchingbefore exercise is to increase muscle–tendon unit length15

and flexibility; the increase in flexibility may help toenhance athletic performance and decrease the risk ofinjury from exercise.13

Types of stretching technique

There are various types of stretching techniques that areused, often depending on athlete choice, training pro-gramme, and the type of sport. An earlier review ofstretching12 indicated that four different methods are com-monly used for sport activities – static, ballistic, proprio-ceptive neuromuscular facilitation (PNF), and dynamic(see Table 1).

MECHANISMS OF STRETCHING

Stretching results in elongation of muscles and soft tis-sues through mechanical and neurological mechanisms.Stretching activities may benefit athletes mentallythrough psychological mechanisms; however, there havebeen no detailed studies on the psychological effects ofstretching.

Biomechanical mechanisms

Muscle-tendon units can be lengthened in two ways –muscle contraction and passive stretching. When muscle

190 WEERAPONG ET AL.

Table 1. Stretching techniques: summary of advantages and disadvantages

Technique Ballistic stretchingDefinition Repetitive bouncing movements at the end of joint range of motion12

Advantages Increased range of motion12

Disadvantages Reduced muscle strength;16 may cause injury17

Technique Proprioceptive neuromuscular facilitation (PNF) stretchingDefinition Reflex activation and inhibition of agonist and antagonist muscles18

Advantages Increased range of motion19

Disadvantages Reduced jump height;20 need experience and practice21

Technique Static stretchingDefinition Passive movement of a muscle to maximum range of motion and then holding it for an extended period12

Advantages Increased range of motion;22 simple techniqueDisadvantages Reduced muscle strength;1,2 may cause injury17

Technique Dynamic stretchingDefinition Slow movement of a joint as a result of antagonist muscle contraction throughout the range of movement12

Advantages UnknownDisadvantages Unknown

contracts, the contractile elements are shortened, and acompensatory lengthening occurs at the passive ele-ments of tissues (tendon, perimysium, epimysium, andendomysium).23 When muscle is lengthening, the musclefibres and connective tissues are elongated because ofthe application of external force.23 Stretching increasesmuscle–tendon unit length by affecting the biomechani-cal properties of muscle (range of motion and viscoelas-tic properties of the muscle–tendon unit).

Range of motion

The majority of previous research on the effects ofstretching on flexibility used range of motion as anindicator.19,22,24–26 The exact physiological mechanismof stretching resulting in increased range of motionstill remains unclear27 as most research has failed toshow changes in muscle properties such as pas-sive8,28–32 or active33,34 stiffness. The increase in rangeof motion is thought to be the influence of increasingstretch tolerance8,22,28,29,35 and pain threshold, or sub-ject bias following an intervention.36 Therefore, theincrease in static flexibility, as indicated by range ofmotion, does not provide clear information on mus-culotendinous behaviour.8,28,29

Viscoelastic properties of the muscle–tendon unit

The viscoelastic properties of muscle result in several phe-nomena when external load is applied. When tissues areheld at a constant length, the force at that length graduallydeclines and is described as the ‘stress relaxation’response.8,28,29,36–38 When tissues are held at a constant force,the tissue deformation continues until approaching a newlength and is termed ‘creep’.15 Creep might represent oneexplanation for the immediate increased range of motionafter static stretching.35 The musculotendinous unit alsoproduces a variation in the load–deformation relationshipbetween loading and unloading curves.15 The area betweenthe loading and unloading curves is termed ‘hysteresis’and represents the energy loss as heat due to internaldamping.15,39 A number of researchers have studied theeffects of stretching on stress–relaxation, creep, and hys-teresis;8,15,22,23,28,29,36,38–42 however, none of the previousresearch has clearly demonstrated the relationship of thesephenomena to the rate of muscle injury or performance.

Tendon, which is the major resistance to the finalrange of motion of the musculotendinous unit, showssimilar characteristics. The mechanics of tendon havebeen recently reviewed.43–45 Passive stiffness refers to thepassive resistance of the muscle–tendon unit in a relaxedstate when external forces are applied. The slope of theforce and deformation curve at any range of motion isdefined as passive stiffness.8,13 Passive torque, whichoccurs during passive movement, is resistance from sta-ble cross-links between actin and myosin, non-contrac-tile proteins of the endosarcomeric (titin) and

exosarcomeric cytoskeletons (desmin), and connec-tive tissues surrounding muscles (endomysium, per-imysium, and epimysium).46 Perimysium is consideredto produce major resistance.28,29 Active stiffness isdefined as the resistance of the contracted muscle totransiently deform when external forces were appliedbriefly, and can be measured by the damped oscilla-tion technique.34,47,48 The oscillation of the contractedmuscle after the application of external force resultsfrom the viscoelasticity of muscle and the level ofmuscle activation.34 Passive and active stiffness pro-vide more information on muscle–tendon unit behav-iour during movement than range of motion alone.

Neurological mechanisms

Biomechanical responses of muscle–tendon units duringstretching are independent of reflex activity15,28,31,37,49 asindicated by the lack of muscle activity (EMG)responses during stretching. However, a decrease in theHoffman reflex response (H-reflex) during50 and afterstretching51–54 has been reported. Some research reportshave found that all stretching techniques affect neuralresponses by reducing neural sensitivity.50–55 The major-ity of research on the effects of stretching on neurologi-cal mechanisms have investigated the changes of theH-reflex – the electrical analogue of the stretch reflex butwithout the effects of gamma motoneurons and musclespindle discharge.56 Electrical stimulation of a mixedperipheral nerve (both sensory and motor axons)56 willevoke the H-reflex. The activation of the motor axonsdirectly induces the M-wave (from the point of stimula-tion to the neuromuscular junction) prior to evoking theH-reflex (from Ia afferents arising from annulospiralendings on the muscle spindle) via a monosynaptic con-nection to the alpha motoneurons.56 H-reflex is widelyused to study changes in the reflex excitability of a groupof muscle fibres.50,53,54,56–59 The depressed amplitude of H-reflex after stretching might be due to several possibili-ties relating to presynaptic and/or postsynaptic change;for example, presynaptic inhibition inducing an auto-genic decrease in Ia afferents and/or an altered capacityfor synaptic transmission during repetitive activation.60

Alternatively, the postsynaptic changes might be due toan autogenic inhibition from the Golgi tendon organ(GTO), recurrent inhibition from the Renshaw loop, orpostsynaptic inhibition of afferents from joint and cuta-neous receptors.60

Mechanisms of each stretching technique

Even though each technique of stretching is expectedto increase muscle and joint flexibility, different

THE MECHANISMS AND BENEFITS OF STRETCHING 191

192 WEERAPONG ET AL.

Table 2. The effects of static stretching on muscle properties: overview of studies

ROM and passive stiffnessReference McHugh et al.37

Trial design PPTSample 9 men and 6 women (hamstrings)Interventions Static stretch (hold 45 s). (i) At onset of EMG; (ii) 5° below the onset of EMG (negligible EMG activity)Outcome measures (i) Peak torque; (ii) ROM; (iii) EMGMain results S: ↓ torque; ↑ ROM

Reference Magnusson et al.36

Trial design PPTSample 10 men (hamstrings)Interventions (i) Static stretch (hold 90 s, rest 30 s) 5 times (stretch 1–5); (ii) repeated static stretch one time (stretch 6)Outcome measures (i) Peak torque; (ii) ROM; (iii) EMGMain results S: ↓ stress relaxation; ↑ ROM

Reference Magnusson et al.28

Trial design PPTSample 7 women (one leg stretch one leg control) (hamstrings)Interventions Static stretch (45 s hold x 15–30 s rest x 5 times), twice daily, 20 consecutive daysOutcome measures (i) Stress relaxation; (ii) energy; (iii) EMG; (iv) ROMMain results S: ↑ ROM

Reference Halbertsma et al.22

Trial design RCTSample 10 men and 6 women with short hamstringsInterventions (i) Static stretching (30 s hold x 30 s rest) for 10 min (n = 10); (ii) control-rest (n = 6)Outcome measures (i) Peak torque; (ii) ROM; (iii) passive stiffnessMain results S: ↑ ROM

Reference Magnusson et al.29

Trial design CCTSample 8 neurological intact and 6 spinal cord injury volunteers (hamstrings)Interventions Static stretch (hold 90 s)Outcome measures (i) Stress relaxation; (ii) passive torque; (iii) EMGMain results NS

Reference Magnusson et al.14

Trial design PPTSample 13 men (hamstrings)Interventions 5 static stretches (hold 90 s, rest 30 s) and repeated 1 h laterOutcome measures (i) Stiffness; (ii) energy; (iii) passive torqueMain results S: ↓ energy, stiffness, and peak torque

Reference Klinge et al.32

Trial design CCTSample 12 men in experimental group, 10 men in control groupInterventions 4 x 45 s static stretchOutcome measures (i) ROM; (ii) passive stiffnessMain results NS

Reference McHugh et al.31

Trial design CCTSample 8 men and 8 women (hamstrings)Interventions SLR stretchOutcome measures (i) Peak torque; (ii) ROM; (iii) EMGMain results S: ↑ ROM

Reference Magnusson et al.8

Trial design CCTSample 12 men (hamstrings)Interventions (i) 90 s static stretches; (ii) continuous movements 10 times at 20°/sOutcome measures (i) ROM; (ii) passive stiffnessMain results S: ↑ ROM

Reference Muir et al.27

Trial design RCTSample 10 men (one leg-stretching, one leg-control) (hamstrings)Interventions (i) static stretching (30 s x 10 s) for 4 times; (ii) control-restOutcome measures (i) Peak torque; (ii) centre range (of hysteric loop)Main results NS

THE MECHANISMS AND BENEFITS OF STRETCHING 193

Reference McNair et al.38

Trial design CBTSample 15 men and 8 women (plantar flexors)Interventions Static stretching. (i) 1 x 60 s hold; (ii) 2 x 30 s hold; (iii) 4 x 15 s hold; (iv) continuous passive movement for 60 sOutcome measures (i) Passive stiffness; (ii) peak torque.Main results Continuous movement – S: ↓ passive stiffness. Hold condition – S: ↓ peak tension

Reference Magnusson et al.42

Trial design PPTSample 20 menInterventions 3 static stretches (hold 45 s, rest 30 s) and repeated 1 h laterOutcome measures (i) Stiffness; (ii) energy; (iii) passive torqueMain results S: ↓ stress relaxation

Reference Kubo et al.39

Trial design CBTSample 7 men (plantar flexors)Interventions Passive stretching to 35° dorsiflexion at 5°/s for 10 minOutcome measures (i) Tendon stiffness; (ii) tendon hysteresis; (iii) MVCMain results S: ↓ tendon stiffness (10%), tendon hysteresis (34%)

Reference Kubo et al.41

Trial design CBTSample 8 men (plantar flexors)Interventions Passive stretching to 35° dorsiflexion at 5°/s for 5 minOutcome measures (i) Tendon stiffness; (ii) tendon hysteresisMain results S: ↓ tendon stiffness (8%), tendon hysteresis (29%)

ROM and active stiffnessReference Wilson et al.47

Trial design CCTSample 16 male weight-lifters (n = 9 in experiment, n = 7 in control group)Interventions Flexibility training (6–9 repeats) of upper extremities, 10–15 min per session, twice a week for 8 weeksOutcome measures (i) Rebound bench press (RBP); (ii) purely concentric bench press (PCBP)Main results S: ↑ ROM (13%). S: ↑ RBP (5.4%). S: ↓ SEC stiffness (7.2%)

Reference McNair & Stanley34

Trial design CCTSample 12 men and 12 women (plantar flexors)Interventions (i) Static stretch (30 s x 30 s); (ii) jogging (60% MHR); (iii) combined (ii) + (i).

Randomly order, each intervention for 10 minOutcome measures (i) ROM; (ii) active stiffnessMain results Jogging group – S: ↓ active stiffness. All groups – S: ↑ ROM

Reference Cornwell et al.33

Trial design PPTSample 10 men (plantar flexors)Interventions Passive stretchingOutcome measures (i) ROM; (ii) active stiffnessMain results S : ↑ ROM

Reference Hunter & Marshall62

Trial design CCTSample 15 men, 15 women (n =15 in experiment and control groups) (plantar flexors)Interventions 10 x 30 s static stretchesOutcome measures Active stiffnessMain results NS

Reference Cornwell et al.1

Trial design PPTSample 10 men (plantar flexors)Interventions Passive stretching (30 s x 6 times)Outcome measures (i) Active muscle stiffness; (ii) EMG; (iii) jump heightMain results S: ↓ jump height (7.4%). ↓ active stiffness (2.8%)

CCT, controlled clinical trial; RCT, randomised controlled trial; PPT, pre- and post-test trial; CBT, counterbalance trial; ROM, range ofmotion; EMG, electromyography; SEC, series elastic components; S, significant; NS, non-significant.

stretching techniques produce increases in flexibilityby different mechanisms.

Static stretching

Static stretching is the most widely used technique by ath-letes due to its simplicity. Static stretching has been foundto affect both mechanical1,28,29,31,36–39,41 and neurologi-cal50–52,60,61 properties of the muscle–tendon unit resultingin increased musculoskeletal flexibility (see Table 2).

Despite static stretching being effective in increas-ing static flexibility as measured by range ofmotion,22,28,29,34,38 it does not affect dynamic flexibilityas measured by passive8,22,38 or active33,34 stiffness, butaffects viscoelastic properties by reducing stress relax-ation.22,27–29,36,38 The reduction of stress relaxation isan acute adaptation of the parallel elastic componentto lower the imposed load across the myotendinousjunction where injury usually occurs.28,29 However,there is no clear evidence that static stretching canreduce the rate of injury. Static stretching has beenreported to produce similar muscle–tendon unit prop-erty responses between neurologically intact partici-pants and those with spinal cord injuries withcomplete motor loss.29 Furthermore, there have beena number of reports of no EMG activity during pas-sive stretching.28,29,31,36,37 Therefore, the effects of staticstretching on muscle properties do not involve theneurological mechanism.

The effects of stretching on muscle properties dependon various factors including the stretching techniquesused, time to stretch, holding duration, time to rest, andthe time gap between intervention and measurement.The majority of research has examined the acute effectsof static stretching on passive properties of the mus-cle–tendon unit.8,28,29,36,38,42 In a series of studies byMagnusson et al.,28,29,36,63 static stretching at 90 s for fiverepetitions reduced muscle resistance measured by pas-sive stiffness, peak torque, and stress relaxation. Thedecline of muscle–tendon unit resistance returned tobaseline within 1 h28,29 except for stress relaxation.36

Unfortunately, shorter stretch holding times (less than60 s), and lower stretching repetitions (less than fourtimes) did not provide such effects.22,38,42 Interestingly,long-term training using ten stretches for 45 s per day(3 weeks)28 and four stretches for 45 s, two sessions perday, 7 days per week for 13 weeks32 did not change themechanical or viscoelastic properties of muscle.28

Therefore, the changes in viscoelasticity of muscle–ten-don units depend more on the duration of stretch ratherthan the number of stretches39 or the length of thestretching training period. Indeed, if a decrease in mus-cle–tendon unit resistance is required long stretch dura-tion (up to 90 s) might be most beneficial.

The protocol used in the study by Magnusson et al.8

showed an acceptable reproducibility (correlation

coefficient; r = 0.91–0.99, 0.92–0.95 and coefficient ofvariation; 5.8–14.5%, 9.5–17.6%) for test and retestwithin 1 h and between days, respectively. The key tosuch high reproducibility was the participants’ trunkbeing fixed perpendicular to the seat used in the experi-ment in order to fix the origin of the hamstring musclesat the pelvis. The rate of stretch was controlled at anangular velocity of 5°/s which does not stimulate muscleactivity. This rate of stretch might be an appropriateprotocol to study passive properties of muscle at rest.During sports activities or injury, however, muscle–ten-don units are stretched at 25–50 times more than thevelocity used in these studies (e.g. 120°/s)39 It was foundthat the tensile strength and energy absorption changedwith different rates in vitro.15 Therefore, the high veloc-ity of stretch as used during sport performance isneeded to be researched to investigate the change of vis-coelastic properties of muscle.

Prolonged static stretch (5–10 min) has been shownto decrease tendon and aponeurosis stiffness (elastic-ity) and hysteresis (viscosity) as measured passively byultrasonography.39,41 The decrease in stiffness fromstretching may be due to an acute change in thearrangement of collagen fibres in tendon.39 However,the holding time in the Kubo et al.39,41 studies is con-sidered very long when compared with those rou-tinely used in stretching (30–60 s hold).Unfortunately, no research on the effects of staticstretching on the tendon for shorter durations hasbeen identified in the published literature. The samegroup of researchers41 also reported that the combina-tion of long-term resistance (70% of one repetitionmaximum, 10 repetitions per set, five sets per day, 4days per week for 8 weeks) and stretching (10 min perday, 7 days per week for 8 weeks) training did notchange tendon elasticity (determined by stiffness) butreduced hysteresis (17%). Unfortunately, the mecha-nisms responsible for the decrease in stiffness and hys-teresis of the tendon are still unknown. The acuteresponse of stretching on tendon and aponeurosisstiffness might be partly responsible for the increasein range of motion. The effects of stretching on serieselastic components are also unclear. The findings ofseveral researchers are in agreement that staticstretching does not affect active stiffness. McNair andStanley34 and Hunter et al.62 each found that soleusstretching (five stretches of 30 s hold and 30 s rest,and 10 stretches of 30 s hold, respectively) did notreduce active stiffness. Similarly, Cornwell andNelson33 reported that stretching did not affect theactive stiffness of musculotendinous units of tricepssurae. Unfortunately, Cornwell and Nelson33 did notreport key details of the research protocol such as thestretching technique, nor the active stiffness measure-ment. Recently, Cornwell et al.1 found a slight, but

194 WEERAPONG ET AL.

significant, reduction of active muscle stiffness of thesoleus muscle (2.8%) after stretching (six stretches of30 s hold and 30 s rest). Whereas McNair andStanley34 and Hunter et al.62 only investigated thesoleus muscle, Cornwell et al. investigated the wholetriceps surae. Total stretching time in the study ofCornwell et al. was comparable to that in McNair andStanley’s study (180 s and 150 s, respectively), but wasmuch shorter than that used by Hunter et al. (300 s).The method of stretching also varied. In the McNairand Stanley and Hunter et al. studies, participantsstretched the plantar flexors by adopting a step-standing position and stretching by flexing bothknees. The participants in the Cornwell et al. studystretched in two ways: in the first case, participantsplaced the foot on an inclined board and maximallydorsiflexed the ankle joint whilst keeping the sole ofthe foot flush with the board surface and the knee

joint fully extended. The second stretching techniqueemployed the same protocol but the knee was flexedin order to increase the stretching force on the soleus.These stretching protocols might provide more strainon series elastic components of muscle than themethod used in the McNair and Stanley study.

In evaluating the effects of long-term stretching,Wilson et al.47 reported that flexibility training of pec-toralis and deltoid muscles (10–15 min per session,twice per week for 8 weeks) reduced active musclestiffness by 7.2%. The decrease in active stiffnessmight be from the long-term adaptation of connectivetissue, sarcomere, contractile tissue, and/or reflexresponses. These effects may not occur with acutestretching.

Despite the fact that it is accepted that a neurologi-cal mechanism is not responsible for the change ofmuscle–tendon unit properties during passive stretching,

THE MECHANISMS AND BENEFITS OF STRETCHING 195

Table 3. The effects of static stretching on neuromuscular activity: overview of studies

Reference Thigpen et al.55

Trial design CCTSample 6 men, 2 womenInterventions 3 x 20 s toe touch stretchingOutcome measures H-reflexMain results S: ↓ H/M ratio 21.49%

Reference Vujnovich & Dawson50

Trial design CCTSample Group A (n = 14): static stretching. Group B (n = 2): similar to A but followed up every 2 min for 10 min.

Group C (n = 5): similar to B but followed by ballistic stretch (1 rad/s for 160 s). Group D (n = 2): stretching at midway between neutral and fully dorsiflexion

Interventions Maximally dorsiflexion for 160 sOutcome measures H-reflexMain results Group A: S ↓ H-reflex (45%). Group B: S ↓ H-reflex during stretching but NS afterwards. Group C: S ↓ H-

reflex during ballistic stretching (84%) greater than static stretching (40%). Group D: S ↓ H-reflex 40%

Reference Rosenbaum et al.52

Trial design CCTSample 50 male athletesInterventions 3 min static stretch: hold 30 sOutcome measures H-reflex of triceps surae: (i) peak force; (ii) force rise rate; (iii) half relaxation rate; (iv) EMG amplitude and

integral; (v) EMG latencies; (vi) impulsesMain results S: ↓ peak force, force rise rate, half relaxation rate, EMG amplitude and integral. S: ↑ EMG latencies

Reference Avela et al.51

Trial design CCTSample 6 men (plantar flexors)Interventions Repeated passive stretching (1 h)Outcome measures (i) MVC; (ii) 50% MVC; (iv) Hmax; (v) motor unit firing rate (ZCR)Main results S: ↓ MVC (23.2 ± 19.7%). S: ↓ H reflex (46.1 ± 38.3%). S: ↓ ZCR (12.2 ± 11.4%)

Reference Guissard et al.60

Trial design PPTSample 7 men, 4 womenInterventions Static stretch at 10° and 20°Outcome measures H/M ratioMain results Dorsiflexion (10°) – S: ↓ H/M ratio (25%). Dorsiflexion (20°) – S: ↓ H/M ratio (55%)

CCT, controlled clinical trial; RCT, randomised controlled trial; PPT, pre- and post-test trial; CBT, counterbalance trial; ROM, rangeof motion; EMG, electromyography; SEC, series elastic components; S, significant; NS, non-significant; MVC, maximum voluntarycontraction; H/M ratio, H-reflex/M-wave ratio.

static stretching has been reported to decrease neuro-muscular sensitivity as indicated by H-reflexresponses.50,51,60 Rosenbaum and Henning52 reportedthat a static stretch of triceps surae (30 s each for threetimes) reduced the peak force of reflex force produc-tion, force rise rate, and EMG activity. Stretchingmight improve muscle compliance (reduced peakforce and force rise rate), reduce muscle spindle sensi-tivity (reduced peak-to-peak amplitude), and reduceexcitation–contraction coupling (increased force-to-EMG ratios). Thigpen et al.55 proposed that thedecrease of evoked H-wave amplitude might be due toinhibitory effects of the Ib afferent from the Golgitendon organ (GTO). Avela et al.51 also reported areduction of H-reflex (46%) after prolonged stretch-ing for 1 h. The reduction of stretch reflex activity(reduced peak-to-peak amplitude) and α-motoneu-ron pool excitability (reduced H-wave/M-wave ratio)were suggested to result from reduced sensitivity ofthe large-diameter afferents. Vujnovich and Dawson50

compared two stretching techniques (static and ballis-tic stretching) on the excitability of the α-motoneu-rons as indicated by amplitude changes in theH-reflex, which was reduced by up to 55% during sta-tic stretching of the soleus muscle (maintained for160 s) but returned to baseline immediately followingthe termination of stretching. The amplitude ofstretching (mid- and full-range of motion) wasreflected in the mean depression of the H-reflexamplitude. Therefore, the receptor that is likely tomediate the inhibitory effect during static musclestretch is the muscle spindle type II afferent. In con-trast, Guissard et al.60 reported that greater stretchingamplitude produced a greater reduction of H-reflex.When the ankle was moved passively for 10°, the H-reflex reduced by 25% but when the ankle was movedup to 20°, the H-reflex was reduced by 54%. Eachstretch was held for 20–30 s. These results suggestedthat a reduction of motoneuron excitation duringstretching resulted from both pre-and post-synapticmechanisms. The pre-synaptic mechanism wasresponsible at small stretching amplitudes, while post-synaptic mechanisms played a dominant role in largerstretching amplitudes.

A reduction of neuromuscular sensitivity duringstretching, reflected in the change in the amplitude ofthe H-reflex, might be due to the tension on stretchedmuscle applied by an external force being higher thanthe resistance from a protective mechanism of musclefrom a stretch reflex. When muscle is stretched fur-ther, neuromuscular excitability from the stretchreflex still works but cannot resist the external force,which might cause reduced sensitivity of the musclespindles and, consequently, of the H-reflex duringstretching. Decreased neuromuscular sensitivity, as

indicated by the amplitude of the H-reflex, was foundto revert to control levels immediately after termina-tion of static stretching in a study by Vujnovich andDawson.50 A neuromuscular mechanism is, therefore,unlikely to underlie increased muscle flexibility afterstatic stretching.

Ballistic stretching

Ballistic stretching is likely to increase flexibilitythrough a neurological mechanism. The stretchedmuscle is moved passively to the end range by anexternal force or agonist muscle: holding a muscle inthis position might reduce muscle spindle sensitivity,with repeated stretch applied at the end range inhibit-ing the GTO. Research has reported an increase inrange of motion,50,64 decrease in EMG65 and decreaseH-reflex50 with ballistic stretching.

Only one published study50 has examined the effects ofballistic stretching on neuromuscular excitability.Ballistic stretch applied following static stretch demon-strated lowered H-reflex mean amplitude than thatobtained during static stretching.50 The lower H-reflexmight be due to the inhibition of GTO or presynapticinhibition from type Ia afferents. However, these resultsshould be interpreted with caution, as the number ofsubjects in the static stretching and ballistic stretchinggroups were different (n = 14 and 5 respectively) andthere was no control group. Additionally, ballisticstretching was performed immediately after staticstretching, therefore the effects of ballistic stretchingalone on H-reflex are still unknown.

It has been suggested that ballistic stretching may bemore harmful than other stretching techniques.During ballistic stretching, muscle is stretched at afast rate and then rebounded back repetitively, result-ing in greater tension and more absorbed energywithin the muscle–tendon unit.15 Muscle, which isreleased immediately after applying a high force, isnot allowed enough time to reduce tension (stressrelaxation) or increase length (creep).15 Surprisingly,scientific evidence did not support the widely heldview that ballistic stretching is potentially more harm-ful than static stretching. Rather, ballistic stretching(60 bounces per minute, 17 stretches per set for threesets) was found to result in less severe muscle sorenessthan static stretching (the same intensity and dura-tion, but with static stretching held for 60 s) in col-lege-age male volunteers.17 However, despite suchevidence, most researchers still recommend slow staticstretch before exercise.12,21

Proprioceptive neuromuscular facilitation (PNF)

Several PNF techniques have been used to increaseflexibility including slow-reversal-hold, contract-relax, and hold-relax techniques.12 These techniques

196 WEERAPONG ET AL.

include the combination of alternating contraction andrelaxation of both agonist and antagonist muscles.12 Thetheory of PNF has been discussed and reviewedrecently.18 The contractility of muscles provides the flexi-bility in the PNF technique, on the basis of the vis-coelastic properties of muscle and neuromuscularfacilitation. The contracted muscle lengthens the non-contractile elements (perimysium, endomysium, ten-don), and consequently, causes a relaxation of themuscle–tendon unit and decreased passive tension in themuscle.23 The contracted muscle also stimulates sensoryreceptors within the muscle: muscle spindles (negativestretch reflex) and GTOs which help to relax the tensedmuscle; as a consequence, the muscle–tendon-unitbecomes more relaxed after the contraction.

Some PNF techniques, such as slow-reversal-hold,require the agonist muscle to contract in order torelax antagonist muscle.12 ‘Reciprocal inhibition’occurs when the excitability signal from the agonistmuscle is transmitted by one set of neurons in thespinal cord to elicit muscle contraction, whilst aninhibitory signal is transmitted through a separate setof neurons to inhibit the antagonist muscle;18 suchreciprocal inhibition helps all antagonistic pairs ofmuscle to make smooth movement. When the antago-nist muscle is inhibited, it will be stretched in theopposite direction more easily.

Isometric contraction is commonly performedprior to passive stretching as part of the PNF tech-nique. It has been found that isometric contractionproduced a brief decrease in H-reflex response (83%by 1 s and 10% by 10 s, respectively);54 furthermore,the depression of H-reflex has been reported as inde-pendent of the intensity of isometric contraction,66

velocities, and amplitude of stretch.53 Researchershave proposed that the decrease in H-reflex after iso-metric contraction could be a result of pre-synapticinhibition.53,54 The suppression of reflex activity hasbeen found to be short-lasting (less than 10 s), indi-cating that passive stretching should be performedimmediately after pre-isometric contraction in orderto gain the maximal efficiency of stretching.

PNF stretching has been reported to result in agreater improvement of range of motion comparedwith static stretching.28,29,67 Post-isometric contractionreduced neuromuscular sensitivity53,54 might help toenhance the effectiveness of stretching (see Table 4).Toft et al.68 compared short-term (90 min afterstretching) and long-term (3 weeks) effects of con-tract–relax stretching (maximal contraction of plan-tar flexors for 8 s, relaxation for 2 s, and passivestretch for 8 s) on stress relaxation of ankle plantarflexors. No difference was found between short-termand long-term effects of PNF stretching. In studies by

THE MECHANISMS AND BENEFITS OF STRETCHING 197

Table 4. The effects of PNF stretching on biomechanics and neuromuscular activity: overview of studies

Reference Sady et al.67

Trial design CCTSample (i) Control (n = 10); (ii) static (n = 10); (iii) ballistic (n = 11); (iv) PNF (n = 12)Interventions (i) Static: 3 x 6 s; (ii) ballistic: repeated movements for 20 times; (iii) PNF: 3 x 6 s. 3 days/week for 6 weeksOutcome measures ROMMain results PNF and control group. S: ↑ ROM

Reference Toft et al.68

Trial design PPTSample 10 menInterventions Contract-relax (8 s maximum contraction, 2 s relax, 8 s static stretch) 6 timesOutcome measures Stress relaxationMain results NS

Reference Magnusson et al.28

Trial design CCTSample 7 women (one leg stretch one leg control) (hamstrings)Interventions Static stretch (45 s hold x 15–30 s rest x 5 times), twice daily, 20 consecutive daysOutcome measures (i) Stress relaxation; (ii) energy; (iii) EMG; (iv) ROMMain results S: ↑ ROM

Reference Magnusson et al.29

Trial design CCTSample 8 neurological intact and 6 spinal cord injury volunteers (hamstring)Interventions Static stretch (hold 90 s)Outcome measures (i) Stress relaxation; (ii) passive torque; (iii) EMGMain results NS

CCT, controlled clinical trial; PPT, pre- and post-test trial; ROM, range of motion; EMG, electromyography; SEC, series elasticcomponents; S, significant; NS, non-significant.

Toft et al.68 and McNair et al.,38 peak torque of theplantar flexors 60 s after the start of the stress–relax-ation phase was reduced by about 15% by PNF stretch-ing,68 and 20% by static stretching.38 Similarly, peaktorque of the hamstrings was found to have declined by18% after PNF stretching and 21% after static stretch-ing by Magnusson et al.28,29 If dynamic flexibility isrequired, the claim that PNF stretching represents abetter flexibility technique than static stretching is stilldebatable.

PNF stretching is a complicated stretching techniquewith a combination of shortening contraction and pas-sive stretching. Therefore, PNF stretching might poten-tially be harmful as it has been found to increase bloodpressure69 and EMG activity70 during the contractionphases. Moreover, PNF technique needs some experi-ence to be performed, and a partner is needed to helpwith stretching. When compared with other stretchingtechniques, the selection of PNF stretching prior toexercise is still questioned; further study is required toinvestigate the effects of PNF stretching on dynamicmuscle properties (e.g. active and passive stiffness), per-formance, and muscle soreness.

Dynamic stretching

In an extensive review of warm-up and stretching,12

Shellock and Prentice reported that ‘dynamic stretchingis important in athletic performance because it is essen-tial for an extremity to be capable of moving through anon-restricted range of motion’. Unfortunately, therewas no published research cited within the review forany aspects of dynamic stretching.

Cyclic stretching, or passive continuous motion, hasbeen demonstrated to be effective for decreasing passivemuscle stiffness.38 A less stiff muscle is believed to absorbgreater energy when forces are applied to it;38 furthermore,less muscle stiffness might be beneficial in reducing theseverity of muscle soreness following exercise as researchhas shown the positive relation between passive stiffnessand the severity of muscle soreness.71 However, there is nodirect evidence that the dynamic movement of stretchingcan reduce the severity of muscle soreness.

Dynamic stretching might be a useful protocol forincreasing flexibility without decreasing athletic perfor-mance. Dynamic contraction of muscle throughout therange of motion is expected to decrease dynamic flexibil-ity as indicated by passive muscle stiffness.38 Movement,without holding at end range of motion, may not reduceneuromuscular sensitivity. If the effect of decreasing pas-sive stiffness, however, is more pronounced than theeffects of neuromuscular sensitivity on performance, areduction in performance might still occur. Furtherresearch is needed to elucidate the benefits of dynamicstretching on flexibility, muscle properties, neuromuscularsensitivity, performance, and injury prevention.

COMBINATION OF STRETCHING WITHOTHER THERAPIES

Warm-up

Stretching is generally performed after warm-up;12

the commonly accepted theory is that warm-up willincrease muscle temperature to help enhance tissueflexibility.42 Warm-up has been shown to increase tis-sue temperature but not to affect muscle properties(passive energy absorption) during stretching.42 In astudy by Magnusson et al.,42 warm-up (jogging) wasperformed at 70% of maximum O2 uptake for 10 minand resulted in elevation of muscle temperature by3°C. After warm-up, four static stretch manoeuvresof 90 s reduced passive energy absorption by 25%,while five static stretching exercises (held for 90 s) atresting temperature (no warm-up) reduced passiveenergy absorption by 30%.28,29 However, there were nodata on other parameters of passive muscle propertiessuch as passive muscle stiffness and peak torque col-lected in this study. In the same way, warm-up (run-ning72 and heel raising73) did not help to improverange of motion of lower limbs regardless of thewarm-up intensity (60, 70, and 80% of VO2 max)72 ortraining period (2, 4, and 6 weeks).73 Therefore,warm-up might not be an effective way to enhanceflexibility of passive properties of muscle.

McNair and Stanley34 reported the effects of warm-up (treadmill jogging for 10 min at 60% of maximumage-predicted heart rate) on series elastic muscle stiff-ness. Surprisingly, warm-up reduced active (serieselastic component) stiffness (6%) more than the com-bination treatments of warm-up and stretching (3%)and stretching alone (–1%). Warm-up might be moreeffective in reducing resistance from various proper-ties associated with active stiffness (passive jointproperties, level of muscle activation, tendon proper-ties, and the effect of stretch reflex) than stretching.

Warm-up might help to increase muscle relaxationby reducing EMG activity. Mohr et al.49 reported thatpost-warm-up EMG activity was significantly lessthan pre-warm-up EMG activity in gastrocnemiusand soleus muscle. The authors proposed that musclearchitecture and arrangement of connective tissuemight influence the effects of warm-up (in this casecycling) on reducing EMG activity of the gas-trosoleus complex. EMG during stretching in theMohr et al.49 study was slightly higher than previousreports8,30,36,38,42,74 due to this study using needle elec-trodes and participants stretching by themselves(weight-bearing position). Participants in previousresearch were investigated by using surface elec-tromyographic electrodes, and were passivelystretched in a relaxed position. Therefore, warm-up

198 WEERAPONG ET AL.

prior to stretching in order to reduce EMG activitymight be important in some muscles (i.e. gastrocne-mius and soleus) which are dense in connective tissue,and using more challenging stretch positions thatneed more control and balance of the body such as astanding bent knee stretch or a standing straight kneestretch with heel overhanging a step.

Heat and cold

Temperature has effects on muscle75 and connectivetissue76 in vivo. In a study of skeletal muscle tensilebehaviour, warm muscle (40°C, measured using anintramuscular probe) showed less stiffness and moreload-to-failure than cold (35°C) muscle.75 A study ofrat tail tendon indicated that an elevated tissue tem-perature (45°C), before the application of low force(one-quarter of full load-to-failure), produced greaterresidual length and reduced tissue damage (indicatedby tissue rupture which was defined as the point atwhich elongation continued with no increase in load)than tissues with a normal temperature.76 The resultsof these animal studies75,76 support the common prac-tice of applying superficial heat before stretching inorder to maximise the effectiveness of treatment.

In humans, results of research on the combinedeffects of stretching with either heat or cold on flexi-bility as measured by range of motion have beeninconclusive.24,77,78 In the study by Henricson et al.,24

application of an electric heating pad (43°C) for20 min before PNF stretching significantly increasedrange of motion of hip flexion and abduction. Thestretching-only group increased range of motion ofhip flexion and external rotation, while the heat-onlygroup did not show any effect on range of motion.The stretching in this study was performed in onlyone direction (hip flexion) while the participants wereinvestigated for hip flexion, abduction, and externalrotation. Taylor et al.77 compared the effects of staticstretch alone (held for 1 min), heat (77°C) and staticstretch, and cold (–18°C) and static stretch. There wasa significant increase in hamstring length regardlessof treatment but no significant difference betweentreatments. Similarly, the comparison among PNFstretching alone (isometric contraction for 10 s, 5 srest for four times), PNF stretching and cold(immersed in a cold-water bath (8°C) for 10 minbefore performing the same PNF protocol), and PNFand heat (immersed in a hot-water bath [44°C] for10 min before performing the same PNF protocol) forfive consecutive days did not lead to any difference inhip range of motion.78 Knight et al.73 studied theeffects of four treatments (static stretching alone,warm-up prior to static stretching, superficial moist

heat for 15 min prior to static stretching, and 7 min ofcontinuous ultrasound prior to static stretching) onplantar flexor flexibility over six consecutive weeks.The use of ultrasound for 7 min prior to stretchingwas the most effective treatment for increasing ankledorsiflexion range of motion. Ultrasound may haveprovided a deeper heat at the muscular level79 com-pared to the hot pack or hot bath that might onlyincrease skin temperature.

Massage

Apart from stretching and warm-up, massage is oftenperformed in sport practice to help prevent muscleinjury. Rodenburg et al.80 reported that the combina-tion of these treatments could reduce some negativeeffects of muscle soreness induced by eccentric exer-cise. The application of warm-up aimed to decreaseviscosity of muscle tissues and stretching aimed toreduce passive tension. These two treatments wereperformed before eccentric exercise, and massage wasperformed after exercise with the aim of increasingblood flow and reducing waste products. The results,however, were not consistent as the maximal force, theflexion of elbow angle, and the creatine kinase level inblood were reduced, while other soreness parameterssuch as soreness sensation, extension elbow angle,and myoglobin in blood did not change. Therefore,the combination of these treatments did not reducethe severity of muscle soreness any more than anyindividual treatment. In other studies, warm-up hasbeen reported as effective in reducing the severity ofmuscle soreness and functional loss,81 massage waseffective in reducing soreness sensation,82–84 whilestretching did not show any effect at all.3,85,86

THE EFFECTS OF STRETCHING ONPERFORMANCE

Stretching is expected to increase flexibility, and, conse-quently, enhance sport performance.13 The effects ofstretching on several key performance parameters havebeen investigated including muscle strength, power, andendurance, as well as the efficiency of exercise (such asrunning economy). However, recent research2,7,20,87,88

still questions whether these interventions provide anybenefit for performance (see Table 5).

Muscle strength, power, and endurance

The acute effects of static stretching, ballistic stretch-ing, and PNF include reduction of muscle strength as

THE MECHANISMS AND BENEFITS OF STRETCHING 199

200 WEERAPONG ET AL.

Table 5. The effects of stretching on performance: overview of studies

Static stretchingReference Kokkonen et al.5

Trial design CBTSample 15 men and 15 women (hamstrings)Interventions 20 min stretching (5 stretches, 3 times assisted, 3 times unassisted, hold 15 s, rest 15 s)Outcome measures (i) Sit and reach score; (ii) maximum strength (1RM)Main results S: ↑ ROM (16%), ↓ strength (7.3%)

Reference Fowles et al.2

Trial design CBTSample 8 men, 4 womenInterventions 13 x 135 s static stretches, total 30 minOutcome measures (i) MVC; (ii) twitch interpolation with EMG; (iii) twitch characteristics at pre-, immediately post, 5, 15, 30, 45,

60 min post-stretchingMain results S: ↓ MVC (28, 21, 13, 12, 10, and 9% (by the time to collect data)). S: ↓ motor unit activation and EMG after

treatment but recovered by 15 min

Reference Knudson et al.6

Trial design CBTSample 10 men and 10 women (quadriceps, hamstrings, plantar flexors)Interventions 3 x 15 s static stretchOutcome measures (i) Peak velocities; (ii) duration of concentric phase; (iii) duration of eccentric phase; (iv) smallest knee angle;

(v) jump heightMain results NS

Reference Nelson et al.4

Trial design PPTSample 10 men and 5 womenInterventions One active and 3 passive stretching for 15 minOutcome measures Peak torque at 1.05, 1.57, 2.62, 3.67, and 4.71 rad/sMain results S: ↓ strength at 1.05 rad/s (7.2%) and 1.57 rad/s (4.5%)

Reference Behm et al.87

Trial design CBTSample 12 menInterventions Quadriceps stretching (45 s held, 15 s rest, for 5 sets)Outcome measures (i) MVC; (ii) EMG; (iii) evoked torque; (iv) tetanic torqueMain results S: ↓ MVC (12%), muscle inactivation (2.8%), EMG (20%), evoked force (11.7%)

Reference Cornwell et al.1

Trial design CCTSample 10 menInterventions 3 x 30 s static stretchOutcome measures (i) Active muscle stiffness; (ii) EMG; (iii) jump heightMain results S: ↓ jump height (7.4%) ↓ active stiffness (2.8%)

Reference Young et al.7

Trial design CBTSample 13 men, 4 women (quadriceps and plantar flexors)Interventions 2 x 30 s for each muscleOutcome measures (i) Concentric force; (ii) concentric jump height; (iii) concentric rate of force developed; (iv) drop jump heightMain results S: ↓ concentric force (4%)

Reference Laur et al.89

Trial design CBTSample 16 men and 16 women (hamstrings)Interventions 3 x 20 s static stretchingOutcome measures Perceived exertionMain results S: ↑ perceived exertion

Ballistic stretchingReference Nelson et al.16

Trial design CBTSample 11 male and 11 female college studentsInterventions Ballistic stretch: 15 bob up and down once per minOutcome measures (i) Sit and reach score; (ii) maximum strength (1RM)Main results S: ↑ ROM (7.5%); ↓ strength (7.3%)

determined by maximum lifting capacity5,16,20 and iso-metric contraction force.2 Nelson et al.4 reported that adecrease in muscle strength for slow velocities of move-ment after static stretching, and decreases in perfor-mance of functional high velocity movements, such asjumping, after static stretching.1,4,6,7 The negative acuteeffect of stretching on performance is probablyexplained by the change in neuromuscular transmissionand/or biomechanical properties of muscle. Severalstudies of the effect of stretching on performance havedemonstrated a reduction of performance associatedwith a decrease in neural activation (H-reflex ampli-tude).2,50,51 Some stretching research has reported anincrease in muscle compliance as investigated by range

of motion,22,26,34,76,77 active muscle stiffness,1,47 and pas-sive stiffness.28,29

A study by Fowles et al.2 assessed strength perfor-mance after prolonged stretch (13 maximal stretches,2 min and 15 s hold, and 5 s rest) by measuring force,EMG activity, and passive stiffness. The greatestdegree of strength loss was immediately after stretch(28%), and this lasted more than 1 h after stretch(9%). Interestingly, muscle activation and EMG activ-ity was significantly depressed after stretching, butwas recovered by 15 min. Passive stiffness recoveredquickly after stretching at 15 min but did not fullyrecover within 1 h. The results imply that theimpaired muscle activation was responsible for

THE MECHANISMS AND BENEFITS OF STRETCHING 201

PNF stretching (short term)Reference Wiktorsson-Moller et al.26

Trial design CBTSample 8 healthy malesInterventions PNF: isometric contraction 4–6 s, relax 2 s, passive stretching 8 sOutcome measures (i) ROMs of lower extremities; (ii) hamstrings and quadriceps strengthMain results S: ↑ ROM of ankle dorsiflexion and plantar flexion, hip flexion, extension, abduction, knee flexion

Reference Church et al.20

Trial design CBTSample 40 womenInterventions (i) Static stretching; (ii) PNFOutcome measures (i) Vertical jump; (ii) ROMMain results S: ↓ jump height (3%)

PNF stretching (long term)Reference Wilson et al.47

Trial design CCTSample 16 male weight-lifters (n = 9 in experimental group, n = 7 in control group)Interventions Flexibility training (6–9 repeats) of upper extremities, 10–15 min per session, twice a week for 8 weeksOutcome measures (i) Rebound bench press (RBP); (ii) purely concentric bench press (PCBP)Main results S: ↑ ROM (3%); ↑ RBP (5.4%); ↓ SEC stiffness (7.2%)

Reference Worrell et al.64

Trial design CCTSample 19 participants with short hamstrings (one leg static, one leg PNF)Interventions Static – 15 s held, 15 s rest; PNF – 5 s isometric, 5 s rest, 4 repeats per day, 5 days/week, 3 weeksOutcome measures (i) ROM; (ii) con/ecc strengthMain results S: ↑ strength Ecc – 60 & 120°/s; Con – 120°/s

Reference Hande et al.88

Trial design CCTSample 16 men (one leg stretch, one leg control)Interventions CR for 8 weeks (isometric at 70% MVC, 1–2 s rest, 10–15 s passive stretching). Follow-up at 0, 4, 8 weeksOutcome measures Knee flexion and extension. Con: 240, 180, 120, 60°/s. Ecc: 60 and 120°/sMain results S: ↑ torque. Extension – ecc at 120 & 60°/s. Flexion – all velocities

Reference Hunter & Marshall90

Trial design RCTSample 60 participants (15 per group)Interventions Static stretching (3 x 20 s) and PNF (submaximal contraction 10 s)Outcome measures (i) Drop jump; (ii) countermovement jumpMain results NS

CCT, controlled clinical trial; RCT, randomised controlled trial; PPT, pre- & post-test trial; CBT, counterbalance trial; ROM, range ofmotion; EMG, electromyography; SEC, series elastic components; S, significant; NS, non-significant; MVC, maximum voluntarycontraction; con, concentric contraction; ecc, eccentric contraction; RM, repetitive maximum.

strength loss after prolonged stretching in the earlyphase, while impaired contractile force was responsi-ble for strength loss throughout the entire period.This is consistent with the results of several studieswhere the mechanism responsible for a decrease inperformance after acute stretching is likely to be neu-romuscular inhibition.1,87 Behm et al.87 investigatedthe effects of static stretching (held for 45 s, and restfor 15 s for five times) on voluntary and evoked force,and EMG activity of quadriceps. Maximal voluntaryand evoked contraction decreased similarly by 12%,and muscle activation and EMG activity decreased by2.8% and 20%, respectively. Similarly, Cornwell et al.1

reported that static stretching of gastrosoleus (180 s)reduced jump height by 7.4% but active stiffness wasreduced by only 2.8%. Other studies have reportedthat static stretching reduced jumping performance(knee bend) by 3%.6,7 A reduction in jumping perfor-mance was consistent with a reduction of EMG activ-ity,7 but there were no changes in biomechanicalvariables (vertical velocity, knee angle, duration ofconcentric and eccentric phases).6 The prolongedstretch in the study of Fowles et al.2 (75 s for 13 times)might have increased muscle compliance more than inany other study as there was evidence that staticstretching held for 90 s for five times could decreasemuscle stiffness.28,29 Shorter durations of stretch couldnot change passive muscle properties,22,42 thus sug-gesting that the change in muscle compliance andmuscle inactivation in the study of Fowles et al.2

caused more strength loss (28%) than that reportedby Behm et al. (12%).87

The detrimental effects of acute stretching exerciseon muscular endurance has been demonstrated byLaur et al.:89 application of acute stretching reducedthe maximal number of repetitions performed with asubmaximal load, and also produced higher per-ceived exertion scores. Although the magnitude ofreduction was small, it was statistically significant.

Interestingly, studies on the effects of long-termstretching reported a positive effect of stretching onperformance.47,64,88 Three weeks of flexibility trainingin both PNF and static stretching increased peaktorque of hamstrings eccentrically (at 60°/s and120°/s) and concentrically (at 120°/s only).64 PNFtraining (contract–relax technique) for 8 weeksincreased maximum torque of knee flexors and exten-sors.88 The increase in muscle strength of knee flexorswas significant at all velocities and might be due tothe contraction phase of the PNF stretching tech-nique showing the same effect as isometric muscletraining. Knee flexors, which are used in normalactivity less than knee extensors, showed the greatestincrease in muscle strength. More functionally,Wilson et al.47 reported that 8 weeks of static flexibil-

ity training increased rebound bench press perfor-mance by 5.4%, concomitantly with a decrease inactive muscle stiffness by 7.2%. The authors proposedthat flexibility-induced performance enhancementmight result from increased musculotendinous com-pliance facilitating the use of energy strain in stretchshort-cycle activities. In contrast, with the acutestretching, Hunter and Marshall90 reported that thecombination of static and PNF training for 10 weeksdid not result in a detrimental effect on countermove-ment jump and drop jump, but helped to increaseknee joint range of motion.90 Therefore, flexibilitytraining of at least three weeks is beneficial to someperformance factors as indicated by increased rangeof motion and muscle strength. Unfortunately, thereis no research on the effects of flexibility training onneuromuscular activity.

There are no studies on the effects of acute stretch-ing after prolonged flexibility training. It is unclearwhether the negative effects of acute stretching willattenuate the positive effects of flexibility training,because the athletes who commonly undertake flexi-bility training, also perform stretching before compe-tition. There are also no studies on the effects ofdynamic stretching on performance. As a range ofstudies have indicated detrimental effects on athleticperformance of all stretching techniques (static, PNF,and ballistic),2,7,20,87,88 dynamic stretching might be auseful protocol to increase flexibility without decreas-ing athletic performance.

The efficiency of exercise

Flexibility is considered to play an important role inthe efficiency of movement,35 by enabling the use ofelastic potential energy in muscle.13 The more compli-ant muscle–tendon unit needs more contractile forceto transmit to the joint and, consequently, causes agreater delay in external force generation.13 A stiffermuscle would provide a more efficient transmission ofcontractile force production,13 but this contradicts theaim of stretching, which intends to increasemuscle–tendon unit compliance.

Craib et al.91 reported that less flexible runnersshowed a reduced aerobic demand during running(better running economy). The positive and signifi-cant correlation between range of motion and the aer-obic demand of running presented in only twomovements (external hip rotation and dorsiflexion)and accounted for 47% of the variance observed inrunning economy. This study91 was cross-sectional,did not control the training programme of the run-ners, and did not consider other factors that mighthave influenced running economy such as kinematic,

202 WEERAPONG ET AL.

anthropometric, physiological, and cellular variables. Incontrast, the flexibility training of hip flexors (3 weeks)92

and lower leg muscles (quadriceps, hamstrings, and gas-trosoleus; 10 weeks)93 resulted in increased range ofmotion but had no effect on running economy. None ofthe published papers in this area has assessed the effectsof flexibility on running time, stride length, stride fre-quency, or the perception of fatigue. The optimal level offlexibility (static and dynamic) for running performanceneeds to be researched.

THE EFFECTS OF STRETCHING ON INJURY PREVENTION

Rate of injury

Despite performance of stretching before exercise gener-ally being recommended to reduce the risk of injury, arecent review of the effects of stretching on the incidenceof injury indicated inconclusive results.10 This might bedue to exercise-related injury being a complex phenome-non with a variety of physiological, psychological, andenvironmental factors involved. Furthermore, themajority of research in this area has been retrospectiveand does not provide a clear relationship between flexi-bility and injury.13,94,95

In one prospective study, Van Mechelen et al.96 pro-vided a standardised programme of stretching exercisesto runners and assessed the number of injuries after16 weeks. There was no reduction in injury incidence per1000 h of running between the experimental (standard-ised programme of stretching exercise) and the control(no stretching information) groups. In a study of armyrecruits,97,98 a pre-exercise stretching programme of11–12 weeks did not reduce the risk of exercise-relatedinjury. Fitness and age,98 and the early detection ofsymptoms of overuse injuries96 were more importantfactors for injury rather than the stretching exercise. Thismight be due to the fact that increases in range ofmotion (or static flexibility) resulting from stretchingmay not confer benefits for a majority of sports such asrunning and swimming which do not require extremeranges of motion. Dynamic flexibility might be moreimportant because it represents the resistance of themuscle–tendon unit during movement; despite this,there is no research on the relationship between dynamicflexibility and rate of injury.

Muscle damage

In contrast to the general belief that stretching helpsreduce the risk of muscle damage from unaccustomedeccentric exercise, previous research has reported that

both prolonged static and ballistic stretching (60 s fortwo stretches for each muscle) actually induced mus-cle damage.17 Moreover, no research has reported anybenefit from stretching on the severity of muscle dam-age or muscle soreness either before3,85,86 or after85

exercise. The effects of stretching before and aftereccentric exercise have been reviewed recently.99 Thelack of evidence for the benefits of stretching may bedue to previous research only investigating staticstretching. If the decrease in passive muscle stiffness isthe key to reducing the severity of muscle damage,static stretching in these studies3,85,86 was not held forlong enough to induce a decrease in passive stiffness(most studies used stretches held less than 90 s).Interestingly, a study by McNair et al.38 reported thatpassive dynamic stretching did reduce muscle stiff-ness, while a study by McHugh et al.71 reported thatpassive stiffness was related to the severity of muscledamage measured by strength loss, pain, muscle ten-derness, and creatine kinase activity. Therefore, anystretching technique that can reduce passive stiffnessmight help to reduce the severity of muscle damage.To date, there are no published papers reporting theeffects of dynamic stretching on the severity of muscledamage. The effects of other stretching techniquessuch as ballistic and PNF on the severity of muscledamage have also not as yet been determined.

SUMMARY AND CONCLUSIONS

Despite stretching commonly being performed beforeexercise to enhance performance and reduce the riskof injury, there is limited scientific data to support thesuggested benefits of stretching. Static and ballisticstretching have been shown to have detrimentaleffects on muscle strength and functional perfor-mances such as jumping, and to have inconclusiveeffects on the incidence of injury, and no effects onthe severity of muscle damage. Even though researchhas indicated that stretching is an effective treatmentto increase static flexibility (range of motion), theeffects on dynamic flexibility (muscle stiffness) areinconclusive given the variation of the length of holdand the number of repetitions used in studies. Theaim of stretching is to increase flexibility, but doesflexibility help to enhance performance? The idealflexibility for the performance of each sports activityis different. Compliant muscle might be beneficial toeccentric contraction while stiffer muscle might bemore suitable for concentric and isometric contrac-tions. Does flexibility help to reduce the rate ofinjury? The majority of research does not support thisstatement. In fact, the majority of movement in sportsrequires repetitive movements within the normal

THE MECHANISMS AND BENEFITS OF STRETCHING 203

range of motion. An increase in range of motion,therefore, is not necessary. The aim to reduce resis-tance during repetitive movement might be more ben-eficial in terms of increasing quality of movement andreducing the risk of overuse injury. Practically, theoptimal level of flexibility is required because theincrease in flexibility (more compliant muscle) mightnot benefit performance but may help to reduce therisk of injury. The compliant muscle–tendon unitabsorbs and requires more energy to shorten, andconsequently delays and reduces external force pro-duction. Nevertheless, the increase in ability to absorbenergy in the compliant muscle might help to reducethe mechanical overload on muscle fibres, and conse-quently reduce the risk of muscle injury and the sever-ity of muscle damage. Further research is needed toinvestigate the appropriate stretching techniques andthe optimal level of flexibility which can maintain orimprove performance, or which can prevent injury.

RECOMMENDATIONS

In order to clarify the effects of stretching, furtherresearch is recommended to:1. Provide information on the relationship of

dynamic flexibility, performance, and rate ofinjury.

2. Examine the effects of several stretchingtechniques such as ballistic, PNF, and dynamicstretching on dynamic flexibility andneuromuscular sensitivity.

3. Compare the effects of several stretchingtechniques such as static, ballistic, PNF, anddynamic stretching on different types ofperformance, the severity of muscle soreness,running economy, and rate of injury.

4. Study the effects of acute stretching after long-term flexibility training.

5. Provide more information on the appropriateflexibility level to enhance performance andreduce the risk of injury.

ACKNOWLEDGEMENTS

The authors would like to acknowledge HuachiewChalermprakiet University (Thailand), TheAmerican Massage Therapy Association Foundation,and the New Zealand Institute of Sport andRecreation Research (Division of Sport andRecreation, Auckland University of Technology, NewZealand) for funding assistance in the preparation ofthis manuscript.

REFERENCES

1 Cornwell A, Nelson A, Sidaway B. Acute effects of stretching onthe neuromechanical properties of the triceps surae musclecomplex. Eur J Appl Physiol 2002;86:428–34

2 Fowles J, Sale D, MacDougall J. Reduced strength after passivestretch of the human plantar flexors. J Appl Physiol2000;89:1179–88

3 Johansson P, Lindstrom L, Sundelin G, Lindstrom B. The effectsof preexercise stretching on muscular soreness, tenderness andforce loss following heavy eccentric exercise. Scand J Med SciSports 1999;9:219–25

4 Nelson A, Guillory I, Cornwell A, Kokkonen J. Inhibition ofmaximal voluntary isokinetic torque production followingstretching is velocity-specific. J Strength Cond Res 2001;15:241–6

5 Kokkonen J, Nelson A, Cornwell A. Acute muscle stretchinginhibits maximal strength performance. Res Q Exerc Sport1998;69:411–5

6 Knudson D, Bennett K, Corn R, Leick D, Smith C. Acute effectsof stretching are not evident in the kinematics of the verticaljump. J Strength Cond Res 2001;15:98–101

7 Young W, Behm D. Effects of running, static stretching andpractice jumps on explosive force production and jumpingperformance. J Sports Med Phys Fitness 2003;43:21–7

8 Magnusson S. Passive properties of human skeletal muscleduring stretch manoeuvres. Med Sci Sports Exerc 1998;8:65–77

9 Witvrouw E, Mahieu N, Danneels L, McNair P. Stretchingand injury prevention: an obscure relationship. Sports Med2004;34:443–9

10 Weldon S, Hill R. The efficacy of stretching for prevention ofexercise-related injury: a systematic review of the literature. ManTher 2003;8:141–50

11 Thacker S, Gilchrist J, Stroup D, Kimsey CJ. The impact ofstretching on sports injury risk: a systematic review of theliterature. Med Sci Sports Exerc 2004;36:371–8

12 Shellock F, Prentice W. Warming-up and stretching for improvedphysical performance and prevention of sports-related injuries.Sports Med 1985;2:267–78

13 Gleim GW, McHugh MP. Flexibility and its effects on sportsinjury and performance. Sports Med 1997;24:289–99

14 Magnusson S, Simonsen E, Aagaard P, Dyhre-Poulsen P,McHugh M, Kjaer M. Mechanical and physical responses tostretching with and without preisometric contraction in humanskeletal muscle. Arch Phys Med Rehabil 1996;77:373–8

15 Taylor D, Dalton J, Seaber A, Garrett W. Viscoelastic propertiesof muscle–tendon units: the biomechanical effects of stretching.Am J Sports Med 1990;18:300–9

16 Nelson A, Kokkonen J. Acute ballistic muscle stretching inhibitsmaximal strength performance. Res Q Exerc Sport 2001;72:415–9

17 Smith L, Brunetz M, Chenier T, McCammon M, Houmard J,Franklin M et al. The effects of static and ballistic stretching ondelayed onset muscle soreness. Res Q Exerc Sport 1993;64:103–7

18 Burke D, Culligan C, Holt L. The theoretical basis of pro-prioceptive neuromuscular facilitation. J Strength Cond Res2000;14:496–500

19 Spernoga S, Uhl T, Arnold B, Gansneder B. Duration ofmaintained hamstring flexibility after one-time, modified hold-relax stretching protocol. J Athl Train 2001;36:44–8

20 Church J, Wiggins M, Moode E, Crist R. Effects of warm-upand flexibility treatments on vertical jump performance. JStrength Cond Res 2001;15:332–6

21 Smith C. The warm-up procedure: to stretch or not to stretch. Abrief review. J Orthop Sports Phys Ther 1994;19:12–7

22 Halbertsma J, van Bolhuis A, Goeken L. Sport stretching: effectson passive muscle stiffness of short hamstrings. Arch Phys MedRehabil 1996;77:688–92

204 WEERAPONG ET AL.

23 Taylor D, Brooks D, Ryan J. Viscoelastic characteristics ofmuscle: passive stretching versus muscular contractions. Med SciSports Exerc 1997;29:1619–24

24 Henricson A, Fredriksson K, Persson I, Pereira R, Westlin N.The effects of heat and stretching on the range of hip motion. JOrthop Sports Phys Ther 1984;6:110–5

25 Zito M, Driver D, Parker C, Bohannon R. Lasting effects of onebout of two 15-second passive stretches on ankle dorsiflexionrange of motion. J Orthop Sports Phys Ther 1997;26:214–21

26 Wiktorsson-Moller M, Oberg B, Ekstrand J, Gillquist J. Effectsof warming up, massage, and stretching on range of motion andmuscle strength in the lower extremity. Am J Sports Med1983;11:249–52

27 Muir I, Chesworth B, Vandervoort A. Effect of a static calf-stretching exercise on the resistive torque during passive ankledorsiflexion in healthy subjects. J Orthop Sports Phys Ther1999;29:421–4

28 Magnusson S, Simonsen E, Aagaard P, Sorensen H, Kjaer M.A mechanism for altered flexibility in human skeletal muscle. JPhysiol 1996;497:291–8

29 Magnusson S, Simonson E, Dyhre-Poulsen P, Aagaard P,Mohr T, Kjaer M. Viscoelastic stress relaxation during staticstretch in human skeletal muscle in the absence of EMGactivity. Med Sci Sports Exerc 1996;6:323–8

30 McHugh M, Magnusson S, Gliem G, Nicholas J. Viscoelasticstress relaxation in human skeletal muscle. Med Sci SportsExerc 1996;24:1375–82

31 McHugh M, Kremenic I, Fox M, Gleim G. The role ofmechanical and neural restraints to joint range of motionduring passive stretch. Med Sci Sports Exerc 1998;30:928–32

32 Klinge K, Magnusson S, Simonsen E, Aagaard P, Klausen K,Kjaer M. The effects of strength and flexibility training onskeletal muscle electromyographic activity, stiffness, andviscoelastic stress relaxation response. Am J Sports Med1997;25:710–6

33 Cornwell A, Nelson A. The acute effects of passive stretchingon active musculotendinous stiffness. Med Sci Sports Exerc1997;29:s281

34 McNair P, Stanley S. Effect of passive stretching and joggingon the series muscle stiffness and range of motion of the anklejoint. Br J Sports Med 1996;30:313–8

35 Gajdosik R, Giuliani C, Bohannon R. Passive compliance andlength of the hamstring muscles of the healthy men andwomen. Clin Biomech 1990;5:23–9

36 Magnusson S, Simonsen E, Aagaard P, Gleim G, McHugh M,Kjaer M. Viscoelastic response to repeated static stretching inthe human hamstring muscle. Scand J Med Sci Sports1995;5:342–7

37 McHugh MP, Magnusson S, Gliem G, Nicholas J. Viscoelasticstress relaxation in human skeletal muscle. Med Sci SportsExerc 1992;24:1375–82

38 McNair P, Dombroski E, Hewson D, Stanley S. Stretching atthe ankle joint: viscoelastic responses to holds and continuouspassive motion. Med Sci Sports Exerc 2000;33:354–8

39 Kubo K, Kanehisa H, Fukunaga T. Is passive stiffness inhuman muscles related to the elasticity of tendon structures?Eur J Appl Physiol 2001;85:226–32

40 Cavagna G, Citterio G. Effects of stretching on the elasticcharacteristics and the contractile component of frog striatedmuscle. J Physiol 1974;239:1–14

41 Kubo K, Kanehisa H, Fukunaga T. Effects of resistance andstretching training programmes on the viscoelastic propertiesof human tendon structures in vivo. J Physiol 2002;538:219–26

42 Magnusson S, Aagaard P, Larsson B, Kjaer M. Passive energyabsorption by human muscle–tendon unit is unaffected byincrease in intramuscular temperature. J Appl Physiol

2000;88:1215–2043 Vanderby R, Provenzano P. Collagen in connective tissue:

from tendon to bone. J Biomech 2003;36:1523–744 Silver F, Freeman J, Seehra G. Collagen self-assembly and the

development of tendon mechanical properties. J Biomech2003;36:1529–53

45 Kjaer M. Role of extracellular matrix in adaptation of tendonand skeletal muscle to mechanical loading. Physiol Rev2004;84:649–98

46 Gajdosik R. Passive extensibility of skeletal muscle: review ofthe literature with clinical implications. Clin Biomech2001;16:87–101

47 Wilson G, Elliott B, Wood G. Stretch shorten cycleperformance enhancement through flexibility training. MedSci Sports Exerc 1992;21:116–23

48 Wilson G, Murphy A, Pryor J. Musculotendinous stiffness: itsrelationship to eccentric, isometric and concentricperformance. J Appl Physiol 1994;76:2714–9

49 Mohr K, Pink M, Elsner C, Kvitne R. Electromyographicinvestigation of stretching: the effect of warm-up. Clin J SportMed 1998;8:215–20

50 Vujnovich A, Dawson N. The effects of therapeutic musclestretch on neural processing. J Orthop Sports Phys Ther1994;20:145–53

51 Avela J, Kyrolainen H, Komi P. Altered reflex sensitivity afterrepeated and prolonged passive muscle stretching. J ApplPhysiol 1999;86:1283–91

52 Rosenbaum D, Henning E. The influence of stretching andwarm-up exercises on Achilles tendon reflex activity. J SportsSci 1995;13:481–90

53 Gollhofer A, Schopp A, Rapp W, Stroinik V. Changes in reflexexcitability following isometric contraction in humans. Eur JAppl Physiol 1998;77:89–97

54 Moore M, Kukulka C. Depression of Hoffmann reflexesfollowing voluntary contraction and implications forproprioceptive neuromuscular facilitation therapy. Phys Ther1991;71:321–33

55 Thigpen L, Moritani T, Thiebaud R, Hargis J. The acute effects ofstatic stretching on alpha motoneuron excitability. In: Winter W,Norman R, Wells R, Hayes K, Patla A. (eds) Biomechanics IX-A.Champaign, IL: Human Kinetics, 1985:353–7

56 Zehr E. Considerations for the use of the Hoffmann reflex inexercise studies. Eur J Appl Physiol 2002;86:455–68

57 Morelli M, Seaborne D, Sullivan S. Changes in H-reflexamplitude during massage of triceps surae in healthy subjects.J Orthop Sports Phys Ther 1990;12:55–9

58 Morelli M, Seaborne D, Sullivan S. H-reflex modulationduring manual muscle massage of human triceps surae. ArchPhys Med Rehabil 1991;72:915–9

59 Morelli M, Chapman C, Sullivan S. Do cutaneous receptorscontribute to the changes in the amplitude of the H-reflexduring massage? Electromyogr Clin Neurophysiol1999;39:441–7

60 Guissard N, Duchateau J, Hainaut K. Mechanisms ofdecreased motoneuron excitation during passive musclestretching. Exp Brain Res 2001;137:163–9

61 Ribot-Ciscar E, Tardy-Gervet M, Vedel J, Roll J. Post-contraction changes in human muscle spindle resting dischargeand stretch sensitivity. Exp Brain Res 1991;86:673–8

62 Hunter D, Coveney V, Spriggs J. Investigation into the effect ofstatic stretching on the active stiffness and dampingcharacteristics of the ankle joint plantar flexors. Phys TherSport 2001;2:15–22

63 Magnusson S, Simonsen E, Aagaard P, Kjaer M.Biomechanical responses to repeated stretches in humanhamstring muscle in vivo. Am J Sports Med 1996;24:622–8

THE MECHANISMS AND BENEFITS OF STRETCHING 205

64 Worrell T, Smith T, Winegardner J. Effects of hamstringstretching on hamstring muscle performance. J Orthop SportsPhys Ther 1994;20:154–9

65 Wiemann K, Hahn K. Influences of strength, stretching andcirculatory exercises on flexibility parameters of the hamstrings.Int J Sports Med 1997;18:340–6

66 Enoka R, Hutton R, Eldred E. Changes in excitability of tendontap and Hoffmann reflexes following voluntary contractions.Electroencephalogr Clin Neurophysiol 1980;48:664–72

67 Sady S, Wortman M, Blanke D. Flexibility training: ballistic,static or proprioceptive neuromuscular facilitation? Arch PhysMed Rehabil 1982;63:261–3

68 Toft E, Sinkjaler T, Kalund S, Espersen G. Biomechanicalproperties of the human ankle relation to passive stretch. JBiomech 1989;22:1129–32

69 Cornelius W, Jensen R, Odell M. Effects of PNF stretchingphases on acute arterial blood pressure. Can J Appl Physiol1995;20:222–9

70 Osternig L, Robertson R, Troxel R, Hansen P. Muscle activationduring proprioceptive neuromuscular facilitation (PNF)stretching techniques. Am J Phys Med Rehabil 1987;66:298–307

71 McHugh MP, Connolly DA, Eston RG, Kremenic I, Nicholas SJ,Gleim GW. The role of passive muscle stiffness in symptoms ofexercise-induced muscle damage. Am J Sports Med 1999;27:594–9

72 Stewart I, Sleivert G. The effects of warm-up intensity on range ofmotion and anaerobic performance. J Orthop Sports Phys Ther1998;27:154–61

73 Knight C, Rutledge C, Cox M, Acosta M, Hall S. Effect ofsuperficial heat, deep heat, and active exercise warm-up on theextensibility of the plantar flexors. Phys Ther 2001;81:1206–14

74 McHugh MP, Connolly DAJ, Eston RG, Gleim GW. Exercise-induced muscle damage and potentials mechanisms for therepeated bout effect. Sports Med 1999;27:157–70

75 Noonan T, Best T. Thermal effects on skeletal muscle tensilebehaviour. Am J Sports Med 1993;21:517–25

76 Warren C, Lehmann J, Koblanski J. Heat and stretch procedures:an evaluation using rat tail tendon. Arch Phys Med Rehabil1976;57:122–6

77 Taylor B, Waring C, Brashear T. The effects of therapeuticapplication of heat or cold followed by static stretch on hamstringmuscle length. J Orthop Sports Phys Ther 1995;21:283–6

78 Burke D, Holt L, Rasmussen R, MacKinnon N, Vossen J, PelhamT. Effects of hot or cold water immersion and modifiedproprioceptive neuromuscular facilitation flexibility exercise onhamstring length. J Athl Train 2001;36:16–9

79 Draper D, Prentice W. Therapeutic ultrasound. In: Prentice W.(ed) Therapeutic Modalities for Physical Therapists, 2nd edn. NewYork: McGraw-Hill, 2002:270–304

80 Rodenburg J, Steenbeek D, Schiereck P, Bar P. Warm-up,stretching and massage diminish harmful effects of eccentricexercise. Int J Sports Med 1994;15:414–9

81 Nosaka K, Clarkson P. Influence of previous concentric exerciseon eccentric exercise-induced muscle damage. J Sports Sci1997;15:477–83

82 Bale P. Massage, warmdown and rest as recuperative measuresafter short term intense exercise. Physiother Sport 1991;13:4–7

83 Tiidus P, Shoemaker J. Effleurage massage, muscle blood flowand long term post-exercise recovery. Int J Sports Med1995;16:478–83

84 Smith M, Jackson C. Delayed onset muscle soreness (DOMS),serum creatine kinase (SCK), and creatine kinase-MB (%CK-MB) related to performance measurements in football. Med SciSports Exerc 1990;22:S:34

85 High D, Howley E, Franks B. The effects of static stretching andwarm-up on prevention of delayed-onset muscle soreness. Res QExerc Sport 1989;60:357–61

86 Lund H, Vestergaard-Poulsen P, Kanstrup I, Sejrsen P. The effectsof passive stretching on delayed onset muscle soreness, and otherdetrimental effects following eccentric exercise. Scand J Med SciSports 1998;8:216–21

87 Behm D, Button D, Butt J. Factors affecting force loss withprolonged stretching. Can J Appl Physiol 2001;26:261–72

88 Handel M, Horstmann T, Dickhuth H, Gulch R. Effects ofcontract-relax stretching training on muscle performance inathletes. Eur J Appl Physiol 1997;76:400–8

89 Laur D, Anderson T, Geddes G, Crandall A, Pincivero D. Theeffects of acute stretching on hamstring muscle fatigue andperceived exertion. J Sports Sci 2003;21:163–70

90 Hunter J, Marshall R. Effects of power and flexibility training onvertical jump technique. Med Sci Sports Exerc 2002;34:478–86

91 Craib M, Mitchell V, Fields K, Cooper T, Hopewell R, MorganD. The association between flexibility and running economy insub-elite male distance runners. Med Sci Sports Exerc1996;28:737–43

92 Godges J, MacRae P, Engelke K. Effects of exercise on hip rangeof motion, trunk muscle performance, and gait economy. PhysTher 1993;73:468–77

93 Nelson A, Kokkonen J, Eldredge C, Cornwell A, Glickman-WeissE. Chronic stretching and running economy. Scand J Med SciSports 2001;11:260–5

94 Watson A. Sports injuries related to flexibility, posture,acceleration, clinical defects, and previous injury, in high-levelplayers of body contact sports. Int J Sports Med 2001;22:222–5

95 Dubravcic-Simunjak S, Pecina M, Kuipers H, Moran J, Haspl M.The incidence of injuries in elite junior figure skaters. Am J SportsMed 2003;31:511–7

96 van Mechelen W, Hlobil H, Kemper H, Voorn W, de Jongh H.Prevention of running injuries by warm-up, cool-down, andstretching exercise. Am J Sports Med 1993;21:711–9

97 Pope R, Herbert R, Kirwan J. Effects of ankle dorsiflexion rangeand pre-exercise calf muscle stretching on injury risk in armyrecruits. Aust J Physiother 1998;44:165–77

98 Pope R, Herbert R, Kirwan J, Graham B. A randomised trial ofpre-exercise stretching for prevention of lower-limb injury. MedSci Sports Exerc 2000;32:271–7

99 Herbert R, Gabriel M. Effects of stretching before and afterexercising on muscle soreness and risk of injury: systemic review.Br J Sports Med 2002;325:1–5

206 WEERAPONG ET AL.

PATRIA A. HUME PhD

Associate Professor, New Zealand Institute of Sport and Recreation Research, Division of Sport and Recreation,

Faculty of Health and Environmental Studies, Auckland University of Technology, Private Bag 92006, Auckland, New Zealand

Tel: +64 9 917 9999 ext 7306; Fax: +64 9 917 9960; E-mail: patria.hume@aut.ac.nz