HAMSTRING STRAIN INJURY

Transcription

1 HAMSTRING STRAIN INJURY The role of strength & voluntary activation Matthew N. Bourne B. App Sci. HMS. (Hons) 2016 Doctor of Philosophy (Thesis by publication) School of Exercise and Nutrition Sciences Faculty of Health Queensland University of Technology

2 i

3 Table of Contents Table of Contents... ii Abstract... iv List of Figures... x List of Tables... xii List of Abbreviations... xiii Statement of Original Authorship... xiv Chapter 1: INTRODUCTION... 1 Chapter 2: LITERATURE REVIEW Definition of hamstring strain injury Incidence of hamstring strain injury in sport Recurrence of hamstring strain injury in sport Hamstring anatomy Mechanism(s) of injury Hamstring function during high-speed running and propensity for injury Proposed risk factors for hamstring strain injury Unalterable risk factors Alterable risk factors Factors underpinning high rates of hamstring strain injury recurrence Mechanism(s) for chronic strength deficits following hamstrings strain injury Neuromuscular Inhibition Evidence for incomplete activation in maximal voluntary contractions Mechanism(s) underpinning neural inhibition The impact of resistance training on skeletal muscle activation The impact of pain and injury on skeletal muscle activation Evidence for neuromuscular inhibition following hamstring strain injury Impact of neuromuscular inhibition on hamstring muscle morphology and architecture Neuromuscular inhibition as a mechanism for high rates of hamstring strain injury recurrence Chapter 3: PROGRAM OF RESEARCH Chapter 4: STUDY 1 ECCENTRIC STRENGTH AND HAMSTRING INJURY RISK IN RUGBY UNION: A PROSPECTIVE COHORT STUDY ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION ii

4 Chapter 5: STUDY 2 REDUCED ACTIVATION OF PREVIOUSLY INJURED BICEPS FEMORIS LONG HEAD MUSCLES IN RUNNING Linking Paragraph ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Chapter 6: STUDY 3 IMPACT OF EXERCISE SELECTION ON HAMSTRING MUSCLE ACTIVATION Linking Paragraph ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Chapter 7: STUDY 4 ADAPTABILITY OF HAMSTRING ARCHITECTURE AND MORPHOLOGY TO TARGETED RESISTANCE TRAINING Linking Paragraph ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION Chapter 8: GENERAL DISCUSSION, LIMITATIONS & CONCLUSION169 Bibliography Appendices iii

5 Abstract Hamstring strain injuries (HSIs) are endemic in sports involving high-speed running. These injuries typically occur when athletes run at maximal or near maximal speeds and upwards of 80% affect the biceps femoris long head (BFLH). High rates of injury recurrence (16-54%) are also concerning, particularly given the tendency for re-injuries to be more severe than the initial insult. These observations suggest that more is to be learnt about the mechanisms underpinning first-time and recurrent HSIs, while also suggesting that prophylactic programs should specifically target the BFLH. This thesis aimed to, firstly, explore the role of eccentric strength and between-limb imbalance in HSI occurrence, and determine if previously injured hamstrings display altered neuromuscular function during high-speed running. The second aim of this thesis was to characterise the activation patterns and the malleability of hamstring muscle architecture and morphology to different strengthening exercises, in an attempt to improve HSI prevention programs by better targeting the site of injury. The aim of study 1 was to determine if lower levels of eccentric knee-flexor strength or greater between-limb imbalances in eccentric strength are risk-factors for HSI. This study found that athletes with between-limb imbalance in eccentric knee-flexor strength of 15% and 20% increased the risk of HSI 2.4 fold (RR = 2.4, 95% CI = 1.1 to 5.5, p = 0.033) and 3.4 fold (RR = 3.4, 95% CI = 1.5 to 7.6, p = 0.003), respectively. Furthermore, the risk of reinjury was augmented in players with strength imbalances (p < 0.001). Study 2 aimed to determine: 1) the spatial patterns of hamstring muscle activation during high-speed overground running in limbs with and without a prior HSI and; 2) whether previously injured hamstring muscles exhibit lasting deficits in cross-sectional area (CSA). iv

6 Ten elite male athletes with a history of unilateral BFLH strain injury underwent functional magnetic resonance imaging before and immediately after a repeat-sprint running protocol. This study demonstrated that previously injured BFLH muscles displayed a significantly lower percentage increase in transverse relaxation time after the running protocol, compared to uninjured contralateral BFLH muscles (mean difference = 12.0%, p < 0.001). However, no between-limb differences in CSA were observed for any hamstring muscles. The purpose of Study 3 was to determine the extent to which different strength training exercises selectively activate the commonly injured BFLH muscle. Part 1 employed surface electromyography (EMG) to measure hamstring activation during 10 common exercises and found that, in eccentric contractions, the largest BF/MH normalised EMG (nemg) ratio was observed in the 45 hip extension exercise (HE) and the lowest was observed in the Nordic hamstring (NHE) and bent-knee bridge exercises. Part 2 used fmri to explore the spatial patterns of hamstring activation in the 45 HE and NHE and revealed that the BFLH was significantly more active in the 45 HE than the NHE (p < 0.001). Based on the results from Study 3, Study 4 aimed to evaluate changes in hamstring muscle volume, anatomical cross-sectional area (ACSA) and BFLH fascicle length following 10- weeks of NHE or HE training, or a period of no training (CON). This study found that BFLH fascicles were significantly longer in the NHE and HE groups after 5 (p < 0.001) and 10 weeks of training (p < 0.001) but remained unchanged for the CON group (p > 0.05). The HE group displayed a greater percentage increase in BFLH volume than the NHE (p < 0.037) and CON (p < 0.001) groups. Similarly, BFLH ACSA increased more in the HE group than the NHE (p = 0.047) and CON groups (p < 0.001). Both exercises induced similar (p > 0.05) increases in semitendinosus volume and ACSA which were greater than those observed for v

7 the CON group (all p 0.002). However, only the NHE group exhibited increased BF short head ACSA, and only the HE group displayed increased semimembranosus volume (p = 0.007) and ACSA (p = 0.015), compared to the CON group. This program of research has contributed new knowledge relating to factors which may predispose to, and manifest as a result of HSI, while also providing novel data which may be used to inform injury preventive and rehabilitation practices. This thesis has provided evidence 1) that between-limb imbalance in eccentric knee flexor strength is a risk factor for HSI; 2) that previously injured hamstrings display a reduced activation capacity following a return to sport; 3) that different strengthening exercises elicit unique patterns of hamstring muscle activation; and 4) that training with different exercises results in heterogeneous architectural and morphological adaptations in the hamstrings. These data highlight the potential importance of ameliorating eccentric strength imbalances and restoring voluntary activation, particularly following HSI, while also providing an evidence base from which to form decisions regarding exercise selection in prophylactic programs. vi

8 List of publications related to thesis 1. Bourne, MN., Opar, DA., Williams, MD., & Shield, AJ. (2015). Eccentric Kneeflexor Strength and Hamstring Injury Risk in Rugby Union: A prospective study. Am J Sports Med, 43(11): doi: / Bourne, MN., Williams, MD., Opar, DA., Al Najjar, A., & Shield, AJ. (2016). Impact of exercise selection on hamstring muscle activation. Br J Sports Med, Accepted. Manuscripts currently under peer review 1. Bourne, MN., Duhig, SJ., Timmins, RG., Williams, MD., Opar, DA., Al Najjar, A., Kerr, G., & Shield, AJ. (2016). Impact of the Nordic hamstring and hip extension exercises on hamstring architecture and morphology: implications for injury prevention. Br J Sports Med, Under Review. Other relevant publications 1. Bourne, MN., Opar,DA., Williams,MD., Al Najjar, A, & Shield, AJ (2015). Muscle activation patterns in the Nordic hamstring exercise: Impact of prior strain injury. Scand J Med Sci Sports doi: /sms Timmins, R., Bourne, MN., Shield, A., Williams, M., Lorenzon, C., & Opar, D. (2015). Short biceps femoris fascicles and eccentric knee flexor weakness increase the vii

9 risk of hamstring injury in elite football (soccer): a prospective cohort study. Br J Sports Med, [Epub ahead of print]. 3. Timmins, RG., Bourne, MN., Shield, AJ., Williams, MD., Lorenzen, C., & Opar, DA (2015). Biceps femoris architecture and strength in athletes with a prior ACL reconstruction. Med Sci Sports Exerc, [Epub ahead of print]. Grants awarded during candidature 1) Institute of Health and Biomedical Innovation Bourne MN, Shield AJ. (2015) The effect of gender on hamstring muscle activity during selected rehabilitation exercises. $8000 2) Queensland Academy of Sport Centre of Excellence Bourne MN, Shield AJ. (2014) Impact of exercise selection of biceps femoris activation and hypertrophy. $ List of conference presentations 1) Bourne, MN, Opar, DA, Williams, MD, Shield, AJ. Eccentric Strength and Hamstring Injury Risk in Rugby Union: A Prospective corhort study. World Congress on Science and Football. Copenhagen, ) Bourne, MN, Opar DA, Williams MD, Al Najjar A, Shield AJ. Reduced activation of biceps femoris long head muscles in running following strain injury. Sports Medicine Australia. Canberra, viii

10 3) Bourne, MN, Opar DA, Williams MD, Al Najjar A, Shield AJ. Impact of previous strain injury on hamstring muscle activation during high-speed overground running. International Olympic Committee World Congress for the Prevention of Injury and Illness in Sport. Monaco, ) Bourne, MN, Opar DA, Williams MD, Al Najjar A, Shield AJ. Previously injured biceps femoris long head muscles display reduced activation during high-speed overground running: an fmri investigation. IHBI Inspires, Gold Coast, ) Bourne, MN, Opar DA, Williams MD, Al Najjar A, Shield AJ. Hamstring muscle activation during the Nordic hamstring exercise and the impact of previous strain injury: an fmri study. XXII International Conference on Sports Rehabilitation and Traumatology: Football Medicine Strategies for Muscle and Tendon Injuries. London, ) Bourne, MN, Opar DA, Williams MD, Al Najjar A, Shield AJ The impact of previous strain injury on hamstring muscle activation during the Nordic hamstring exercise. American College of Sports Medicine Annual Meeting. Indianapolis, ) Bourne, MN, Opar DA, Williams MD, Al Najjar A, Shield AJ Spatial activation patterns of the knee flexors during the Nordic hamstring exercise: an fmri study. Sports Medicine Australia. Phuket, ) Bourne, MN, Preventing hamstring strain injuries in elite athletes. Queensland Academy of Sport Injury Management Seminar. Brisbane, ) Bourne, MN. The role of neuromuscular inhibition in hamstring strain injury recurrence. IHBI Inspires. Brisbane, ix

11 List of Figures Figure 2-1. Phases of the running gait cycle Figure 2-2. Comparison of knee flexion torque-velocity relationships between previously injured hamstrings, contralateral uninjured hamstrings and reference values from uninjured control subjects Figure 2-3. The torque/force-velocity relationships of electrically stimulated (red) and voluntarily activated (blue) skeletal muscle Figure 2-4. A simplified scheme of the afferent synaptic inputs to alpha (ά) and gamma (γ) motoneurones Figure 2-5. Percentage change in fmri T2 relaxation times of each hamstring muscle for both the previously injured (inj) and uninjured (uninj) limbs Figure 2-6. MRI image illustrating a previously injured BFLH (right limb) and uninjured contralateral BFLH (left limb) Figure 2-7. Architectural characteristics of the injured BFLH Figure 2-8. Conceptual model for the development of neuromuscular inhibition following hamstring strain injury Figure 4-1. The Nordic hamstring exercise Figure 4-2. The relationship between eccentric knee flexor strength imbalances and probability of future hamstring strain injury Figure 5-1. Mean percentage change in fmri T2 relaxation times after running for each hamstring muscle in previously injured (Inj) and uninjured (Uninj) limbs Figure 5-2. A. Parametric map of transverse (T2) relaxation times for the previously injured and uninjured contralateral limbs of a single participant Figure 5-3. Mean CSAs (cm 2 ) of each hamstring muscle for both the previously injured (Inj) and uninjured (Uninj) contralateral limbs Figure 5-4. Percentage change in fmri T2 relaxation times of each hamstring muscle in the uninjured limb Figure 6-1. The 10 examined exercises Figure 6-2. Biceps femoris (BF) to medial hamstring (MH) normalised EMG (nemg) relationship for the (a) concentric and (b) eccentric phases of each exercise Figure 6-3. Percentage change in fmri T2 relaxation times of each hamstring muscle following the 45 hip extension exercise Figure 6-4. Percentage change in fmri T2 relaxation times of each hamstring muscle following the Nordic hamstring exercise Figure 6-5. Ratio of biceps femoris long head (BFLH) to semitendinosus (ST) (BFLH/ST) percentage change in fmri T2 relaxation times following the 45 hip extension and the Nordic hamstring exercise x

12 Figure 7-1. (a) The 45 0 hip extension (HE) exercise and (b) the Nordic hamstring exercise (NHE) Figure 7-2. T1-weighted image (transverse relaxation time = 750ms; echo time = 12ms, slice thickness = 10mm), depicting the regions of interest for each hamstring muscle Figure 7-3. Biceps femoris long head (BFLH) fascicle lengths before (baseline), during (mid-training) and after (post-training) the intervention period Figure 7-4. Percentage change in volume (cm 3 ) for each hamstring muscle after the intervention Figure 7-5. Percentage change in anatomical cross sectional area (ACSA) (cm 2 ) for each hamstring muscle after the intervention Figure 7-6. Eccentric knee flexor force measured during the Nordic strength test before (baseline) and after (post-training) the intervention period Figure 7-7. Hip extension three-repetition maximum (3RM) before (baseline) and after (post xi

13 List of Tables Table 2-1. Morphometric and architectural data of the hamstring muscles... 9 Table 4-1. Pre-season Nordic hamstring exercise force variables for each level of competition and player position Table 4-2. Pre-season Nordic hamstring exercise force variables for hamstring strain injured and uninjured rugby union players Table 4-3. Univariate relative risk of suffering a future hamstring strain injury Table 4-4. Multivariate logistic regression model using prior hamstring strain injury (HSI) and between-limb imbalance in eccentric knee flexor strength Table 5-1. Hamstring strain injury details for all participants (n=10) Table 6-1. Mean normalised EMG (nemg) amplitudes for the biceps femoris (BF) and medial hamstring (MH) muscles during the concentric and eccentric phases of 10 hamstring strengthening exercises Table 7-1. Training program variables Table 7-2. Participant characteristics xii

14 List of Abbreviations BFLH BFSH CI CSA EMG fmri HSI kg MRI MTU MVIC N NHE ROI RR SD SE SM ST T2 VA biceps femoris long head biceps femoris short head confidence interval cross-sectional area electromyography functional magnetic resonance imaging hamstring strain injury kilograms of body mass magnetic resonance imaging musculotendinous unit maximal voluntary isometric contraction newtons of force Nordic hamstring exercise region of interest risk ratio standard deviation standard error semimembranosus semitendinosus transverse relaxation time voluntary activation xiii

15

16 Chapter 1: INTRODUCTION Hamstring strain injury (HSI) is characterised by a partial to complete disruption of muscle fibres in the hamstring muscle group and it represents the most common injury in sports involving high-speed running (Brooks, Fuller, Kemp, & Reddin, 2005c, 2006; Drezner, Ulager, & Sennet, 2005; Ekstrand, Hagglund, & Walden, 2011b; Orchard, James, & Portus, 2006; Woods et al., 2004). Comparatively high rates of HSI recurrence (Brooks, et al., 2006; Heiser, Weber, Sullivan, Clare, & Jacobs, 1984; Orchard & Seward, 2010; Woods, et al., 2004) are perhaps the most concerning aspect of these injuries, as re-injuries are typically more severe (Brooks, et al., 2006; Ekstrand, et al., 2011b; Koulouris, Connell, Brukner, & Schneider-Kolsky, 2007) and demand greater periods of convalescence (Koulouris, et al., 2007) than first-time insults. Despite significant efforts in recent years to reduce the burden of HSI in sport, longitudinal data from the elite Australian football league (Orchard & Seward, 2002; Seward, Orchard, Hazard, & Collinson, 1993), professional rugby union (Brooks, Fuller, Kemp, & Reddin, 2005a, 2005b; Brooks, et al., 2006), elite level soccer (Hagglund, Walden, & Ekstrand, 2009; Woods, et al., 2004) and athletics (Opar et al., 2013), suggest that HSI rates have not declined over several years. This is particularly concerning in light of evidence that other common injuries such as ankle sprains in soccer (Ekstrand & Gillquist, 1983) and posterior cruciate ligament injuries in Australian football (Orchard & Seward, 2010), have shown reduced injury rates following the implementation of preventive measures. These data suggest that the effectiveness of conventional HSI prevention and rehabilitation practices might be overstated and that there is more to be learnt about the mechanisms underpinning injury occurrence. Chapter 1: INTRODUCTION 1

17 Significant time lost from training and competition, from first time and recurrent HSIs (Brooks, et al., 2006; Orchard & Seward, 2011; Woods, et al., 2004), is not only challenging for the athlete, but also imparts a significant financial burden on professional sporting clubs. For example, HSIs were estimated to cost English premier league clubs 74.4 million in wages paid to unavailable players during the seasons (Woods, Hawkins, Hulse, & Hodson, 2002). In the elite Australian football league, HSI cost clubs AUD$1.5m in lost wages throughout the 2009 competitive season (Opar, Williams, & Shield, 2012) and between 2002 and 2012, the average yearly cost of HSIs per Australian football club increased by 71% (Hickey, Shield, Williams, & Opar, 2013). Over the same time period the average financial cost of a single HSI increased by 56% from AUD$ in 2003 to AUD $ in 2012, despite little change in the rate of injuries during that period (Hickey, et al., 2013). Given the performance and economic based implications of HSI, further exploration of the mechanisms responsible for this injury is warranted. This review aims to provide the reader with a summary of HSI literature through an overview of HSI prevalence in sport, a description of anatomical and morphological factors which predispose this muscle group to injury, and a discussion of the proposed causes and risk factors for HSI. It will conclude by proposing a conceptual framework for the role of neuromuscular inhibition in HSI recurrence and suggest some novel direction for future research. Chapter 1: INTRODUCTION 2

18 Chapter 2: LITERATURE REVIEW 2.1 DEFINITION OF HAMSTRING STRAIN INJURY HSI is characterised by a partial to complete disruption of muscle fibres in the hamstring muscle group and is typically associated with the instantaneous onset of posterior thigh pain (Heiderscheit, Sherry, Silder, Chumanov, & Thelen, 2010). HSI or rupture occurs most often at the proximal aponeurosis of the BFLH (Askling, Tengvar, Saartok, & Thorstensson, 2007; De Smet & Best, 2000; Garrett, 1990) but can also affect the proximal bony origin, musculotendinous junction (MTJ), muscle belly or the point of distal bony insertion of any of the hamstring muscles (Agre, 1985). The severity of injury appears to be dependent on the location, the magnitude of force applied, the physical integrity of the muscle at the time of injury (Agre, 1985) and the magnitude of strain experienced (Askling, Saartok, & Thorstensson, 2006). The American Medical Association has identified three grades of severity (Craig, 1973): grade 1 injuries involve a minor tear of only a few muscle fibres with minimal loss of function; grade 2 injuries are more severe partial tears of the muscle-tendon unit (MTU) signified by some loss of function; and grade 3 injuries involve a complete rupture of the MTU and severe functional deficits. 2.2 INCIDENCE OF HAMSTRING STRAIN INJURY IN SPORT HSI is common in a range of sports including athletics (Bennell & Crossley, 1996; D'Souza, 1994; Drezner, et al., 2005; Opar, Drezner, et al., 2013), American football (Elliott, Zarins, Powell, & Kenyon, 2011; Feeley et al., 2008), Australian football (Gabbe, Finch, Wajswelner, & Bennell, 2002; Orchard & Seward, 2002, 2010; Seward, Orchard, Hazard, & Collinson, 1993b), rugby union (Brooks, et al., 2005a, 2005b) and soccer (Ekstrand, 3

19 Hagglund, & Walden, 2010, 2011a; Woods, et al., 2002; Woods, et al., 2004). Within athletics, HSI accounts for 75% of all lower limb strains and represents 24.1% of all injures in the sport (Opar, Drezner, et al., 2013). In the American National Football League, HSI represents 11.6% of all injuries and is the most severe injury subtype with, on average, 8.3 days lost per incident (Feeley, et al., 2008). HSI is the most common injury in the elite Australian football league (Gabbe, et al., 2002; Orchard & Seward, 2002, 2010; Seward, et al., 1993). On average, six new injuries per club per season over the past 10 years have been attributed to HSI, representing 16% of all injuries sustained in the sport (Orchard & Seward, 2010). HSI accounted for 28% of all AFL games missed in the 2009 season, which is far more than quadriceps strains (7.8%) or groin strains (17.9%) (Orchard & Seward, 2010). Large-scale studies in English professional rugby union (Brooks, et al., 2005a, 2005b) have reported that hamstring strains account for 6 15% of all injuries suffered during match play and result in, on average, 17 days of absence from training and/or playing. This contrasts with 12 days lost following quadriceps or hip flexor strain and 10 days lost as a result of hip adductor strain injury (Brooks, et al., 2005c). HSIs are the single most common injury in soccer and account for 12% of all injuries in this game (Ekstrand, et al., 2010, 2011a; Woods, et al., 2002; Woods, et al., 2004). This is more than quadriceps strains (7%), groin injuries (9%) and ankle sprains (7%) (Ekstrand, et al., 2011a) and equates roughly to a squad of 25 incurring 7 HSIs each season. Further, HSI represents the most common type of severe injury (those resulting in >28 days of absence from training and playing) (Ekstrand, et al., 2011a). 2.3 RECURRENCE OF HAMSTRING STRAIN INJURY IN SPORT Arguably the most concerning aspect of HSIs is their tendency to re-occur (Brooks, et al., 2006; Croisier, 2004; Heiser, et al., 1984; Orchard & Seward, 2002, 2010, 2011; Seward, et al., 1993; Woods, et al., 2004), often with greater severity than the original insult (Brooks, et 4

20 al., 2006) (Brooks, et al., 2006; Ekstrand, et al., 2011b). Two decades ago, Seward et al. (Seward, et al., 1993) reported that 34% of all HSIs in professional Australian football, rugby union and rugby league players were recurrences of previous injuries. Heiser and colleagues (Heiser, et al., 1984) noted similar recurrence rates in American football (31.7%) close to 30 years ago. More recent epidemiological studies suggest that recurrent HSI is still an issue as 27% of all HSIs across a 20-year period in elite Australian football were recurrent injuries (Orchard & Seward, 2011). In elite rugby union, 21.3% of all HSIs were reported to be recurrences of previous injuries (Brooks, et al., 2006). Interestingly, HSI recurrence rates in Australian football have declined moderately in recent years with the reduction attributed to more cautious treatment and increased convalescence rather than to improved rehabilitation practices (Orchard & Seward, 2010). However, HSI recurrence rates remain significantly higher than groin strain recurrences (23.3%) and quadriceps strain recurrences (17%) (Orchard & Seward, 2011) in elite Australian footballers. High recurrence rates across a number of sports suggest that conventional rehabilitation practices are not fully addressing the underlying risk factors leading to HSI or the maladaptations associated with the previous insult. 2.4 HAMSTRING ANATOMY The hamstring muscle compartment collectively describes a group of three muscles located on the posterior thigh semimembranosus (SM), semitendinosus (ST), and biceps femoris (BF); BF is further divided into a long head (BFLH) and a short head (BFSH). With the exception of BFSH, the hamstrings are biarticular muscles because they cross both the posterior aspects of the knee and hip joints. This organisation allows the biarticular hamstrings to perform flexion at the knee and extension of the hip during concentric contraction. Although these muscles share relatively common actions, they exhibit significant 5

21 differences in morphology, architecture and function (Markee et al., 1955). A thorough understanding of the anatomical and architectural characteristics of these muscles, as well as differences in their patterns of innervation, is necessary to comprehend the potential impact hamstring structure may play in HSI occurrence. Semitendinosus ST constitutes one-half of the medial hamstrings and morphologically is considered a single muscle, although it may be termed digastric given the presence of a tendinous inscription dividing the muscle belly into superior and inferior portions (Markee, et al., 1955). Proximally, its long fibres originate from three distinct locations: the posteromedial portion of the ischial tuberosity; the medial border of the proximal BFLH tendon; and an aponeurosis which arises from the proximal tendon of the BFLH (Woodley & Mercer, 2005). Distally, fibres converge into a tendon which then inserts onto the proximal medial surface of the tibia (Marieb & Hoehn, 2007). ST has a dual innervation (a result of its digastric structure) with each nerve originating from the tibial portion of the sciatic nerve (Markee, et al., 1955). The uppermost nerve arises from the tibial division almost opposite the ischial tuberosity and innervates motor units proximal to the tendinous inscription. The second nerve comes from the tibial division at the level of the upper and middle thirds of the thigh and supplies the inferior portion of the muscle. Architecturally, ST demonstrates characteristics of a strap-like muscle with long and thin fibres, which are the longest of all the hamstring muscles (Table 2-1) (Woodley & Mercer, 2005). Semimembranosus SM is the second of the medial hamstrings and is a single muscle with a bipennate architectural arrangement (Markee, et al., 1955). Its proximal tendon originates from the 6

22 lateral facet of the posterior ischial tuberosity (Marieb & Hoehn, 2007). Distally, the tendon has multiple attachment sites; one portion inserts on the posterior medial surface of the medial tibial condyle and a second segment expands to the medial condyle of the femur, the capsule of the knee joint and proximally onto the medial collateral ligament (Markee, et al., 1955). SM is innervated by a single nerve branch arising from the tibial portion of the sciatic nerve; it shares the same nerve as that which supplies the distal section of ST. This nerve travels inferiorly, dispersing into five branches in succession, the first of which lies near the proximal origin and last of which lies near the distal aponeurosis (Markee, et al., 1955). Architecturally, SM displays the greatest physiological cross-sectional area (PCSA) of all hamstring muscles (Table 2-1) (Woodley & Mercer, 2005). Biceps femoris BF is a two-segment muscle with both a short (BFSH) and a long head (BFLH). BFSH is the only single joint muscle of the hamstring group, acting only to flex the knee during concentric contraction. Its fibres originate from three locations: the lateral lip of the linear aspera; the upper two-thirds of the lateral supracondylar line; and the lateral intermuscular septum (Marieb & Hoehn, 2007). Distally, its parallel fibres converge on a common tendon with that of BFLH, which inserts onto the head of fibula and lateral condyle of the femur (Marieb & Hoehn, 2007). BFSH is innervated by nerve branches originating from the sciatic nerve or common peroneal nerves, depending on anatomical variations (Markee, et al., 1955; Woodley & Mercer, 2005). Structurally, BFSH has the smallest PCSA relative to all other hamstring muscles (Table 2-1) (Woodley & Mercer, 2005), although it comprises the longest mean fascicle lengths. 7

23 BFLH, a bipennate and biarticular muscle (Marieb & Hoehn, 2007), originates partly from a common tendon with ST, which originates from the medial portion of the superior half of the ischial tuberosity (Woodley & Mercer, 2005). BFLH also attaches directly to the sacrotuberous ligament. The long proximal tendon passes distally until it forms a narrow aponeurotic musculotendinous junction; BFLH fibres emerge on the lateral border of this tendon along with some ST fascicles (Woodley & Mercer, 2005). Distally, the large muscle belly passes inferolaterally to converge onto a common tendon with BFSH, inserting on the superior extremity of the head of the fibula and the lateral femoral condyle (Markee, et al., 1955; Woodley & Mercer, 2005). BFLH is innervated by the tibial division of the sciatic nerve, which divides into an upper and lower portion and supplies the deep and superficial motor units, respectively (Markee, et al., 1955). Architecturally, BFLH displays the second greatest PCSA (behind SM) of all the hamstring muscles (Table 2-1) (Woodley & Mercer, 2005). 8

24 Table 2-1. Morphometric and architectural data of the hamstring muscles (SM, Semimembranosus; ST, Semitendinosus; BFlh, Biceps femoris long head; BFsh, Biceps femoris short head). Adapted from Woodley and Mercer (2005). 9

25

26 2.5 MECHANISM(S) OF INJURY HSI occurs most often as a result of trauma to the MTU (Craig, 1973). This may be in the form of forceful eccentric contraction or from excessive stretch of the MTU (Agre, 1985). A number of potential mechanisms have been proposed for HSIs however, there is no clear consensus in the literature. High degrees of muscular strain, defined as the change in length of the muscle during movement (Garrett, 1990), and high degrees of muscular stress, quantified as force per unit of cross-sectional area (CSA) (Garrett, 1990) are both present during forceful eccentric contractions. However, it is unclear which of these is the major contributor to HSI (Garrett, 1990). Garrett and colleagues (1987) examined the effects of strain on in situ animal muscles that were stimulated maximally and submaximally and then stretched to the point of MTU strain-induced failure. The resultant load deformation curves demonstrated that, regardless of the stimulation level, all muscles failed at the same MTU length (Garrett, Safran, Seaber, Glisson, & Ribbeck, 1987). These data imply that strain, not stress, is the primary determinant of strain injury occurrence. However, various authors have noted that HSI is most likely to occur during the terminal swing phase of running (Brooks, et al., 2005c; Ekstrand, et al., 2011a; Woods, et al., 2002; Yu et al., 2008) where stress is high (Schache, Dorn, Blanch, Brown, & Pandy, 2012) but strain is moderate (Thelen, Chumanov, Best, Swanson, & Heiderscheit, 2005). Clearly, there is some degree of both stress and strain whenever muscles are active and it is likely that with lower levels of stress, a higher level of strain is needed to cause rupture, while at higher levels of stress comparatively lower levels of strain will bring about injury (Opar, et al., 2012). 11

27 2.5.1 Hamstring function during high-speed running and propensity for injury The running gait cycle for each leg can be divided into two major phases: a stance phase comprising the initial contact, mid-stance and take-off; and a swing phase comprising initial, mid and terminal-swing of each leg (Figure 2-1) (Subotnick, 1985). During the mid and terminal-swing phases, the hamstrings primarily contract eccentrically to decelerate hip flexion and knee extension (Montgomery, Pink, & Perry, 1994). With increasing running speed, the duration of the terminal swing phase is reduced (Agre, 1985), increasing the angular velocity at both the hip and knee which requires increased torque generation by the hamstrings to control the movement at each joint (Agre, 1985). The biomechanical demands of high-speed running may explain the propensity of HSI to occur during sprint running compared with slow-speed running (jogging). Figure 2-1. Phases of the running gait cycle To date, there have only been two case-studies which captured the time-occurrence and biomechanical response to an HSI during running. Each of these studies concluded that injury occurred while the hamstrings were actively lengthening during the terminal-swing phase of the running cycle (Heiderscheit et al., 2005; Schache, Wrigley, Baker, & Pandy, 2009). This might be explained by evidence that the biarticular SM, ST and BFLH all produce peak force 12

28 while lengthening during the late-swing phase of running (Chumanov, Heiderscheit, & Thelen, 2007, 2011; Schache, et al., 2012; Thelen et al., 2005). Schache and colleagues (2012) suggest that during this phase, the BFLH exhibits the greatest peak strain, ST exhibits the fastest lengthening velocity, and SM displays the greatest peak force, produces the most power and completes the greatest amount of work. Given the reported pre-eminence of muscle strain in HSI (Opar, et al., 2012), the differences in peak muscle length during terminal-swing has received significant interest (Chumanov, et al., 2007, 2011; Schache, et al., 2012; Thelen, Chumanov, Hoerth, et al., 2005); BFLH increases in length by 9.5% relative to the MTU length in the anatomical position in contrast with 7.4% lengthening by SM and 8.1% by ST (Chumanov, et al., 2007, 2011; Schache, et al., 2012; Thelen, Chumanov, Hoerth, et al., 2005). That BFLH experiences the greatest strain magnitude may explain its propensity for injury relative to the other hamstring (Connell et al., 2004; Koulouris, et al., 2007), however, whether these small differences can account for discrepant injury rates is yet to be determined. 2.6 PROPOSED RISK FACTORS FOR HAMSTRING STRAIN INJURY Several unalterable and alterable risk factors for HSI have been proposed in the literature. Unalterable risk factors include previous HSI (Arnason et al., 2004; Bennell et al., 1998; Gabbe, Bennell, Finch, Wajswelner, & Orchard, 2006; Hagglund, Walden, & Ekstrand, 2006; Orchard, 2001; Verrall, Slavotinek, Barnes, Fon, & Spriggins, 2001), increasing age (Arnason, et al., 2004; Gabbe, Bennell, & Finch, 2006; Verrall, et al., 2001) and ethnicity (Brooks, et al., 2006; Verrall, et al., 2001; Woods, et al., 2004). Alterable risk factors include muscular weakness (Aagaard, Simonsen, Magnusson, Larsson, & Dyhre-Poulsen, 1998; Arnason, Andersen, Holme, Engebretsen, & Bahr, 2008; Askling, Karlsson, & Thorstensson, 2003; Brockett, Morgan, & Proske, 2001; Burkett, 1970; Croisier, Forthomme, Namurois, 13

29 Vanderthommen, & Crielaard, 2002; Croisier, Ganteaume, Binet, Genty, & Ferret, 2008c; Friden & Lieber, 1992; Gabbe, Bennell, Finch, et al., 2006; Garrett, et al., 1987; Lee, Reid, Elliott, & Lloyd, 2009; Mair, Seaber, Glisson, & Garrett, 1996; Orchard, Marsden, Lord, & Garlick, 1997b; Petersen, Thorborg, Nielsen, Budtz-Jørgensen, & Hölmich, 2011; Sugiura, Saito, Sakuraba, Sakuma, & Suzuki, 2008; Yamamoto, 1993; Yeung, Suen, & Yeung, 2009), poor flexibility (Bradley & Portas, 2007; Henderson, Barnes, & Portas, 2009; McHugh et al., 1999; Witvrouw, Danneels, Asselman, D'Have, & Cambier, 2003), fatigue (Brooks, et al., 2006; Heiser, et al., 1984; Mair, et al., 1996; Opar, et al., 2012; Woods, et al., 2004), poor lumbopelvic control (Chumanov, et al., 2007; Sherry & Best, 2004) and short BFLH fascicle lengths (Timmins, Bourne, et al., 2015). A comprehensive understanding of each of these risk factors is necessary for identifying individuals most susceptible to HSI and re-injury Unalterable risk factors Previous injury Previous HSI appears to be the best independent predictor of future HSI (Bennell, et al., 1998a; Gabbe, Bennell, Finch, et al., 2006; Hagglund, et al., 2006; Orchard, 2001; Verrall, et al., 2001). Prospective studies in elite soccer players have found that players who sustained a HSI in the previous season were up to 11.6 times more likely to experience a recurrent injury in the following season (Arnason, et al., 2004; Hagglund, et al., 2006). Similarly, elite (Gabbe, Bennell, Finch, et al., 2006; Orchard, 2001) and community-level (Verrall, Slavotinek, Barnes, & Fon, 2003) Australian footballers with a history of HSI are at significantly greater risk of future HSI. The mechanism(s) by which previous HSI increases future injury risk is unclear, but is likely to reflect a number of maladaptations following HSI (Opar, et al., 2012) or the persistence of pre-existing risk factors (Bradley & Portas, 2007; Brockett, Morgan, & Proske, 2004; Croisier, et al., 2002; Jonhagen, Nemeth, & Eriksson, 14

30 1994; Silder, Heiderscheit, Thelen, Enright, & Tuite, 2008; Silder, Reeder, & Thelen, 2010; Witvrouw, et al., 2003). Because of a lack of prospective data on these maladaptations, it is unknown whether these changes result from the injury or whether they were present before, and were possibly the cause of, the original insult. Age Increasing age has been identified as a significant predictor of future HSI in Australian football (Gabbe, Bennell, & Finch, 2006; Opar et al., 2014) and soccer players (Timmins, Bourne, et al., 2015; Verrall, et al., 2001). The risk of HSI appears to increase by 10% per year in elite Icelandic soccer players (Arnason, et al., 2004). Older elite Australian football players (>25 years) are more than four times more likely to experience an injury than are younger players (<20 years) (Gabbe, Bennell, & Finch, 2006). Similar findings have been found in community-level AFL, with older players (>23 years) up to four times more likely to incur an HSI than younger players (<23 years) (Gabbe, Finch, Bennell, & Wajswelner, 2005). Relatively little is known about the mechanism(s) responsible for the age-related changes that heighten HSI risk. However, recently we (Opar, et al., 2014; Timmins, Bourne, et al., 2015) have provided evidence to suggest that this risk might be modulated by one or more modifiable risk factors. For example, older (>23 years) elite Australian footballers are only at an elevated risk of HSI if the athlete also has low levels of eccentric knee flexor strength. Further, in professional soccer players, higher levels of eccentric knee flexor strength and longer BFLH fascicles appear to offset the elevated risk of HSI associated with increasing age (Timmins, Bourne, et al., 2015). Others have suggested that age-related reductions in hip flexor flexibility and increased body mass index (BMI) (Gabbe, Bennell, & Finch, 2006), or hypertrophy of the lumbosacral ligament (Orchard et al., 2004) may explain the increased 15

31 susceptibility to HSI. Further work is required to advance our knowledge about how increasing age influences HSI risk Alterable risk factors Alterable risk factors refer to those functional characteristics that can be modified with training. For the purposes of this review these will include: muscular weakness or imbalances in strength (Aagaard, et al., 1998; Arnason, et al., 2008; Askling, et al., 2003; Brockett, et al., 2001; Burkett, 1970; Croisier, et al., 2002; Croisier, et al., 2008c; Friden & Lieber, 1992; Gabbe, Bennell, Finch, et al., 2006; Garrett, et al., 1987; Lee, et al., 2009; Mair, et al., 1996; Orchard, et al., 1997b; Petersen, et al., 2011; Sugiura, et al., 2008; Yamamoto, 1993; Yeung, et al., 2009); poor flexibility (Bradley & Portas, 2007; Henderson, et al., 2009; McHugh, et al., 1999; Witvrouw, et al., 2003); fatigue (Brooks, et al., 2006; Heiser, et al., 1984; Mair, et al., 1996; Opar, et al., 2012; Woods, et al., 2004); and poor lumbopelvic control (Chumanov, et al., 2007; Sherry & Best, 2004). Weakness and strength imbalances For the purposes of this review, strength imbalances are between-leg asymmetries in knee flexor strength and/or a low ratio of knee flexor to knee extensor strength, otherwise known as the hamstring-to-quadriceps (H:Q) ratio. Early in situ studies demonstrated that maximally stimulated muscles can tolerate more stress than those stimulated submaximally (Garrett, et al., 1987; Mair, et al., 1996). A heightened ability to tolerate stress enables these muscles to absorb more energy before strain injury occurs (Mair, et al., 1996). Assuming that these in situ results reflect in vivo function, one can 16

32 infer that stronger muscles are more resistant to strain injury than are weaker muscles (Garrett, et al., 1987). Between-limb strength imbalances The concept of a relationship between leg-to-leg (bilateral) strength asymmetries and injury risk is a logical assumption given the evidence suggesting that weakness may predispose to strain injury. Between-limb strength imbalances of the knee flexors have been associated with an increased risk of HSI in several sports (Burkett, 1970; Croisier, et al., 2002; Croisier, et al., 2008c; Orchard, Marsden, Lord, & Garlick, 1997a). Although the underlying mechanism remains unknown, it has been proposed that biomechanical alterations arise from these imbalances and may increase the strain experienced by the weaker hamstrings during running, thereby increasing HSI risk (Croisier, et al., 2008c). The largest study examining bilateral strength asymmetries, by Croisier and colleagues (2008c), used isokinetic dynamometry to determine whether asymmetry could predict future HSI. The study tested 462 professional soccer players during the preseason period to assess hamstring strength asymmetry using a standardised concentric and eccentric isokinetic protocol. At the end of the competitive season, athletes who did not present with strength imbalances in the pre-season or those who underwent an intervention to correct imbalances and then completed a subsequent re-test, displayed similarly low incidence rates of HSI (Croisier, et al., 2008c). In comparison, those with detected asymmetry who chose not to undergo the intervention were up to four times more likely to suffer an HSI (Croisier, et al., 2008c). However, it should be acknowledged that only severe injuries (> 30 days to return to sport) were reported in this study and epidemiological data (Ekstrand, et al, 2011) suggests that the average return to play time from HSI in professional football is significantly less than 17

33 this and severe injuries (> 28 days) constitute only a very small percentage (~10%) of total injuries in this cohort. These data suggest the possibility that a large number of HSIs were not reported by Croisier and colleagues (2008c) and renders it impossible to determine if athletes with less severe injuries were also at a greater risk of injury is they presented with isokinetic strength imbalances. It should also be acknowledged that isokinetic testing was conducted by multiple clinicians at various sites, using different equipment and different cut-points to determine pass or fail (Croisier, et al, 2008c) and the reliability of such measures is unclear. Given these limitations, the results of Croisier and colleagues study (2008c) should be interpreted with caution. Nevertheless, prospective studies of elite-level sprinters with no history of HSI have also demonstrated that athletes with isokinetically derived knee flexor asymmetries were more likely to sustain a strain injury in the following months (Sugiura, et al., 2008; Yamamoto, 1993). Similarly, isokinetic strength testing of Australian footballers showed that players who exhibited unilateral hamstring muscle weakness in the preseason were significantly more likely to experience HSI throughout the following season (Orchard, et al., 1997b). Although little is known about the degree of knee flexor asymmetry required to elevate HSI risk, various guidelines have been suggested. Early research found that asymmetries of more than 10% are associated with an increased risk of HSI in American footballers (Burkett, 1970) and track and field athletes (Heiser, et al., 1984). More recent evidence has shown that Australian footballers with asymmetries of 8% or greater (Orchard, et al., 1997b) and soccer players with imbalances exceeding 15% (Croisier, et al., 2008c) are at increased risk of HSI. It should be noted that some prospective studies have found no relationship between isokinetic strength imbalances and hamstring injury risk (Bennell, et al., 1998a; Yeung, et al., 2009). However, the size and therefore statistical power of these studies is almost always inadequate to rule out a link between strength imbalance and risk (Bahr & Holme, 2003). 18

34 Future research should continue to explore the risk factors for HSI with an emphasis on defining the level of knee flexor strength imbalance associated with an increased risk of HSI across different sports. The strength asymmetries among other muscles which act during terminal swing, for example the hip flexors, should also be explored because these are known to affect hamstring mechanics during running (Lee, et al., 2009). Implementation of practical field-based measures of between-limb hamstring strength may also help to reduce injury rates in sport. 19

35 H:Q ratio Throughout the terminal swing phase of running (Figure 2.3), the hamstrings act primarily as a brake by rapidly decelerating the extending knee and flexing hip. This exposes the hamstrings to moderate degrees of strain and high degrees of stress, which may increase their susceptibility to injury (Garrett, 1990). The H:Q ratio is calculated as the peak (concentric or eccentric) hamstring strength divided by peak concentric quadriceps stength (Aagaard, et al., 1998). Theoretically, an individual with relatively stronger knee extensors and comparatively weaker knee flexors may have a lesser ability to overcome the inertia imparted on the shank by the quadriceps during the swing phase. Initially, the H:Q ratio was measured using concentric knee flexor and concentric knee extensor strength and this is now termed the conventonal hamstring:quadriceps ratio (H:Qconv) (Burkett, 1970; Orchard, et al., 1997a). However, this method has been criticised for disregarding the eccentrically biased role of the hamstrings during high-speed running. As a result, a functional ratio involving eccentric hamstring to concentric quadriceps strength (H:Qfunc) has been proposed (Aagaard, et al., 1998) and popularised (Croisier, et al., 2008c; Gabbe, Bennell, Finch, et al., 2006; Sugiura, et al., 2008; Yeung, et al., 2009). Early small-scale research examining the H:Qconv and its relationship with HSI risk in Australian footballers found that players with H:Qconv ratios of <0.61 were at significantly increased risk of sustaining an HSI in the following competitive season (Orchard, et al., 1997a). Similar studies exploring the H:Qconv ratios in American footballers concluded that those with a ratio of <0.50 were more likely to experience an HSI during the season (Heiser, et al., 1984). The most statistically powerful research available (n=462) found that low H:Qfunc ratios (< ) and low H:Qconv ratios (< ) were associated with a significantly increased incidence of HSI (Croisier, et al., 2008c). However, there are 20

36 conflicting data, and several authors have reported no relationship between H:Q ratios and increased HSI risk (Yeung, et al., 2009; Bennell, et al., 1998). Yeung and colleauges (2009), for example, found no significant relationship between either H:Qconv or H:Qfunc and subsequent HSI in sprinters. Bennell and colleagues (Bennell, et al., 1998) also reported no significant association between either the H:Qconv or the H:Qfunc ratio and future HSI in Australian footballers. However, these and many studies investigating H:Q ratios have used relatively small sample sizes and different methodologies. Bahr and colleagues (2003) argue that this presents an obvious limitation and that sample sizes of >300 are necessary to accurately detect small-sized associations between H:Q ratios and HSI risk. This must be considered when interpreting the current literature and future research should explore these relationships using larger-scale prospective studies. In prospective studies which examine eccentric hamstring strength and associated strength ratios as a risk factor for future HSI, isokinetic dynamometry has been the chosen strength testing methodology (Bennell et al., 1998; Croisier, Ganteaume, Binet, Genty, & Ferret, 2008b; Sugiura, Saito, Sakuraba, Sakuma, & Suzuki, 2008; Yeung, et al., 2009). Whilst isokinetic dynamometry is considered the gold standard tool for assessing eccentric hamstring strength, its wide spread application is limited due to the device being largely inaccessible and expensive to purchase. Further to this, the time taken to complete an assessment of an individual athlete (up to 20 minutes) normally at an off-site location is often prohibitive, particularly in elite sporting environments (Opar, Piatkowski, Williams, & Shield, 2013a). Our group has recently developed a field testing device for the assessment of eccentric hamstring strength to overcome the limitations of isokinetic dynamometry (Opar, Piatkowski, et al., 2013a). This device measures eccentric knee flexor force during the performance of the Nordic hamstring exercise. We recently employed this device in a large- 21

37 scale prospective study conducted in the elite AFL (n=210) during the 2013 competitive season (Opar, et al., 2014). Footballers were tested during the pre-season and at three timepoints throughout the season and injury data were collected prospectively. Results demonstrated that limbs that went on to sustain a HSI were significantly weaker than the limbs of uninjured athletes at the start and end of preseason, and players with a two-limb average eccentric strength lower than 256 N at the start of preseason or 279 N at the end of preseason were at 2.7 and 4.3 fold greater risk of sustaining an HSI compared to players above these thresholds. However, the protective effect of extra strength appeared to diminish at around ~ N. This suggests that there might be no relationship between strength and HSI risk in running-based sports that are typically characterised by higher levels of strength, for example rugby union, and this should be a focus of future investigations. Angle of peak knee flexor torque (T-JA) Previously injured hamstrings have been reported to generate peak torque at greater knee joint angles (shorter hamstring muscle lengths) compared with the uninjured contralateral limb (Brockett, et al., 2004). As described in section 2.4.1, if the optimum angle for torque generation is at a shorter relative muscle length, more of that muscle s working range is on the descending limb of the force length relationship and the muscle may be more susceptibile to damage (Morgan, 1990). It is suspected that those who generate peak torque at shorter muscle lengths would be more prone to accumulated microscopic damage in the form of sarcomere over-extension (also known as sarcomere popping ), which may, if it accumulates sufficiently, give rise to macroscopic strain injury (Brockett, et al., 2004). Early retrospective work by Brockett and colleagues (2004) found that athletes with a history of unilateral HSI produced peak knee flexor torque at markedly shorter muscle lengths than 22

38 uninjured athletes. In a group of elite Australian footballers and sub-elite track and field athletes, peak torque was shifted by 12.1 ± 2.7 towards a more flexed knee in the previously injured limb when compared to the uninjured contralateral limb (Brockett, et al., 2004). However, given the retrospective nature of this study, one cannot determine whether the altered angle of peak torque was the cause or result of the injury. A subsequent small and therefore underpowered prospective study of national and international-level sprinters found no relationship between the angle of peak knee flexor torque and the incidence of HSI throughout the competitive season (Yeung, et al., 2009). A larger scale study is required to ascertain whether the angle of peak torque is a significant predictor of HSI risk. Biceps femoris long head (BF LH) fascicle length Recently, Timmins et al., (2014) demonstrated that athletes with a history of HSI display shorter BFLH fascicles coupled with increased pennation angles in their previously injured limb when compared to their uninjured contralateral limb. A subsequent prospective study in elite soccer players demonstrated that short BFLH fascicles (<10.56cm) increased the risk of HSI four-fold (Timmins, Bourne, et al., 2015). Moreover, longer fascicles appeared to off-set the increased risk of HSI associated with increasing age and prior HSI, which were previously considered to be non-modifiable risk factors. Reduced fascicle lengths most likely result from the shedding of in-series sarcomeres (Lieber & Friden, 2000), although it is difficult to determine whether this causes or results from injury. Fewer serial sarcomeres would be expected to shift the muscles force-length relationship to the left, thereby increasing its susceptibility to muscle damage at longer lengths. Fortunately, muscle architecture has the potential to be altered with appropriately structured resistance training. For example, Timmins et al., (2015) recently demonstrated a significant increase in BFLH fascicle length coupled with a reduction in pennation angle in response to six weeks of eccentric-only 23

39 isokinetic knee flexor training. However, few athletes train exclusively with isokinetic devices and further work is needed to determine the adaptability of muscle architecture to different strength training interventions. Flexibility Increased flexibility has long been considered important in injury prevention, despite little prospective evidence (Witvrouw, et al., 2003; Worrell, Smith, & Winegardner, 1994). It has been proposed that high forces produced during eccentric contractions are absorbed by the active contractile components and the passive in-series elements of muscle (Bennell, et al., 1998a). Early research suggested that greater compliance of the passive elastic structures may increase the energy absorption capabilities of the MTU (Worrell, et al., 1994), thereby reducing the loads placed on the active contractile elements of the muscle and mitigating the risk of strain injury (Garrett, et al., 1987; Garrett, 1990). However, large-scale prospective studies using objective measures of flexibility have identified no relationship between flexibility and future HSI in either elite Australian Football players (Gabbe, Bennell, Finch, et al., 2006; Orchard, et al., 1997b) or elite American footballers (Burkett, 1970). However, these studies used only the sit-and-reach test to measure flexibility, which is limited by its inability to identify between-leg differences because both legs are tested simultaneously. Furthermore, the test can be influenced by lumbar spine flexibility and upper to lower limb length ratios. Subsequent studies using the toe-touch test, also found no relationship between poor flexibility and injury risk in elite Australian footballers (Bennell, Tully, & Harvey, 1999). Moreover, there is no correlation between active (Gabbe, et al., 2005) and passive (Arnason, et al., 2004) knee flexor stiffness and HSI incidence in Australian footballers (Gabbe, et al., 2005) or professional soccer players (Arnason, et al., 2004). By contrast, a prospective study by Witvrouw and colleagues (Witvrouw, et al., 2003) found that elite 24

40 soccer players (n=146) who exhibited a passive straight-leg raise of less than 90 were at significantly greater risk of a subsequent HSI. However, the use of the straight-leg raise has been criticised for being more indicative of neural extensibility than hamstring muscle flexibility (Devlin, 2000). Despite an unclear relationship between reduced flexibility and subsequent HSI risk, there is some evidence to suggest that individuals with a history of HSI are less flexible than uninjured people (McHugh, et al., 1999; Witvrouw, et al., 2003). Several studies have demonstrated that sprinters (Jonhagen, et al., 1994), American footballers (Worrell & Perrin, 1992) and elite soccer players (Ekstrand & Gillquist, 1983) with a history of HSI exhibit ongoing deficits in hamstring flexibility. This may be a result of scar tissue formation, a known maladaptation to previous HSI (Silder, et al., 2010), however these studies are again limited by the use of a retrospective study design (Bahr & Holme, 2003). In addition, the subjective measures of flexibility used in these studies lack validity (Opar, et al., 2012), which may affect their interpretation (Bahr & Holme, 2003) Fatigue Fatigue is commonly implicated as a potential cause of HSI (Heiser, et al., 1984; Mair, et al., 1996). Epidemiological evidence shows an increased incidence of HSI during the latter stages of practice or competition (Brooks, et al., 2006; Woods, et al., 2004). Fatigue reduces the capacity of the muscle to generate contractile force, which limits the energy-absorbing ability of the MTU in vivo (Mair, et al., 1996). Muscular fatigue occurs through both central and peripheral mechanisms. Central fatigue can be defined as a reduction of maximum force generating capacity due to a diminished capacity to voluntarily activate a muscle (Shield & Zhou, 2004), whereas peripheral fatigue is a reduction in the maximum capacity of the 25

41 muscle itself caused by a number of changes occurring distal to the neuromuscular junction (Aagaard et al., 2000). In situ animal studies were the first to suggest that the energy-absorbing capability of a muscle is diminished in a fatigued state and that this is related to a reduction in contractile strength (Garrett, et al., 1987; Mair, et al., 1996). While both fatigued and non-fatigued muscles failed at the same relative muscle length (Mair, et al., 1996) the non-fatigued in situ muscle is able to absorb more energy before stretch-induced MTU failure, which suggests that fatigue may limit a muscle s ability to prevent over lengthening. Fatigue has also been shown to induce a number of proprioceptive changes in humans (Allen, Leung, & Proske, 2010; Brown, Child, Donnelly, Saxton, & Day, 1996; Skinner, Wyatt, Hodgdon, Conard, & Barrack, 1986). Allen and colleagues (Allen, et al., 2010) recently demonstrated that after fatiguing exercise of the knee flexors, subjects perceived their knee to be in a more flexed position than it actually was. The authors suggested that alterations at the level of the sensorimotor cortex were responsible (Allen, et al., 2010). In a separate study, hamstring fatigue led to alterations in hamstring muscle kinematics during running (Pinniger, Steele, & Groeller, 2000). Following a fatiguing intermittent running protocol and knee flexor resistance training session, subjects displayed a significant reduction in hip flexion and a concomitant increase in knee extension during the swing phase of high-speed overground running (Pinniger, et al., 2000). Concurrent semg demonstrated an increase in the duration of hamstring electromyographical activity throughout the swing phase of gait (Pinniger, et al., 2000). An inability to sense knee position accurately may cause an underestimation of hamstring length, potentially increasing susceptibility to repeated over-lengthening which may require greater muscle activity to correct. 26

42 Interestingly, hamstring fatigue following intermittent running has been reported to be confined primarily to reductions in eccentric strength (Greig, 2008; Opar, Williams, Porter, & Raj, 2009). This eccentric-specific decline was prevalent without any deficits in concentric knee flexor or extensor strength and exhibited a high degree of individual variability (Greig, 2008; Opar, et al., 2009). These findings are particularly interesting given the propensity for HSI to occur during forceful eccentric contractions (Brockett, et al., 2004; Croisier, 2004; Garrett, 1990). Lumbopelvic stability Poor neuromuscular control of the lumbopelvic region, specifically the uniarticular hip flexors (psoas major and minor and iliacus) is a recently proposed factor in HSI (Sherry & Best, 2004). Shortening of the uniarticular hip flexors during the early swing phase of running has been shown to induce significant stretch on the contralateral hamstrings, most notably the BFLH, through increased anterior pelvic tilt (Chumanov, et al., 2007). Chumanov (Chumanov, et al., 2007) proposed that perturbations in coordination of the uniarticular hip flexors increase strain on the biarticular hamstrings during running, particularly as speeds increase (Chumanov, et al., 2007). To date, research on lumbopelvic stability is limited to biomechanical models of joint kinematics, and further randomised controlled trials are required to establish the validity of this theory. 27

43 2.7 FACTORS UNDERPINNING HIGH RATES OF HAMSTRING STRAIN INJURY RECURRENCE HSI is known to trigger a number of structural and functional maladaptations which may augment the risk of re-injury. Of particular interest to this review is the interrelationship between previous HSI and muscular weakness. Previously injured hamstrings have been reported to possess significant deficits in eccentric strength, in the presence of smaller or absent concentric strength deficits when compared to the uninjured contralateral limb (Figure 2-2) (Croisier, 2004; Croisier, et al., 2002; Dauty, Potiron-Josse, & Rochcongar, 2003; Jonhagen, et al., 1994; Lee, et al., 2009). Croisier and colleagues (2002) were the first to suggest that isokinetically-derived strength imbalances increase the risk of hamstring strain re-injury. The comprehensive testing battery determined that 18 of 26 athletes with a previous HSI who experienced ongoing hamstring pain also exhibited knee flexor strength asymmetries (Croisier, et al., 2002). These strength imbalances were defined as leg to leg differences in knee flexor strength of >15%, concentric knee flexor to concentric knee extensor strength (H:Qconv) <0.47, and eccentric knee flexor to concentric knee extensor strength (H:Qfunc) <0.80 (Croisier, et al., 2002). Of interest was the preferential reduction of eccentric peak torque (~22%) compared with concentric torque (~11%) (Figure 2-2) (Croisier, et al., 2002). Athletes with predetermined strength deficits were prescribed individualised rehabilitation programs to restore a normalised isokinetic strength profile and correction of these abnormalities resulted in a marked reduction in pain and discomfort. Furthermore, none of the athletes sustained a clinically diagnosed HSI in the 12 months following the intervention and all were successfully able to return to their pre-injury levels of competition (Croisier, et al., 2002). 28

44 Figure 2-2. Comparison of knee flexion torque-velocity relationships between previously injured hamstrings, contralateral uninjured hamstrings and reference values from uninjured control subjects. Note the greater deficit in eccentric compared to concentric strength in previously injured hamstrings (Croisier & Crielaard, 2000). While it is possible that strength deficits may be present prior to the original insult (Croisier, et al., 2008c; Sugiura, et al., 2008; Yeung, et al., 2009), evidence suggests that bilateral asymmetries are amplified following HSI. For example, Sugiara et al. (Sugiura, et al., 2008) reported that eccentric strength deficits of ~4.5% were predictive of future HSI in uninjured elite sprinters, while Croisier et al. (2002) identified much larger deficits of 22-24% in previously injured elite soccer players. Compounding the issue further is the observation that these deficits appear to be long-lasting, with several studies identifying deficits months to years post-injury (Croisier, et al., 2002; Dauty, et al., 2003; Jonhagen, et al., 1994; Lee, et al., 2009). 29

45 2.8 MECHANISM(S) FOR CHRONIC STRENGTH DEFICITS FOLLOWING HAMSTRINGS STRAIN INJURY Given several lines of supportive evidence, previously injured hamstrings appear to be considerably weaker during eccentric contractions compared to uninjured hamstrings (Croisier, 2004; Croisier, et al., 2002; Jonhagen, et al., 1994; Lee, et al., 2009). However, the mechanism(s) responsible for greater eccentric strength loss, compared with concentric strength loss following HSI, remains to be determined. Recently it has been proposed that chronic deficits in hamstring voluntary activation, which can be defined as the completeness of skeletal muscle activation during voluntary contractions (Shield & Zhou, 2004), may manifest following HSI as a result of persistent neuromuscular inhibition (Opar, et al., 2012).The role of neuromuscular inhibition after other injuries is well established. For example, substantial and long-lasting deficits in quadriceps maximal activation have been observed following traumatic knee injury (Hurley, 1997; Urbach, Nebelung, Becker, & Awiszus, 2001). Likewise, injury to the ankle joint has been linked to reductions in plantar flexor activation (Hurley, 1997). In addition, experimentally induced joint and muscle pain has been shown to reduce voluntary drive to nearby muscles as well as alter inter-muscular coordination patterns (Diederichsen et al., 2009). The presence of significant eccentric strength deficits combined with smaller or absent losses in concentric strength is suggestive of a severe contraction mode-specific decline in voluntary activation (Croisier & Crielaard, 2000). Previous investigations from this student s Honours project demonstrated significant muscle-specific reductions in activation of previously injured hamstring muscles, relative to uninjured contralateral hamstring muscles during a common rehabilitation exercise (Bourne, D., Williams, Al-Nijjar, & Shield, 2015). Furthermore, evidence of hamstring remodelling 5-23 months post-hsi, specifically, atrophy of the previously injured muscle (most likely via limited activation) and concomitant hypertrophy of synergists (Silder, et al., 2008), suggests a 30

46 chronic de-loading of the previously injured muscle and the existence of compensatory activation strategies (Silder, et al., 2008). 2.9 NEUROMUSCULAR INHIBITION Discrepancies between the force-velocity relationships of isolated, electrically stimulated muscles (in vitro) and voluntarily activated muscle (in vivo) indicate the inability of healthy but untrained individuals, to maximally activate certain muscles during eccentric contractions (Figure 2-3) (Westing, Cresswell, & Thorstensson, 1991; Westing, Seger, Karlson, & Ekblom, 1988). The force-velocity relationship of electrically stimulated human muscle demonstrates that during concentric contractions maximal force generation declines as the rate of shortening increases (Katz, 1939; Westing, et al., 1988), while eccentric contractions are characterised by markedly higher maximal forces that plateau at levels up to 100% greater than peak isometric force (Katz, 1939; Westing, et al., 1991). 31

47 Figure 2-3. The torque/force-velocity relationships of electrically stimulated (red) and voluntarily activated (blue) skeletal muscle. Note the divergence in maximal eccentric force/torque, indicating the existence of a tension-limiting mechanism(s) during volitional eccentric contractions. Although the force-velocity curves of electrically stimulated and voluntarily activated skeletal muscle display similarities in their concentric and isometric portions (Thorstensson, Grimby, & Karlsson, 1976; Westing, et al., 1988), clear differences can be noted in the eccentric portion of these relationships (Edman, Elzinga, & Noble, 1978; Katz, 1939). Specifically, during eccentric contractions voluntarily activated muscle fails to reach the levels of maximal force obtained by isolated muscle (Edman, et al., 1978; Westing, et al., 1988) (Figure 2-3). This evidence suggests that voluntarily activated muscles are not fully activated during active lengthening despite maximal voluntary effort (Westing, et al., 1991). The divergence of the voluntarily activated force-velocity curve from that observed for 32

48 isolated muscle appears to be the result of a tension-limiting mechanism(s) that acts to reduce the extent of force produced during eccentric contractions (Babault, Pousson, Michaut, Ballay, & Hoecke, 2002; Westing, et al., 1991). By limiting the development of excessive force within the MTU, this mechanism may protect the musculoskeletal system from an injury that could result if the muscle was to be fully activated (Babault, et al., 2002; Westing, et al., 1991) Evidence for incomplete activation in maximal voluntary contractions Studies exploring voluntary activation of the knee extensors and flexors through the use of the interpolated twitch technique (Amiridis et al., 1996; Babault, et al., 2002; Beltman, Sargeant, Mechelen, & Haan, 2004; Westing, et al., 1991) and semg (Aagaard, et al., 2000; Amiridis, et al., 1996; Onishi et al., 2002; Ono, Higashihara, & Fukubayashi, 2011; Ono, Okuwaki, & Fukubayashi, 2010; Westing, et al., 1991) have demonstrated incomplete activation during maximal eccentric (Kellis & Baltzopoulos, 1998) and slow concentric (Aagaard, et al., 2000) contractions. Twitch interpolation is the most commonly used method for assessing the completeness of skeletal muscle activation (Belanger & McComas, 1981; Shield & Zhou, 2004). Twitch interpolation typically involves the application of a supramaximal electrical stimulus to an active muscle(s), during voluntary isometric and/or dynamic contractions (Shield & Zhou, 2004). While some studies suggest that maximal voluntary activation is possible during concentric contractions of muscles such as the biceps brachii (Gandevia, 1998), there is evidence for deficits in maximal activation during eccentric actions of the knee extensors (Babault, et al., 2002; Beltman, et al., 2004). For instance, Babault (2002) compared knee extensor activity during maximal isometric, concentric and eccentric isokinetic contractions 33

49 and reported voluntary activation levels of 95.2% during isometric contractions, in contrast with 88.3% and 89.7% for maximal eccentric and concentric contractions, respectively. A subsequent study employing a similar protocol reported even greater activation deficits during eccentric contractions (79%), compared with concentric (92%) and isometric (93%) contractions (Beltman, et al., 2004). These results suggest that activation of the knee extensors is reduced primarily during eccentric contractions, however to date no studies have validated the interpolated twitch technique for assessing this in the knee flexors. Surface EMG provides a global measure of the electrical contributions made by the active motor units (MUs), as detected by electrodes placed on the skin overlying the active muscle (Farina, Merletti, & Enoka, 2004). This technique provides further evidence for incomplete activation during maximal voluntary contractions, as semg amplitude is reduced during eccentric and slow concentric contractions, when compared to faster concentric contractions (Aagaard, et al., 2000; Amiridis, et al., 1996; Kellis & Baltzopoulos, 1998; Westing, et al., 1991). Interestingly, this reduction in semg is seen despite an increased torque-generating capacity associated with eccentric contractions (Katz, 1939; Westing, et al., 1991) (Figure 2-3). For instance, Kellis and Baltzopoulos (1998) measured semg of the knee extensors and flexors during maximal eccentric and concentric isokinetic contractions at a range of velocities. Results demonstrated that normalised EMG of both extensor and flexor muscle groups was significantly lower during eccentric contractions, when compared to their respective concentric values (Kellis & Baltzopoulos, 1998). Similarly, Westing et al. (1991) explored the semg-velocity relationship of the knee extensors and found that semg amplitudes were significantly reduced during eccentric contractions when compared to concentric contractions, despite greater torque values at all eccentric testing velocities. This 34

50 evidence is suggestive of neuromuscular inhibition of the knee flexors and extensors during maximal voluntary eccentric contractions. Limitations of inferring activation strategies from surface electromyography (semg) While semg is the only tool currently used in the assessment of knee flexor voluntary activation, it should be acknowledged that this technique has a number of limitations. The amplitude of semg is proportional to the net motor unit (MU) activity and therefore reflects the recruitment and discharge rates of active MUs (Farina, et al., 2004). While this is theoretically an index of the extent of muscle voluntary activation, semg amplitude is influenced by several other factors including electrode location (Farina, Cescon, & Merletti, 2002), subcutaneous tissue (Farina, et al., 2004), the distribution of MU conduction velocities (Arendt-Nielsen & Zwarts, 1989) and the degree to which MU firing is synchronous (Yao, Fuglevand, & Enoka, 2000). Indeed, the coefficient of variation for repeated semg measurements has been reported to be as high as 23% (Veiersted, 1991). Furthermore, while semg displays excellent temporal resolution, its spatial resolution is relatively poor. Given, the anatomical complexity of the hamstring muscle group (Woodley & Mercer, 2005), and the observation that the hamstrings display non-uniform activation during running (Schache, et al., 2012) and during different strengthening exercises (Ono, et al., 2011; Ono, et al., 2010), it is possible that semg is limited in its capacity to assess activation in this muscle group. Fine wire EMG overcomes some of these limitations and has been used previously to assess the hamstring EMG-joint angle relationship (Onishi, et al., 2002). While it should be acknowledged that this technique potentially allows for a more accurate spatiotemporal assessment of electromyographical activity (Ciccotti, Kerlan, Perry, & Pink, 1994), its use not widespread because of its invasive nature. 35

51 Assessing spatial activation via functional magnetic resonance imaging (fmri) Functional magnetic resonance imaging (fmri) is a unique technique that allows for a highresolution spatial assessment of the size and functional characteristics of muscles. For two decades this method has been established to noninvasively assess patterns and quantify the extent of skeletal muscle activation during exercise (Adams, Duvoisin, & Dudley, 1992; Foley, Jayaraman, Prior, Pivarnik, & Meyer, 1999; Jayaraman et al., 2004; Kinugasa, Kawakami, & Fukunaga, 2006; Mendiguchia et al., 2004; Ono, et al., 2011; Ono, et al., 2010). The premise of using fmri to assess voluntary activation is that exercise increases the proton transverse (spin-spin) (T2) relaxation time of skeletal muscle in fmri images (T2 1/2<20 min) (Jayaraman, et al., 2004) and this shift has been shown to increase proportionately with exercise intensity (Adams, et al., 1992; Adams, Harris, Woodard, & Dudley, 1993; Fisher, Meyer, Adams, Foley, & Potchen, 1990; Fleckenstein et al., 1991). Exercise-induced increases in T2 relaxation times are proportional to EMG measures of neuromuscular activation (Adams, et al., 1992) and to isometric torque evoked by electrical stimulation of skeletal muscle (Adams, et al., 1993). Functional MRI also provides high-resolution assessment of muscle cross-section and volume. Given the complexity of muscle architecture and neural innervations, fmri presents a unique advantage in its ability to sample the entire length of the muscle. Recently, fmri has been used to assess spatial patterns of knee flexor muscle activation during different hamstring strengthening exercises (Mendiguchia, Arcos, et al., 2013a; Mendiguchia, Garrues, et al., 2013; Ono, et al., 2011; Ono, et al., 2010). Further, this student s Honours project was the first to use fmri to explore the impact of previous HSI on spatial activation patterns of the hamstring muscles (Bourne, et al., In review). 36

52 2.9.2 Mechanism(s) underpinning neural inhibition Currently, the precise mechanism(s) responsible for inhibition of hamstring motoneuron activation remain unclear (Aagaard, et al., 2000). It is known that maximal voluntary activation is moderated by both reflex sensory (afferent) pathways and central descending pathways (Gandevia, 1998). Sensory afferent pathways originate from muscle spindles (group Ia and II), Golgi tendon organs (group Ib) and group III and IV afferents from the skin and joint receptors (Figure 2-4). It has been postulated that the reduction in maximal voluntary activation observed during eccentric and slow concentric contractions is most likely a result of inhibitory feedback from Golgi Tendon organs (group IIb afferents) and excitatory feedback from group Ia muscle spindles (Aagaard, et al., 2000). This is because these afferents converge onto the entire motoneuron pool from both agonist (homonymous afferent pathways) and antagonist muscles (heteronymous pathways) and therefore possess the greatest ability to monitor tension throughout the entire MTU (Gordon, 1991) and throughout the entire physiological range of force graduation (Houk, Crago, & Rymer, 1980). However, conflicting evidence suggests that sensory output from group IIb afferents in particular, is maximised at approximately 20-50% of MVC which would indicate that their involvement at higher levels of force development is limited (Rymer, Houk, & Crago, 1979). Indeed, it is now recognised that group IIb afferents may also exert excitatory effects on the muscles they innervate in certain movement contexts (Rio, Kidgell, Purdam, Gaida, Moseley, Pearce & Cook, 2015; Pratt, 1995). Reflex pathways from joint and ligament receptors (group II and III afferents) may also inhibit motoneuron activation in conditions of high tensile loading (Aagaard, et al., 2000). Certainly, evidence of an inhibitory reflex pathway from the human ACL to the quadriceps muscles has been reported previously (Dyhre-Poulsen & Krogsgaard, 2000). 37

53 Figure 2-4. A simplified scheme of the afferent synaptic inputs to alpha (ά) and gamma (γ) motoneurones (MN). Open and filled circles represent excitatory and inhibitory neurones, respectively. Shown are the type Ia and type II afferents from the muscle spindles, type Ib afferents from Golgi tendon organs and type III and IV afferents from the agonist muscle and the skin and joint structures (From Gandevia, 1998) The impact of resistance training on skeletal muscle activation It appears that resistance training may modulate the mechanism(s) responsible for limiting voluntary activation during high force contractions (Aagaard, et al., 2000; Amiridis, et al., 1996). For instance, heavy progressive-intensity resistance training has been shown to increase maximal voluntary activation of quadriceps in untrained men (Aagaard, et al., 2000). In this study (Aagaard, et al., 2000) pre-training EMG activity of the quadriceps was considerably lower during maximal voluntary eccentric and slow concentric contractions, when compared to fast concentric contractions. However, following a 14-week (38 sessions) 38

54 period of high-intensity resistance training, inhibition of quadriceps activation was either reduced or completely removed. In addition, the increase in activation was accompanied by significant increases in quadriceps strength during slow concentric and eccentric contractions (Aagaard, et al., 2000). In a separate isokinetic knee extension investigation (Amiridis, et al., 1996), elite-level high jumpers displayed significantly greater maximal activation of quadriceps during eccentric contractions when compared to healthy untrained individuals, suggesting that long-term conditioning to high-force loading may inhibit the tension-limiting mechanism(s) in human quadriceps. More recently, Rio et al (2015) employed EMG and transcranial magnetic stimulation on six athletes with patella tendinopathy, to determine the effect of an acute bout of isometric resistance exercise for the quadriceps on corticospinal excitability. Following isometric exercise, these athletes displayed a significant reduction in corticospinal inhibition to the quadriceps, which was accompanied by an increase in isometric knee extensor strength and a concurrent reduction in pain. These changes, which persisted for at least 45 min after the cessation of the training stimulus, highlight the effectiveness of resistance training on modulating voluntary drive, and demonstrate that the time-course of these adaptations may be quite rapid (Rio, et al, 2015). The role of resistance training in modulating the tension-regulating mechanism(s) of skeletal muscle can be further explained by the fact that strength increases significantly more than muscle CSA in the early weeks of resistance training (Higbie, Cureton, Warren, & Prior, 1996; Narici, Roi, Landoni, Minetti, & Cerretelli, 1989). Following short bouts of resistance training, several studies have reported an increase in MU firing rates (Van Cutsem, Duchateau, & Hainaut, 1998), enhanced reflex potentiation (motoneurone excitability) (Milner-Brown & Lee, 1975; Sale, Upton, McComas, & MacDougall, 1983), and increased EMG activity during MVC s (Hakkinen, Alen, & Komi, 1985; Hakkinen & Komi, 1986; 39

55 Komi & Buskirk, 1972; Narici, et al., 1989). These adaptations have also been shown to be reversed with detraining (Hakkinen, et al., 1985; Narici, et al., 1989). Other studies have demonstrated a reduction of the bilateral deficit (Häkkinen et al., 1996) and cross-education from trained limbs to untrained contralateral limbs (Hortobagyi et al., 1996; Moritani & DeVries, 1979; Weir, Housh, Housh, & Weir, 1995). All of these findings are suggestive of positive adaptations in neural control of muscle function as a consequence of training The impact of pain and injury on skeletal muscle activation The pain-adaptation model proposed by Lund, Donga, Widmer and Stohler (1991) suggests a relationship between musculoskeletal pain and motor activity. This theory suggests that in painful conditions, voluntary activation of agonistic muscles is reduced, while voluntary activation of antagonistic muscles is increased (Lund, et al., 1991). For example, following ACL rupture, significant and long-term deficits in quadriceps activity have been observed (Hurley, 1997; Urbach, et al., 2001) and in one study these alterations persisted despite the return of joint stability two years after reconstruction (Urbach, et al., 2001). Deficits in quadriceps activation have also been reported in individuals with knee (Hurley, 1997) and hip (Suetta et al., 2007) osteoarthritis (OA), patella tendinopathy (Rio, et al, 2015) and anterior knee pain (Suter, Herzog, & Bray, 1998) and those with lower back pain (Suter & Lindsay, 2001). Similar reductions in voluntary activation have been reported for the plantar flexors following traumatic ankle injury (Behm & St-Pierre, 1997). Experimentally-induced pain has also been shown to inhibit muscle activation and impair coordination patterns during static and dynamic motor functions. Intramuscular injection of hypertonic saline is associated with reduced torque production and movement velocity 40

56 (Svensson, Houe, & Arendt-Nielsen, 1997), reduced semg amplitude (Graven-Nielsen, Svensson, & Arendt-Nielsen, 1997; Svensson, et al., 1997) and a decline in MU firing rates (Sohn, Graven Nielsen, Arendt Nielsen, & Svensson, 2000) in the painful muscle. For example, saline injections into the supraspinatus muscle have been shown to reduce agonist (deltoid and upper trapezius) activity and increase antagonist muscle activity (latissimus dorsi and lower trapezius) during shoulder elevation (Diederichsen, et al., 2009). Hodges and colleagues (2009) demonstrated similar reductions in quadriceps activation following saline injections into the infrapatellar fat pad. In this study participants displayed delayed activation of vastus medialis obliquus during stair climbing, relative to the ipsilateral pain-free vastus lateralis. They also displayed a reduction in vastus lateralis activation when compared to the homologous vastus lateralis in the pain-free contralateral limb (Hodges, et al, 2009). Such adaptations lend further support to the notion that voluntary activation is inhibited to avoid pain provocation from muscle contraction Evidence for neuromuscular inhibition following hamstring strain injury Findings from this student s Honours project (Appendix A) suggest that previously injured hamstring muscles display long-lasting deficits in activation compared to uninjured contralateral muscles during a common eccentric conditioning exercise, known as the Nordic hamstring exercise (NHE) (Bourne, et al, 2015) (Figure 2-5). During a bout of NHEs previously injured muscles (mean time of 9.8 months post-injury) were approximately 40% less active than homonymous BFLH muscles in the uninjured contralateral limb as assessed by fmri (Figure 2-5). These observations support recent findings from our laboratory of reduced levels of semg activity in previously injured BF muscles when compared to the uninjured contralateral BF muscles during eccentric but not concentric isokinetic contractions 41

57 (Opar, Williams, Timmins, Dear, & Shield, 2013a). This potentially explains the mechanism(s) for why previously injured hamstrings display considerable deficits in eccentric strength in the presence of smaller or absent concentric strength deficits (Croisier, 2004; Croisier, et al., 2002; Jonhagen, et al., 1994; Lee, et al., 2009). Figure 2-5. Percentage change in fmri T2 relaxation times of each hamstring muscle for both the previously injured (inj) and uninjured (uninj) limbs. Values are expressed as a mean percentage change compared to the values at rest. * indicates a significant difference between limbs for individual muscles (p<0.05). Error bars depict standard deviation. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. 42

58 2.9.6 Impact of neuromuscular inhibition on hamstring muscle morphology and architecture Previously injured hamstrings have been reported to exhibit chronic alterations in hamstring muscle and tendon morphology (Silder, et al., 2008). Silder and colleagues (2008) conducted MRI on 14 athletes with a history of HSI 5-23 months post injury and five uninjured controls to investigate between-limb differences in hamstring morphology following injury. Athletes who had previously strained the BFLH (n=13) displayed significant reductions in BFLH muscle volumes (Figure 2-6) and this residual atrophy was often accompanied by concomitant hypertrophy of the ipsilateral BFSH. The authors suggested that these morphological differences may have been influenced by several factors including the severity of the insult, the frequency, intensity and modality of exercises employed in rehabilitation, as well as the intensity of training upon return to competition (Silder, et al., 2008). However, it was concluded that chronic hypertrophy of BFSH was most likely an exercise-induced compensation for atrophy of BFLH, in an effort to conserve global knee flexion strength (Silder, et al., 2008). This evidence is strongly supportive of persistent neuromuscular inhibition of the previously injured BFLH. 43

59 Figure 2-6. MRI image illustrating a previously injured BFLH (right limb) and uninjured contralateral BFLH (left limb). Note the atrophy of the previously injured BFLH with corresponding hypertrophy of the BFSH, relative to the uninjured contralateral limb (Silder et al., 2008). Neuromuscular inhibition following HSI has also been reported to account for fascicular shortening in the previously injured muscle (Fyfe, Opar, Williams, & Shield, 2013b). Timmins et al. (2014) have recently reported that previously injured BFLH muscles display significantly shorter fascicles, coupled with increased pennation angles, compared to homologous muscles in the uninjured contralateral limb. The authors proposed that a reduced capacity to activate the previously injured muscle throughout the rehabilitation process (particularly during eccentric contractions) (Opar, Williams, et al., 2013a), may result in a shedding of serial sarcomeres. This reduction in serial sarcomeres should result in a leftward shift of the BFLH s force-length relationship (Reeves, Narici, & Maganaris, 2004), which may increase its susceptibility to damage at longer lengths such as those experienced in the terminal-swing phase of high-speed running (Brockett, et al., 2001). The restoration of BFLH fascicle lengths is an important component of rehabilitation, in light of recent evidence showing that professional soccer players with short BFLH fascicles are four times more likely to suffer an HSI than those with longer fascicles (Timmins, Bourne, et al., 2015). 44

60 Figure 2-7. Architectural characteristics of the injured BF LH and the contralateral uninjured BF LH in the previously injured group at all contraction intensities (p<0.05) Neuromuscular inhibition as a mechanism for high rates of hamstring strain injury recurrence Recently, our group proposed a novel conceptual framework for the development of neuromuscular inhibition following an HSI and its potential role in HSI recurrence (Figure 8) (Fyfe, Opar, Williams, & Shield, 2013a; Opar, et al., 2012). Given the known relationships between experimentally-induced pain and deficits in maximal voluntary activation of local musculature, it is logical to expect a degree of acute neuromuscular inhibition immediately following an HSI. We hypothesise that the pain associated with the initial insult triggers an acute alteration of neural control designed to protect the injured muscle fibres and fascicles and connective tissues from further damage. A reduced ability to activate the previously 45

61 injured muscle, particularly during eccentric actions and at longer muscle lengths, would result in a number of acute and potentially chronic changes to muscle structure and function, including a reduction of peak eccentric torque (Graven Nielsen, Lund, Arendt Nielsen, Danneskiold Samsøe, & Bliddal, 2002), a loss of serial sarcomeres (Brockett, et al., 2004), and potentially, resultant atrophy of the injured muscle (Silder, et al., 2008). Although deficits in activation appear to be an acute response to the pain associated with injury, compensatory activation strategies learned during rehabilitation may result in a chronic rewiring of neural pathways (Hodges & Tucker, 2011). These changes are likely to occur at multiple levels of the nervous system in an effort to redistribute motor activity within and between muscles (Hodges & Tucker, 2011). Indeed, the resolution of experimentally-induced pain or injury does not automatically trigger a return to initial motor patterns (Hodges & Tucker, 2011), and it is possible that these pain-induced patterns may become ingrained during rehabilitation and may still persist even when athletes are deemed fit to return to competition. If neuromuscular inhibition is not addressed during the rehabilitative process then the athlete is left weaker during eccentric contractions and more susceptible to muscle damage and together, these factors are likely to predispose them to a heightened risk of reinjury. 46

62 Figure 2-8. Conceptual model for the development of neuromuscular inhibition following hamstring strain injury and its potential role in injury recurrence. * Particularly at long muscle lengths, # biceps femoris (BF) specific (Fyfe et al., 2013). 47

63 Chapter 3: PROGRAM OF RESEARCH The goals of this program of research are to 1) further examine the role of eccentric knee flexor strength and between-limb imbalances in hamstring injury occurrence; 2) explore the neuromuscular mechanism(s) which may underpin high rates of HSI recurrence; and 3) in an effort to improve HSI prevention strategies, characterise the activation patterns and the architectural and morphological adaptations of the hamstrings to difference strength training exercises. One major aim of this program of research is to further examine the association between eccentric hamstring strength and injury risk in sport. Recently our group has developed a novel field testing device for the assessment of eccentric hamstring strength which overcomes some of the limitations of isokinetic dynamometry (Opar, Piatkowski, Williams, & Shield, 2013). Using the commonly employed NHE, the device is able to record maximal eccentric hamstring strength and between limb imbalances in less than two minutes. In two recent large scale prospective studies conducted by our group, elite Australian footballers (Opar, et al., 2014) and professional soccer players (Timmins, Bourne, et al., 2015) who displayed eccentric hamstring weakness during the NHE, were ~4 times more likely to suffer a future HSI compared to stronger athletes. It is therefore of practical importance to prospectively explore these relationships in other sports with high incidence rates of HSI so as to reduce the burden of this troublesome injury. It is also important to further explore the effects of previous hamstring injury on knee flexor strength. 48

64 We recently proposed that high rates of HSI recurrence might be partly explained by chronic neuromuscular inhibition of the BF (Fyfe, et al., 2013; Opar, et al., 2012) which has been observed during eccentric but not concentric knee flexor efforts (Bourne, et al.,2015; Opar, Williams, et al., 2013a). These contraction mode-specific deficits in BF activation persist despite rehabilitation and return to sport and may mediate preferentially eccentric hamstring weakness (Croisier & Crielaard, 2001; Croisier, et al., 2002; Jonhagen, et al., 1994), reduced rates of knee flexor torque development (Opar, Williams, Timmins, Dear, & Shield, 2013b), persistent BFLH atrophy (Silder, et al., 2008), and altered BFLH architecture (Timmins et al., 2015) all of which have been observed months to years after an HSI. These activation deficits that persist throughout rehabilitation would reduce the injured muscle s loading, particularly during eccentric contractions at longer muscle lengths (Opar, Williams, et al., 2013a; Sole, Milosavljevic, Nicholson, & Sullivan, 2011a, 2011b) and this likely compromises hypertrophy and sarcomerogenesis (Brockett, et al., 2001), both of which are thought to be important in allowing muscles to adapt to the demands of sprinting and strengthening exercises. While inhibition of BF muscles appears to be a robust and persistent phenomenon in previously injured athletes, to date it has only been explored during eccentric isokinetic tasks (Opar, Williams, et al., 2013a; Sole, et al., 2011a) and the NHE (Bourne, et al., 2015). It remains to be seen whether activation deficits are also present during the presumably injurious (Brooks, et al., 2005c; Ekstrand, et al., 2011a; Woods, et al., 2002; Yu, et al., 2008) terminal-swing phase of high speed running. With respect to the restoration of neuromuscular function following an HSI, heavy, progressive intensity resistance training has been shown to improve activation of skeletal muscle (Aagaard, et al., 2000; Amiridis, et al., 1996; Carolan & Cafarelli, 1992; Deschenes & Kraemer, 2002; Dudley, Tesch, Miller, & Buchanan, 1991). While no studies have measured 49

65 voluntary hamstring activation following a period of strength training, interventions aimed at improving eccentric hamstring strength appear very effective in reducing the incidence of both first-time and recurrent HSIs (Arnason, et al., 2008; Askling, et al., 2003; Askling, Tengvar, & Thorstensson, 2013; Croisier, et al., 2008c; Petersen, Thorborg, Nielsen, Budtz- Jørgensen, & Hölmich, 2011). However, different exercises appear to target different hamstring muscles and different portions of those muscles and the BF is not always heavily recruited (Bourne, et al., 2015; Zebis et al., 2012). As BFLH injuries represent up to 80% of all hamstring strains (Connell et al., 2004; Koulouris, et al., 2007), an improved understanding of which exercises preferentially activate and stimulate adaptations in the BFLH, will be paramount in designing interventions aimed at improving eccentric strength and activation in this muscle. Objectives The major questions to be addressed in this program of research include: 1) Are eccentric strength deficits or between-limb imbalances predictive of future HSI in athletes? 2) Do athletes with a history of HSI display altered patterns of muscle activation during high-speed overground running? 3) What is the optimal exercise to improve voluntary activation and strength in the commonly injured biceps femoris long head (BFLH)? 50

66 4) How does hamstring architecture and morphology adapt to a targeted progressive intensity resistance training intervention? 51

67

68 Chapter 4: STUDY 1 ECCENTRIC STRENGTH AND HAMSTRING INJURY RISK IN RUGBY UNION: A PROSPECTIVE COHORT STUDY Publication statement This chapter is comprised of the following paper published in the American Journal of Sports Medicine: Bourne, MN., Opar, DA., Williams, MD., & Shield, AJ. (2015). Eccentric Knee-flexor Strength and Hamstring Injury Risk in Rugby Union: A prospective study. Am J Sports Med, 43(11): doi: /

69 Statement of Contribution of Co-Authors for Thesis by Published Paper The authors listed below have certified* that: 1. They meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. They take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. There are no other authors of the publication according to these criteria; 4. Potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit 5. They agree to the use of the publication in the student s thesis and its publication on the Australasian Research Online database consistent with any limitations set by publisher requirements. Contributor Matthew Bourne 08/03/2016 David Opar Morgan Williams Anthony Shield Statement of contribution* Experimental design, ethical approval, data collection and analysis, statistical analysis, manuscript preparation Aided in experimental design and manuscript preparation Assisted with statistical analysis and manuscript preparation Aided in experimental design and manuscript preparation Principal Supervisor Confirmation I have sighted from all co-authors confirming their certifying authorship. Dr Anthony Shield 54

70 4.1 ABSTRACT BACKGROUND: Hamstring strain injuries represent the most common cause of lost playing time in rugby union. Eccentric knee-flexor weakness and between-limb imbalances in eccentric knee-flexor strength are associated with a heightened risk of hamstring injury in other sports; however these variables have not been explored in rugby union. PURPOSE: To determine if lower levels of eccentric knee-flexor strength or greater between-limb imbalance in this parameter during the Nordic hamstring exercise are risk-factors for hamstring strain injury in rugby union. STUDY DESIGN: Cohort study; level of evidence, 3. METHODS: This prospective study was conducted over the 2014 Super Rugby and Queensland Rugby Union seasons. In total, 178 rugby union players (age, 22.6 ± 3.8 years; height, 185 ± 6.8 cm; mass, 96.5 ± 13.1 kg) had their eccentric knee-flexor strength assessed using a custom-made device during the pre-season. Reports of previous hamstring, quadriceps, groin, calf and anterior cruciate ligament injury were also obtained. The main outcome measure was prospective occurrence of hamstring strain injury. RESULTS: Twenty players suffered at least one hamstring strain during the study period. Players with a history of hamstring strain injury had 4.1 fold (RR = 4.1, 95% CI = 1.9 to 8.9, p = 0.001) greater risk of subsequent hamstring injury than players without such history. Between-limb imbalance in eccentric knee-flexor strength of 15% and 20% increased the risk of hamstring strain injury 2.4 fold (RR = 2.4, 95% CI = 1.1 to 5.5, p = 0.033) and 3.4 fold (RR = 3.4, 95% CI = 1.5 to 7.6, p = 0.003), respectively. Lower eccentric knee flexor strength and other prior injuries were not associated with increased risk of future hamstring strain. Multivariate logistic regression revealed that the risk of re-injury was augmented in players with strength imbalances. CONCLUSION: Previous hamstring strain injury and between-limb imbalance in eccentric knee-flexor strength were associated with an increased risk of future hamstring strain injury 55

71 in rugby union. These results support the rationale for reducing imbalance, particularly in players who have suffered a prior hamstring injury, to mitigate the risk of future injury. 56

72 4.2 INTRODUCTION Rugby union is a physically demanding contact game with one of the highest reported incidences of match injuries of all sports (Brooks, et al., 2005c; Fuller, Sheerin, & Targett, 2013; Williams, Trewartha, Kemp, & Stokes, 2013). The unique nature of the sport exposes athletes of varying anthropometric characteristics (Zemski, Slater, & Broad, 2015) to frequent bouts of high-intensity running, kicking, and unprotected collisions, interspersed with periods of lower intensity aerobic work (Duthie, Pyne, & Hooper, 2003). Hamstring strain injury represents the most common cause of lost playing and training time at the professional level (Brooks, et al., 2005a, 2005b) and a significant portion of these injuries re-occur, resulting in extended periods of convalescence (Brooks, et al., 2006). Despite the prevalence of HSIs in rugby union (Brooks, et al., 2005a), efforts to identify risk factors and to optimise injury prevention strategies are limited (Brooks, et al., 2006; Upton, Noakes, & Juritz, 1996). It is generally agreed that the aetiology of HSI is multifactorial (Mendiguchia, Alentorn-Geli, & Brughelli, 2012) and injuries result from the interaction of several modifiable (Burkett, 1970; Croisier, et al., 2002; Croisier, et al., 2008c; Heiser, et al., 1984; Opar, et al., 2014; Orchard, et al., 1997b; Sugiura, et al., 2008) and non-modifiable (Arnason, et al., 2004; Gabbe, Bennell, & Finch, 2006; Hagglund, et al., 2006; Verrall, et al., 2001) risk factors. In rugby union (Brooks, et al., 2006), as well as several other sports (Askling, et al., 2007; Opar, et al., 2014; Woods, et al., 2004), HSIs most frequently result from high-speed running which potentially explains why the incidence of HSI is significantly higher for backline rugby players (Brooks, et al., 2006), who perform longer and more frequent sprints than forwards. During running, the biarticular hamstrings play a crucial role in decelerating the forward swinging shank during terminal-swing (Yu, et al., 2008) and in generating horizontal force upon ground contact (Mann, Moran, & Dougherty, 1986). Given 57

73 the active lengthening role of the hamstrings it has been proposed that eccentric weakness (Opar, et al., 2014) or between-limb imbalances in eccentric strength may predispose to HSI, and both factors have been associated with the risk of HSI in other sports (Croisier, et al., 2008c; Fousekis, Tsepis, Poulmedis, Athanasopoulos, & Vagenas, 2011; Heiser, et al., 1984; Yamamoto, 1993). Furthermore, interventions aimed at improving eccentric strength with the Nordic hamstring exercise reduce the incidence and severity of HSIs in soccer (Arnason, et al., 2008; Petersen, et al., 2011) while professional rugby union teams employing the exercise have been reported to suffer fewer HSIs than those which do not (Brooks, et al., 2006). Still, the role of eccentric strength in HSI occurrence remains a controversial issue with contradictory results reported in the literature (Bennell, et al., 1998a; Zvijac, Toriscelli, Merrick, & Kiebzak, 2013) and a recent meta-analysis suggested that isokinetically-derived measures of strength do not represent a risk factor for HSI (Freckleton & Pizzari, 2013). Nevertheless, the authors are not aware of any study that has examined the relationship between eccentric knee-flexor strength, between-limb imbalance, and HSI incidence in rugby union. Given the unique anthropometric characteristics of rugby union players (Zemski, et al., 2015) and the diverse physical demands of the game (Duthie, et al., 2003; Williams, et al., 2013), it may not be appropriate to generalise the findings from other sports to this cohort. It has been shown that eccentric knee flexor strength can be reliably measured during the performance of the Nordic hamstring exercise (Opar, Piatkowski, et al., 2013a). In a recent prospective study of elite Australian footballers (Opar, et al., 2014), players with low Nordic strength measures in the pre-season training period were significantly more likely to sustain an HSI in the subsequent competitive season. However, it remains to be seen if the same measures can identify rugby union players at risk of future HSI. 58

74 An improved understanding of risk factors for HSI in rugby union represents the first step (van Mechelen, Hlobil, & Kemper, 1992) towards optimising injury prevention strategies and reducing the high rates of HSI occurrence in the sport (Brooks, et al., 2005a, 2005b). The aim of this study was to determine whether pre-season eccentric knee-flexor strength and between-limb imbalance in strength measured during the Nordic hamstring exercise, were predictive of future HSI in rugby union players. In addition, given the multifactorial aetiology of HSI (Mendiguchia, Alentorn-Geli, et al., 2012) and the potential for various risk factors to interact (Thorborg, 2014), a secondary aim was to determine the association between measures of eccentric strength, imbalance and other previously identified risk factors, such as prior HSI (Brooks, et al., 2006; Thorborg, 2014). The a priori hypotheses were that subsequently injured players would display lower levels of eccentric knee-flexor strength and greater between-limb imbalances in this measure than players who remained free from HSI. 4.3 METHODS Participants & study design This prospective cohort study was approved by the Queensland University of Technology s Human Research Ethics Committee and was completed during the 2014 Super 15 and Queensland Rugby Union (QRU) seasons. In total, 194 male rugby players (age, 22.6 ± 3.8 years; height, 185 ± 6.7 cm; weight, 97 ± 13.1 kg) from three professional Super 15 clubs (n=75) and two local QRU clubs (n=119) provided written informed consent to participate. The QRU clubs included players in both sub-elite (n=79) and U 19 premier-grade teams (n=40). Prior to the commencement of data collection, retrospective injury details were collected for all players which included their history of hamstring, quadriceps and calf strain injuries and chronic groin pain within the preceding 12 months as well as history of anterior cruciate ligament (ACL) injury at any stage in their career. Demographic (age) and 59

75 anthropometric (height, body mass) data were also collected in addition to player position (forward, back). For all Super 15 players these data were obtained from team medical staff and the national Australian Rugby Union registry. All sub-elite players completed a standard injury history form with their team physiotherapist and injuries were confirmed with information from each club s internal medical reporting system. Subsequently, players had their eccentric knee flexor strength assessed at a single time point within the 2014 pre-season (Super 15, November 2013; sub-elite, January 2014). At the discretion of team medical staff, some players (n=16) were excluded from strength testing because they had an injury or illness at the time of testing that precluded them from performing maximal resistance exercise Eccentric knee-flexor strength assessment The assessment of eccentric knee-flexor strength during the Nordic hamstring exercise has been reported previously (Opar, Piatkowski, et al., 2013a; Opar, et al., 2014). Participants knelt on a padded board, with the ankles secured immediately superior to the lateral malleolus by individual ankle braces which were attached to custom made uniaxial load cells (Delphi Force Measurement, Gold Coast, Australia) with wireless data acquisition capabilities (Mantracourt, Devon, UK) (Figure 1). The ankle braces and load cells were secured to a pivot which allowed the force generated by the knee flexors to always be measured through the long axis of the load cells. Immediately prior to testing, players were provided with a demonstration of the Nordic hamstring exercise from investigators and received the following instructions: gradually lean forward at the slowest possible speed while maximally resisting this movement with both limbs while keeping the trunk and hips in a neutral position throughout, and the hands held across the chest (Opar, et al., 2014). Subsequently, players completed a single warm-up set of three repetitions followed by one 60

76 set of three maximal repetitions of the bilateral Nordic hamstring exercise. All trials were closely monitored by investigators to ensure strict adherence to proper technique and players received verbal encouragement throughout each repetition to encourage maximal effort. A repetition was deemed acceptable when the force output reached a distinct peak (indicative of maximal eccentric strength), followed by a rapid decline in force which occurred when the athlete was no longer able to resist the effects of gravity acting on the segment above the knee joint. All eccentric strength testing was performed in a rested state, prior to the commencement of scheduled team training. Figure 4-1. The Nordic hamstring exercise performed on the testing device (progressing from right to left). Participants were instructed to lower themselves to the ground as slowly as possible by performing a forceful eccentric contraction of their knee flexors. Participants only performed the eccentric portion of the exercise and after catching their fall, were instructed to use their arms to push back into the starting position (not shown here). The ankles are secured independently. Data analysis Force data for the left and right limbs were transferred to a personal computer at 100Hz through a wireless USB base station receiver (Mantracourt, Devon, UK). Eccentric strength, determined for each leg from the peak force during the best of three repetitions of the NHE, 61

77 was reported in absolute terms (N) and relative to bodyweight (N.kg -1 ). For the uninjured group, between limb imbalance in peak eccentric knee-flexor force was calculated as a left:right limb ratio and for the injured group, as an uninjured:injured limb ratio. The between limb imbalance ratio was converted to a percentage difference as per previous work (Opar, et al., 2014) using log transformed raw data followed by back transformation. Prospective hamstring strain injury reporting An HSI was defined as acute pain in the posterior thigh which caused immediate cessation of training or match play and damage to the hamstring muscle-tendon unit (Opar, et al., 2014) which was later confirmed with magnetic resonance imaging (for all Super 15 players) or clinical examination by the team physiotherapist (for all sub-elite and U 19 players). For all injuries that satisfied the inclusion criteria, team medical staff provided the following details to investigators: limb injured (left / right), muscle injured (biceps femoris long or short head/semimembranosus/semitendinosus, injury severity (grade 1-3), injury mechanism (ie, running, kicking, collision, change of direction), the date of injury and whether it was a recurrence and the total time taken to resume full training and competition. Statistical analysis All statistical analyses were performed using JMP (SAS Institute, Inc). Mean and standard deviations (SD) of age, height, weight, eccentric knee-flexor strength for the left and right limb and between-limb imbalance (%) in strength were determined. Because the player and not the leg was the unit of measure in some analyses, it was necessary to have a single measure of eccentric knee-flexor strength for each athlete and this was determined by averaging the peak forces from each limb (two-limb-average strength). Univariate analysis was used to compare age, height, weight and between-limb imbalance between the injured 62

78 and uninjured groups. Eccentric knee-flexor strength of the injured limb was compared to the uninjured contralateral limb and to the average of the left and right limbs from the uninjured control group. In addition, eccentric knee-flexor strength was compared between elite, subelite and U 19 players and between player positions (forwards vs. backs). All univariate comparisons were made using independent samples t tests with Bonferroni corrections to control for Type 1 error. To calculate univariate relative risk (RR) and 95% confidence intervals (95% CI), players were grouped according to: whether they did or did not have a history of o HSI in the previous 12 months o quadriceps strain injury in the previous 12 months o chronic groin pain in the previous 12 months o calf strain injury in the previous 12 months o or ACL injury at any stage; Two-limb-average eccentric knee-flexor strength above or below 267.9N or 3.18N.kg -1 (these cut-offs were determined using receiver operator characteristic (ROC) curves based on the force and relative force values that maximised the difference between sensitivity and 1 specificity). between-limb eccentric strength imbalance above or below a 10, 15 or 20% cut-off; whether they were above or below the 25 th, 50 th and 75 th percentiles for: 63

79 o age o height o weight Any variable associated with subsequent HSI according to univariate analysis was entered into a univariate logistic regression model to determine its predictive value as a risk factor for future HSI. Furthermore, given the multifactorial nature of HSI, a multivariate logistic regression model was constructed (using prior HSI and between-limb imbalance) to explore the potential interaction between risk factors (Opar, et al., 2014) and eliminate any confounding effects (Orchard, 2001). Alpha was set at p<0.05 and for all univariate analyses the difference between limbs and groups is reported as mean difference and 95% CI. 4.4 RESULTS Cohort and prospective hamstring strain injury details In total, 178 players (age, 22.6 ± 3.8 years; height, 185 ± 6.8 cm; weight, 96.5 ± 13.1 kg) had their eccentric knee-flexor strength assessed in the pre-season period. Of these, 75 were elite (age, 24.4 ± 3.1 years; height, 186 ± 7.2 cm; weight, 101 ± 11.3 kg), 65 were sub-elite (age, 21.3 ± 3.7 years; height, 184 ± 6.4 cm; weight, 93 ± 13.4 kg) and 38 were in the U 19 division (age, 18.1 ± 0.8 years; height, 183 ± 6.8 cm; weight, 91 ± 14.9 kg). Twenty athletes suffered at least one HSI during the 2014 competitive season (age, 22.8 ± 3.2 years; height, ± 5.5 cm; weight, 97.4 ± 12.4 kg) and 158 remained free of HSI (age, 22.5 ± 3.8 years; height, ± 7.0 cm; weight, 96.4 ± 13.3 kg). No significant differences were observed in terms of age, height or body mass between the subsequently injured and uninjured players (p>0.05). Hamstring strains resulted in an average of 21 days (range = 7 to 64

80 49 days) absence from full training and match play. Forty-five percent were recurrences from the previous season and 25% of those reported during the observation period recurred. Of the 20 injuries, 80% affected the BF as the primary site of injury and 85% resulted from highspeed running. The majority of HSIs were sustained by backs (60%) compared to forwards (40%). No injuries were sustained during the assessment of eccentric knee-flexor strength. Comparison of strength between playing level and position Eccentric strength measures for each level of play and player position can be found in Table 4-1. In terms of eccentric strength, there was no significant difference between elite and subelite players (mean difference = 21N, 95% CI = -7.8 to 49.9N, p = 0.154) or between elite and U 19 players (mean difference = 24.1N, 95% CI = to 55.0 N, p = 0.126) however, sub-elite players were significantly stronger than U 19 players (mean difference 45.1N, 95% CI = 8.1 to 82.0N, p = 0.017). When expressed relative to bodyweight, both sub-elite (mean difference = 0.35, 95% CI = 0.08 to 0.63, p = 0.013) and U 19 players (mean difference = 0.38N, 95% CI = 0.07 to 0.70, p = 0.017) were significantly stronger than elite players although no difference was observed between sub-elite and U 19 players (mean difference = , 95% CI = -0.4 to 0.34, p = 0.870). In absolute terms, forward line players were significantly stronger than backs (mean difference = 35.3N, 95% CI = to 60.5N, p= 0.006) however, no difference was observed when strength was normalised to bodyweight (mean difference = -0.1, 95% CI = to 0.16, p = 0.583). 65

81 Table 4-1. Pre-season Nordic hamstring exercise force variables for each level of competition and player position. Playing n Absolute eccentric knee flexor Relative eccentric knee flexor group strength (N) strength (N.kg -1 ) Elite ± ± 0.71** Sub-elite ± 96.3* 4.00 ± 0.93 U ± ± 0.92 Forward ± 95.5 # 3.81 ± 0.92 Back ± 74.9 # 3.9 ± 0.80 Data is presented as mean ± standard deviation. * Indicates significantly different from the U 19 group; ** indicates significantly different from both the U 19 and sub-elite groups; # indicates a significant difference between forwards and backs. 66

82 Univariate analysis of factors associated with hamstring strain injury Eccentric knee-flexor strength and between-limb imbalances for the injured and uninjured groups can be found in Table 4-2. Limbs that went on to be injured were significantly weaker in pre-season than uninjured contralateral limbs both in absolute terms (mean difference = 55.1N, 95% CI = to 98.5N, p = 0.016) and when normalised to body mass (mean difference = 0.55 N.kg -1, 95% CI = 0.13 to 0.98N.kg -1, p = 0.013). Players who went on to sustain an HSI displayed higher levels of between-limb imbalance than those players who remained free from HSI (mean difference = -7.4%, 95% CI = to -2.4%, p = 0.004). However, there was no difference between the subsequently injured limb and the average of the left and right limbs from the uninjured group either in absolute strength (mean difference = -14.9N, 95% CI = to 25.6N, p = 0.470) or strength relative to body mass (mean difference = N.kg -1, 95% CI = to 0.33 N.kg -1, p = 0.710). No significant differences were observed in age (mean difference = 0.18yrs, 95% CI = -1.5 to 1.9yrs, p = 0.235), height (mean difference = 0.86cm, 95% CI = -2.3 to 4.1cm, p = 0.457), or weight (mean difference = 0.97kg, 95% CI = -5.2 to 7.4kg, p = 0.632) between the injured and uninjured groups. 67

83 Table 4-2. Pre-season Nordic hamstring exercise force variables for hamstring strain injured and uninjured rugby union players. Group Limb Absolute eccentric knee flexor strength (N) Relative eccentric knee flexor strength (N.kg-1) Between-limb imbalance (%) Injured Injured (n=20) Uninjured (n=20) ± 80.5* 3.65 ± 0.67* ± 132.4* 4.21 ± 1.14* ± 16.1# Uninjured Average of left and right (n=158) ± ± ± 9.8# Data is presented as mean ± standard deviation. * Indicates significant differences between limbs in the injured group (p<0.05). # Significant differences between injured and uninjured players. 68

84 Relative risk Players with a history of HSI in the previous 12 months had 4.1 (RR = 4.1, 95% CI = 1.9 to 8.9, p = 0.001) times greater risk of suffering a subsequent HSI than players with no HSI in the same period (Table 4-3). Between-limb imbalance in eccentric knee-flexor strength of 15% increased the risk of HSI 2.4 fold (RR = 2.4, 95% CI = 1.1 to 5.5, p = 0.033) while an imbalance 20% increased that risk 3.4 fold (RR = 3.4, 95% CI = 1.5 to 7.6, p = 0.003). However, players with two-limb-average eccentric knee-flexor strength of less than 267.9N were not at elevated risk of HSI (RR = 0.17, 0.0 to 2.7, p = 0.204) compared to stronger players (area under the ROC curve = 0.52; specificity= 0.86; sensitivity = 1.0). Similarly, having normalised strength values of less than 3.18N.kg -1 did not increase the risk of HSI (RR = 0.97, 95%CI = 0.3 to 2.7, p = 0.957). 69

85 Table 4-3. Univariate relative risk of suffering a future hamstring strain injury (HSI) using eccentric strength and between-limb imbalance, injury history and demographic data as risk factors. Risk factor n % from each group that sustained HSI Relative risk (95%CI) P Prior injury HSI (1.9 to 8.9) No HSI (0.1 to 0.5) ACL (0.3 to 4.6) No ACL (0.2 to 3.3) Calf strain (0.2 to 8.5) 0.56 No calf strain (0.1 to 4.8) Quadriceps strain (0.1 to 6.2) No quadriceps strain (0.2 to 7.3) Chronic groin pain (0.1 to 5.2) No chronic groin pain (0.2 to 9.0) Pre-season eccentric hamstring strength <267.9N (0.01 to 2.7) N (0.4 to 96.0) <3.18N.kg (0.3 to 2.7) N.kg (0.4 to 2.9) Pre-season between-limb imbalance <10% (0.3 to 1.6) % (0.6 to 3.3) <15% (0.2 to 0.9) % (1.1 to 5.5) <20% (0.1 to 0.7) % (1.5 to 7.6) 70

86 Age (years) (0.2 to 2.0) > (0.5 to 5.5) (0.4 to 2.2) > (0.5 to 2.5) (0.4 to 3.4) > (0.3 to 2.4) Height (cm) (0.2 to 1.6) > (0.6 to 5.1) (0.3 to 1.7) > (0.6 to 2.9) (0.3 to 1.8) > (0.5 to 3.3) Weight (kg) (0.15 to 1.6) > (0.6 to 6.5) (0.5 to 2.2) > (0.5 to 2.2) (0.4 to 2.8) > (0.4 to 2.4) * Indicates a significant difference (p<0.05) in the relative risk of future HSI between groups. 95%CI, 95% confidence interval; HSI, hamstring strain injury; ACL, anterior cruciate ligament; N, newtons; cm, centimetre; kg, kilograms. Logistic regression Players with a history of HSI in the previous 12 months were, according to the odds ratio, 5.3 times more likely (OR = 5.3, 95%CI = 1.84 to 15.0, p = 0.003) to suffer a subsequent HSI than players who had remained injury free in that time. In addition, a relationship was observed between the magnitude of between-limb imbalance in eccentric knee-flexor strength 71

87 and the risk of subsequent HSI; where, for every 10% increase in between-limb imbalance, the odds of HSI increased by a factor of 1.34 (95%CI = 1.03 to 1.75, p = 0.028) (Figure 4-2). Multivariate logistic regression revealed a significant (p < 0.001) relationship between both prior HSI and between-limb imbalance and the risk of subsequent HSI (Table 4-4), however, no interaction effect was observed between these variables. This model suggests that for players with a history of HSI, the risk of re-injury is amplified when they also have betweenlimb imbalances in eccentric knee flexor strength (Figure 4-2). Figure 4-2. The relationship between eccentric knee flexor strength imbalances and probability of future hamstring strain injury (HSI) for players with and without a history of HSI in the previous 12 months. Errors bars depict 95% confidence intervals. 72

88 Table 4-4. Multivariate logistic regression model using prior hamstring strain injury (HSI) and between-limb imbalance in eccentric knee flexor strength as input variables ChiSquare p AUC Sensitivity 1-Specificity Whole Model < Prior HSI Between-limb imbalance (%) AUC, area under the curve 73

89

90 4.5 DISCUSSION The aim of this study was to determine if rugby union players with lower levels of eccentric strength or larger between-limb imbalances in this measure, as determined during the Nordic hamstring exercise, were at increased risk of HSI. Higher levels of between-limb imbalance were found to significantly increase the risk of subsequent HSI and this was amplified in athletes who had suffered the same injury in the previous 12 months. However, while the limbs that went on to be injured were significantly weaker than the uninjured contralateral limbs in pre-season testing, weaker players were no more likely to suffer injury than stronger players when strength was determined by averaging the peak eccentric forces from left and right limbs. The observation that higher levels of between-limb strength imbalance increase an athlete s risk of HSI is consistent with previous reports (Croisier, et al., 2008; Fousekis, et al., 2011; Heiser, et al., 1984; Orchard, et al., 1997b; Yamamoto, 1993). Croisier and colleagues (2008) reported that professional soccer players with isokinetically-derived knee-flexor strength imbalances in pre-season had a 4.66 fold greater risk of subsequent HSI than athletes without such imbalances. More recently, Fousekis and colleagues found that elite soccer players with imbalances in eccentric knee-flexor strength 15% in the pre-season had a significantly greater (OR = 3.88) risk of HSI than athletes with no asymmetry (Fousekis, et al., 2011). Still, contradictory results have been reported in Australian footballers (Bennell, et al., 1998a; Opar, et al., 2014) and it remains unclear as to the exact mechanism(s) by which significant imbalances increase the risk of HSI. It is plausible that between-limb imbalances in eccentric knee-flexor strength may alter running biomechanics (Chumanov, et al., 2007) or reduce the capacity of the weaker limb to decelerate the forward swinging shank during terminal-swing 75

91 (Opar, et al., 2012). However, it should also be noted that the assessment of between-limb imbalance in the current study was performed during a bilateral Nordic hamstring exercise, whereas typical assessments involve maximal unilateral contractions performed on an isokinetic dynamometer (Bennell, et al., 1998a; Fousekis, et al., 2011). For this reason, direct comparisons to previous work should be made with caution. A bilateral Nordic hamstring exercise was employed in the current study as previous work has shown that this is more a more reliable test of eccentric knee-flexor strength than unilateral Nordics (Opar, Piatkowski, et al., 2013a). The finding that weaker players were no more likely to sustain an HSI than stronger players is in line with a recent systematic review and meta-analysis which suggested that isokinetically-derived measures of strength were not a risk factor for HSI in sport (Freckleton & Pizzari, 2013). However, the results of the current study differ from a recent investigation (Opar, et al., 2014) using the Nordic hamstring test which reported that elite Australian footballer s with eccentric strength < 256N at the start of preseason and < 279N at the end of preseason had a 2.7 and 4.3 fold greater risk of HSI, respectively. The disparity between studies might reflect the vastly different anthropometric characteristics of rugby union (Zemski, et al., 2015) and Australian football players (Bilsborough et al., 2014), or the unique physical demands of each sport (Duthie, et al., 2003; Seward, & Orchard, 2013). However, it is also important to consider that the rugby players in the current study were substantially stronger than the Australian footballer s studied previously (Opar, et al., 2014). It is possible that the protective benefits conferred by greater levels of eccentric strength may plateau at higher ends of the strength spectrum as they appear to in Australian footballer s (see Figures 1 & 2 in Opar et al.) (Opar, et al., 2014). It should also be acknowledged that while some studies have found an association between low levels of knee-flexor strength and subsequent 76

92 HSI (Heiser, et al., 1984; Opar, et al., 2014; Yamamoto, 1993), prior injury is also associated with knee-flexor weakness (Croisier, et al., 2002; Opar, Piatkowski, et al., 2013a; Opar, Williams, et al., 2013a; Timmins, Shield, et al., 2014), and this may confound results (Orchard, 2001). The current study supports prior HSI as a risk factor for re-injury which is consistent with earlier observations in rugby union (Brooks, et al., 2006; Upton, et al., 1996) Australian football (Bennell, et al., 1998a; Freckleton, Cook, & Pizzari, 2014; Orchard, et al., 1997b; Warren, Gabbe, Schneider-Kolsky, & Bennell, 2010) and soccer (Arnason, et al., 2004). While the mechanism(s) explaining why prior HSI augments the risk of re-injury remain(s) unclear, this study revealed a significant relationship between prior HSI and between-limb imbalance in eccentric knee-flexor strength. This novel finding suggests that rugby union players with a history of HSI have a significantly greater risk of re-injury if they return to training and match play with one limb weaker than the other (Figure 4-2). For example, an athlete with a prior HSI and a 30% between-limb imbalance in eccentric strength is twice as likely to suffer a recurrence as a previously injured athlete with no imbalance. In light of this interaction, there is a growing body of evidence to suggest that between-limb imbalance in knee-flexor strength (Croisier, et al., 2002; Croisier, et al., 2008c; Jonhagen, et al., 1994; Orchard, et al., 1997b) is a risk-factor for HSI recurrence. These data highlight the multifactorial nature of HSIs and suggest that the amelioration of between-limb imbalances in eccentric knee-flexor strength should be a focus of rehabilitative strategies following HSI. There are some limitations that should be acknowledged in the current study. Firstly, the assessment of eccentric knee-flexor strength and between-limb imbalance was only performed at a single time point in the pre-season period. While this is consistent with other prospective studies exploring the impact of strength variables on HSI risk (Croisier, 77

93 Ganteaume, Binet, Genty, & Ferret, 2008a; Fousekis, et al., 2011; Heiser, et al., 1984; Orchard, et al., 1997b; Yamamoto, 1993), it is important to consider that strength may change over the pre-season and in-season periods (Opar, et al., 2014). The assessment of strength at multiple time points may provide a more robust measure of player risk however, the geographic diversity of the Super 15 competition precluded follow-up assessments by the investigators. Eccentric strength was measured as a force output (N) rather than a joint torque (Nm) which makes direct comparison to isokinetically-derived measures difficult. Further, this mode of testing does not allow for an assessment of the angle at which the knee flexors produce maximum torque (Brockett, et al., 2004), and did not permit force to be expressed relative to quadriceps (Croisier, et al., 2008c) or hip flexor (Yeung, et al., 2009) strength, which may provide additional information on an athlete s risk of HSI. The lack of player exposure data also prevents HSI rates being expressed relative to the amount of training and match-play that athletes were undertaking. Future work should seek to clarify the effect of total exposure time (particularly to high-speed running) on the incidence of HSI in rugby union players (Brooks, et al., 2006). Finally, it should be acknowledged that the area under the curve (AUC) for the multivariate logistic regression model was 0.69 and while it displayed a moderate to high ability to identify those athletes at risk of HSI (sensitivity = 0.6), the specificity (0.2) of the model was low and this should be considered when interpreting the current data. In conclusion, this study suggests that both between-limb imbalances in eccentric knee-flexor strength and prior HSI are associated with an increased risk of future HSI in rugby union. However, lower levels of eccentric knee-flexor strength and a recent history of other lower limb injuries do not significantly increase the risk of future HSI in this cohort. This study, along with previous findings (Opar, et al., 2014), highlights the multifactorial nature of HSI 78

94 and supports the rationale for reducing imbalance, particularly in players who have suffered a prior injury within the previous 12 months. 79

95

96 Chapter 5: STUDY 2 REDUCED ACTIVATION OF PREVIOUSLY INJURED BICEPS FEMORIS LONG HEAD MUSCLES IN RUNNING Publication statement This chapter is comprised of the following paper which was submitted for review at Medicine and Science in Sports and Exercise: Bourne, MN., Opar, DA., Williams, MD., Al Najjar, A., Kerr, G., & Shield, AJ. (2015). Reduced activation of previously injured biceps femoris long head muscles in running. Med Sci Sports Ex, Submitted. 81

97 Statement of Contribution of Co-Authors for Thesis by Published Paper The authors listed below have certified* that: 1. They meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. They take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. There are no other authors of the publication according to these criteria; 4. Potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit 5. They agree to the use of the publication in the student s thesis and its publication on the Australasian Research Online database consistent with any limitations set by publisher requirements. Contributor Matthew Bourne 08/03/2016 David Opar Aiman Al Najjar Morgan Williams Graham Kerr Anthony Shield Statement of contribution* Experimental design, ethical approval, data collection and analysis, statistical analysis, manuscript preparation Aided in experimental design and manuscript preparation Aided in data collection Assisted with statistical analysis and manuscript preparation Assisted with manuscript preparation Aided in experimental design and manuscript preparation Principal Supervisor Confirmation I have sighted from all co-authors confirming their certifying authorship. Dr Anthony Shield 82

98 5.1 LINKING PARAGRAPH The study in Chapter 4 demonstrated that athletes with between-limb imbalances in eccentric knee flexor strength, and those with a history of HSI, are at a significantly elevated risk of future HSI. Moreover, for those athletes who have been injured previously, the risk of reinjury is amplified when they return to sport with between-limb strength imbalances. This study highlighted the multifactorial nature of HSI and supports the rationale for reducing strength imbalances, particularly in those players who have a history of injury. However, the mechanism(s) underpinning these strength imbalances are poorly understood. Using fmri, we have recently found that previously injured hamstring muscles are activated less completely than uninjured contralateral muscles during the Nordic hamstring exercise (Bourne, et al, 2015). However, it remains to be seen if these deficits also exist during the presumably injurious task of high-speed overground running. 83

99 5.2 ABSTRACT PURPOSE: It is unknown whether elite athletes with a history of strain injury display altered patterns of hamstring muscle activation during maximal speed sprinting, or reduced muscle size. The goals of this study were to determine: 1) the spatial patterns of hamstring muscle activation during high-speed overground running in limbs with and without a prior hamstring strain injury and; 2) whether previously injured hamstring muscles exhibit lasting deficits in cross-sectional area (CSA). METHODS: Ten elite male athletes with a history of unilateral biceps femoris long head (BFLH) strain injury underwent functional magnetic resonance imaging before and immediately after a repeat-sprint running protocol. Transverse relaxation times of the BFLH and short head, semitendinosus, and semimembranosus were measured before and immediately after exercise and CSA was measured at rest. RESULTS: Previously injured BFLH muscles displayed a significantly lower percentage increase in transverse relaxation time after the running protocol than uninjured contralateral BFLH muscles (mean difference = 12.0%, p < 0.001). In the uninjured control limb the biceps femoris short head was significantly less active than the BFLH (mean difference = 18.1%, p < 0.001) and the semitendinosus (mean difference = 18.1%, p < 0.001). No participant reported any pain in the posterior thigh before or after exercise. No between-limb differences in CSA were observed for any hamstring muscles. CONCLUSION: Elite athletes with a prior strain injury to the BFLH display altered patterns of muscle activation in their previously injured limbs during maximal speed overground running; these differences exist in the absence of pain and atrophy and despite a full return to pre-injury levels of training and competition. 84

100 5.3 INTRODUCTION Hamstring strains are endemic in sports that involve high-speed overground running and represent the most common injury in track and field (Opar, Drezner, et al., 2013), Australian rules football (Orchard, et al., 2013) and soccer (Ekstrand, et al., 2011b) and the most prevalent non-contact injury in rugby union (Brooks, et al., 2006). High rates of recurrence are arguably the most concerning aspect of these injuries, particularly given the tendency for recurrences to result in more time-loss than the initial insult (Koulouris, et al., 2007). Hamstring strain injury is commonly suffered when athletes run at maximal speeds (Askling, et al., 2007) and ~80% of these injuries effect the BFLH (Connell et al., 2004; Koulouris, et al., 2007). Studies employing surface electromyography (semg) suggest that the hamstrings are most active (Thelen, Chumanov, Hoerth, et al., 2005) during the ostensibly injurious (Schache, et al., 2009; Thelen, Chumanov, Hoerth, et al., 2005) late-swing, where they actively lengthen to decelerate the forward swinging shank. However, while these studies have provided important insight into the temporal patterns of hamstring muscle use during high-speed running, the contribution of individual hamstring muscles is not well understood. Further, it remains unclear as to whether the spatial patterns of muscle activation are altered following an HSI. Fyfe and colleagues (Fyfe, et al., 2013b) have proposed that high rates of HSI recurrence might be partly explained by chronic neuromuscular inhibition of the previously injured muscle. In support of this, reduced semg activity has been observed in previously injured BF muscles during maximal (Opar, Williams, et al., 2013a) and rapid (Opar, Williams, et al., 2013b) eccentric isokinetic knee flexor contractions. Further, previously injured hamstring muscles were found to be significantly less active than uninjured contralateral muscles during 85

101 the performance of the bilateral eccentric Nordic hamstring exercise (Bourne, et al, 2015). It is plausible that activation deficits, that persist long after rehabilitation and the return to training and competition, might mediate preferential eccentric knee flexor weakness (Croisier, et al., 2008c; Lee, et al., 2009), reduced rates of eccentric knee flexor torque development (Opar, Williams, et al., 2013b), lasting BFLH atrophy (Silder, et al., 2008) and a chronic shortening of BFLH fascicles (Timmins, Shield, et al., 2014), all of which have been reported for previously injured hamstring muscles. However, these activation deficits have only been noted during single joint exercises that do not replicate the high-velocity and multijoint demands of high-speed overground running. An improved understanding of the spatial patterns of hamstring muscle activation during high-speed running, particularly in previously injured limbs, will be important in optimising rehabilitation programs and may have implications for understanding the mechanisms of running-induced HSI. Functional magnetic resonance imaging (fmri) is a validated (Adams, Duvoisin, & Dudley, 1992; Cagnie et al., 2011; Fleckenstein, Canby, Parkey, & Peshock, 1988) and highly reliable (Cagnie et al., 2008; Cagnie, et al., 2011) measure of skeletal muscle activation during exercise that allows for a concurrent assessment of muscle morphology. The premise of using fmri to assess muscle activation is based on signal intensity changes in fmr images resulting from a transient increase in the transverse (T2) relaxation time of muscle water following exercise (Cagnie, et al., 2011). These T2 shifts increase proportionately to exercise intensity (Fleckenstein, et al., 1988) and are consistent with electromyographical measures of muscle activity (Adams, et al., 1992; Cagnie, et al., 2011). However, the unique ability of fmri to non-invasively assess deep muscles at multiple sites within a single scan overcomes several spatial limitations associated with EMG (Adams et al., 1992). As such, fmri has become a popular tool for the assessment of muscle 86

102 use during exercise (Bourne, et al, 2015; Cagnie, et al., 2008; Mendiguchia et al., 2012; Ono, et al., 2011; Schuermans, Van Tiggelen, Danneels, & Witvrouw, 2014; Sloniger, Cureton, Prior, & Evans, 1997) with great potential to demonstrate aberrant activation patterns following injury (Bourne, et al, 2015; Patten, Meyer, & Fleckenstein, 2003). This study employed fmri on elite athletes with a history of unilateral HSI to the BFLH who had since undergone apparently successful rehabilitation and returned to their pre-injury level of competition. The primary purpose of this investigation was to determine the spatial patterns of hamstring muscle activation during high-speed overground running in limbs with and without a history of HSI. A secondary aim was to determine whether previously injured hamstrings exhibit chronic alterations in muscle morphology. We hypothesised that the hamstrings of uninjured limbs would be activated non-uniformly during high-speed running and that previously injured muscles would show reduced activation, and reduced crosssectional area (CSA), relative to uninjured contralateral muscles. 5.4 METHODS Experimental Design This observational study employed a cross-sectional design in which all participants completed a single testing session. Prior to testing, participants provided a detailed injury history to investigators with reference to imaging findings and clinical notations from the practitioner who diagnosed and treated their most recent HSI. Subsequently, participants underwent an fmri scan of their thighs before and immediately after a repeat-sprint running protocol. Participants were asked to rate their level of perceived pain in the posterior thigh before and after running using a visual analogue scale (VAS). 87

103 Participants Ten elite male athletes (age, 23.2 ± 5.3 years; height, ± 4.2 cm; weight, 82.2 ± 5.4 kg) currently competing in a running based sport and who had suffered a unilateral grade II strain injury to the BFLH within the previous 12 months (mean time of 8.7 ± 3.2 months since the last insult) were recruited (Table 5-1). A sample size of 10 was deemed sufficient to detect an effect size of 1.0 in T2 relaxation time and CSA between muscles and limbs, at a power of 0.80 and with p<0.05 (Bourne, et al, 2015). All athletes had returned to their pre-injury levels of training and competition (involved in 18 ± 4 hours of structured training per week) with eight participants competing at a national to international level in track and field, one competing at a state level in rugby union (backline) and one at a state level in soccer at the time of testing. Participants completed an injury history questionnaire with reference to clinical notes provided by their physical therapist. All had their diagnosis confirmed with MRI (n=8) or ultrasound (n=2) at the time of injury and had subsequently completed a standard progressive intensity rehabilitation program (Heiderscheit, et al., 2010) under the guidance of their physical therapist. Participants were free of orthopaedic abnormalities of the lower limbs, had no history of neurological or motor disorders and had no other soft tissue injuries to the thighs at the time of testing. All completed a cardiovascular risk factor questionnaire (Appendix B) to ensure it was safe for them to perform intense exercise and a standardised MRI screening questionnaire (Appendix C) provided by the imaging facility to make certain that it was safe for them to enter the magnetic field. All athletes provided written informed consent to participate in this study, which was approved by the Queensland University of Technology Human Research Ethics Committee and the University of Queensland Medical Research Ethics Committee. 88

104 Table 5-1. Hamstring strain injury details for all participants (n=10) Participant Muscle Dominant Severity of Months Rehabilitation injured limb injured last injury since last period (grade 1-3) injury (weeks) 1 BFLH Yes BFLH No BFLH No BFLH Yes BFLH No BFLH Yes BFLH No BFLH No BFLH Yes BFLH Yes Rehabilitation was defined by a return to pre-injury levels of training and competition. BFLH, biceps femoris long head. Experimental Session Repeat-sprint running protocol Participants completed three sets of six maximal intensity 40m sprints (with an additional 10m acceleration and 15m deceleration distance) on a flat grass sports field adjacent to the imaging facility. Participants were provided with 30s of rest between sprints and one minute rest between sets. Investigators verbally encouraged maximal effort throughout each interval. Participants ran towards the entrance of the imaging facility on the final repetition of the 89

105 running protocol and were returned to the scanner immediately (<30s) following the cessation of exercise. Localiser adjustments began within 98 ± 8s (mean ± SD) and post-exercise T2- weighted imaging began within 2min of exercise. Functional magnetic resonance imaging All fmri scans were performed using a Siemens 3-Tesla (3T) TrioTim imaging system with a spinal coil. The participant was positioned supine in the magnet bore with their knees fully extended and hips in neutral and straps were positioned around both limbs to prevent any undesired movement. Consecutive T2- and contiguous T1-weighted transaxial MR images were taken of both limbs beginning at the level of the iliac crest and finishing distal to the tibial plateau. All images were collected using a 180 x 256 image matrix and a 400 x 281.3mm field of view. T2-weighted images were used to assess the extent of hamstring activation during exercise and were acquired pre- and immediately post-exercise using a Car- Purcel-Meiboom-Gill (CPMG) spin-echo pulse sequence (transverse relaxation time = 2000ms; echo time = 10, 20, 30, 40, 50 and 60ms; number of excitations = 1; slice thickness = 10mm; interslice gap = 10mm) (Bourne, et al, 2015). T1-weighted spin-echo images were acquired only during the pre-exercise scan (transverse relaxation time = 1180ms; echo time = 12ms; field of view = 400 x mm; number of excitations = 1; slice thickness = 10mm; interslice gap = 0mm); these images have substantially greater contrast than T2-weighted images and were used to accurately assess muscle CSA. The total acquisition time for preexercise images was 15 minutes and 10s and for post-exercise images, 10 minutes. To minimise any inhomogeneity in MR images caused by dielectric resonances at 3T, a B1 filter was applied to all scans (de Sousa, Vignaud, Fleury, & Carlier, 2011); this is a postprocessing image filter that improves the image signal intensity profile without affecting the 90

106 image contrast. In addition, participants were asked to avoid strength training of the lower limbs for 72 hours prior to scanning as muscle damage may augment resting T2 values. Lastly, to reduce the effects of intramuscular fluid shifts before the pre-exercise scans, participants were seated for a minimum of 15 minutes before data acquisition (Bourne, et al, 2015). Visual analogue scale Before and immediately following the cessation of the repeat-sprint running protocol, participants were asked to rate their level of pain and discomfort in the posterior thigh (if any) on a VAS. Participants were instructed to choose a number between 0 (no pain) and 10 (unbearable pain). Data analysis All fmr images were transferred to a personal computer in the DICOM file format and image analysis software (Sante Dicom Viewer and Editor, Cornell University) was used for subsequent analysis. The analysis of T2 relaxation time and muscle CSA was performed on each hamstring muscle (BFLH, BFSH, ST and SM) for both the previously injured and uninjured contralateral limbs in 10mm thick slices corresponding to 30, 40, 50, 60 and 70% of the distance between the inferior margin of the ischial tuberosity (0%) and the superior border of the tibial plateau (100%) (Ono, et al., 2011). The T2 relaxation times of each hamstring muscle were measured in T2-weighted images acquired before and after exercise to evaluate the degree of muscle activation during the repeat-sprint running protocol. At each slice, the signal intensities of the BFLH, BFSH, ST and SM were measured in both limbs using 10mm 2 rectangular regions of interest (ROI) (Mendiguchia, Garrues, et al., 2012) in each muscle (total of eight ROIs in each slice). Each ROI was placed in a homogenous region of 91

107 muscle tissue, with great care taken to avoid aponeurosis, tendon, bone and blood vessels and was selected at the same coordinates within each muscle for the pre- and post-exercise scans. An ROI approach was deemed most appropriate as this method allowed investigators to avoid any areas of residual scar tissue associated with prior HSI (Silder, et al., 2008). The signal intensity reflected the mean value of all pixels within the ROI and was determined for each ROI across six echo times (10, 20, 30, 40, 50 and 60ms). T2 relaxation time for each ROI was then calculated by fitting signal intensity values at each echo time to a mono-exponential decay model using a least squares algorithm: [(SI= M exp(echo time / T2) (Bourne, et al, 2015; Ono, et al., 2011) where SI is the signal intensity at a specific echo time, and M represents the pre-exercise fmri signal intensity. To assess the degree of muscle activation during exercise, the mean percentage change in T2 for each ROI was calculated as: [(mean post-exercise T2 / mean pre-exercise T2) x 100]. To provide a meaningful measure of whole-muscle activation, the percentage change in T2 relaxation time for each hamstring muscle was evaluated using the average value of all ROIs (at all five thigh levels). Previous studies have demonstrated excellent intertester reliability of T2 relaxation time measures with intra-class correlation coefficients ranging from 0.87 to 0.94 (Cagnie, et al., 2008; Cagnie, et al., 2011). Muscle CSAs obtained from pre-exercise T1-weighted images were analysed to examine differences in hamstring muscle morphology in limbs with and without a history of strain injury to the BFLH. The muscle boundaries of BFLH, BFSH, ST and SM were traced manually on each limb (at slices 30, 40, 50, 60 and 70% of thigh length)(24) and CSA was calculated as the total cm 2 within each trace. For each muscle, the average CSA of all slices in the 92

108 previously injured limb was compared with that of the uninjured contralateral limb to evaluate between-limb differences following injury. Statistical Analysis All statistical analyses were performed using JMP version (SAS Institute Inc, 2012). A repeated measures design linear mixed model fitted with the restricted maximum likelihood (REML) method was used to compare transient exercise-induced percentage changes in T2 relaxation times and resting values of CSA for each muscle in the previously injured and uninjured contralateral limbs. Muscle (BFLH, BFSH, ST or SM), limb (injured/uninjured) and muscle by limb interaction were the fixed factors with participant identity (ID), participant ID by muscle and participant ID by limb as the random factors. When the muscle by limb interaction was significant (p<0.05) for the percentage change in T2 relaxation time or for CSA, post-hoc t-tests with Bonferroni corrections were used to report the mean difference between limbs for each muscle with 95% confidence intervals (95%CI). To determine the spatial activation patterns of healthy uninjured hamstrings, the percentage change in T2 relaxation time was compared between each hamstring muscle in the uninjured limb. Again, when a significant main effect was detected, post hoc t tests with Bonferroni corrections were used to report the mean difference (and 95% CI) between muscles. The adjusted alpha level for all post hoc t tests was set at p < To assess the potential impact of acute posterior thigh pain on between-limb differences in muscle activation, VAS scores obtained from participants before and after the repeat-sprint running protocol were reported descriptively as means ± SD. Finally, given the potential for muscle activation to improve over time, the magnitude of between-limb differences in T2 93

109 relaxation time were correlated with the time since injury and duration of the rehabilitation period using the coefficient of determination. 5.5 RESULTS Comparison of muscle activation between previously injured and uninjured contralateral limbs The between-limb analysis of muscle activation revealed a significant muscle by limb interaction (p < 0.001). Previously injured BFLH muscles displayed a smaller running-induced percentage increase in T2 relaxation time (Figure 5-1 & 5-2A) than the homonymous muscles in the uninjured contralateral limb (mean difference = 12.0%, 95% CI = 2.4 to 21.6%, p < 0.001). In contrast, there were no significant differences in the percentage change in T2 relaxation times for ST (mean difference = 7.4%, 95% CI = -2.2% to 17.0%, p = 0.022), BFSH (mean difference = 2.9%, 95% CI = -6.7 to 12.5%, p = 0.334) or SM (mean difference = - 2.6%, 95% CI = to 7.3%, p = 0.396) muscles in injured and uninjured limbs (Figure 5-1). 94

110 Figure 5-1. Mean percentage change in fmri T2 relaxation times after running for each hamstring muscle in previously injured (Inj) and uninjured (Uninj) limbs. * indicates a significant between limb difference for individual muscles (p<0.005). Error bars depict 95% CIs. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. 95

111 Figure 5-2. A. Parametric map of transverse (T2) relaxation times for the previously injured and uninjured contralateral limbs of a single participant, acquired immediately following the high-speed running protocol. Colour spectrum illustrates the absolute T2 value in milliseconds (ms). Note the divergence between the previously injured and uninjured contralateral limbs. 2B. Typical T1-weighted image at 50% of thigh length (transverse relaxation time = 1180ms; echo time = 12ms; slice thickness = 10mm), depicting the regions of interest for each hamstring muscle. For both A & B, the right side of the image corresponds to the participant s left side as per radiology convention. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. 96

112 Comparison of muscle cross-sectional area between previously injured and uninjured contralateral limbs For muscle CSA, the interaction of muscle by limb was not significant (p = 0.345) when comparing homonymous muscles in uninjured and previously injured limbs. There were no significant between-limb differences in CSAs of homonymous muscles (BFLH mean difference = -0.7 cm 2, 95% CI = -2.6 to 1.2 cm 2, p = 0.248; BFSH mean difference = -0.3 cm 2, 95% CI = -2.1 to 1.50 cm 2, p = 0.589); ST mean difference = -0.6 cm 2, 95% CI = -2.5 to 1.3 cm 2, p = 0.303); SM mean difference = -0.6 cm 2, 95% CI = -2.4 to 1.2 cm 2, p = 0.293) (Figure 5-3). Figure 5-3. Mean CSAs (cm 2 ) of each hamstring muscle for both the previously injured (Inj) and uninjured (Uninj) contralateral limbs (BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus). Error bars depict 95% CIs. 97

113 Spatial activation patterns of the uninjured limb For the analysis of muscle activation in the uninjured limb, there was a significant main effect for muscle (p<0.001) (Figure 5-4). Post-hoc t-tests revealed that BFSH was significantly less active than BFLH (mean difference = 18.1%, 95% CI = 3.3 to 32.9%, p < 0.001) and ST (mean difference = 18.1%, 95% CI = 3.3 to 32.9%, p < 0.001). No significant differences in the percentage change in T2 relaxation time were observed for BFSH vs. SM (mean difference = 9.8%, 95% CI = -5.0 to 24.6%, p = 0.048), SM vs. BFLH (mean difference = 8.3%, 95% CI = -6.5 to 23.1%, p = 0.090), SM vs. ST (mean difference = 8.3%, 95% CI = -6.5 to 23.1%, p = 0.090), or ST vs. BFLH (mean difference = 0.0%, 95% CI = -9.7 to 9.7% p = 0.999). Figure 5-4. Percentage change in fmri T2 relaxation times of each hamstring muscle in the uninjured limb. Values are expressed as a mean percentage change compared to the values at rest. * indicates significantly different from BFSH (p<0.005). Error bars depict 95% CIs. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. 98

114 Pain and discomfort No participant reported any pain in the posterior thigh before or immediately following the repeat-sprint running protocol (mean VAS scores pre- and post-exercise = 0.0 ± 0). Time elapsed since injury and injury severity There was no relationship between the time elapsed since injury and the magnitude of between-limb differences in the percentage change in T2 relaxation time for BFLH (R 2 = 0.001) or ST (R 2 = 0.02). Furthermore, there was no correlation between the duration of rehabilitation and the extent of between-limb differences in T2 change for BFLH (R 2 = 0.05) or ST (R 2 = 0.002). 99

115 5.6 DISCUSSION This study is the first to use fmri to map the spatial activation patterns of hamstring muscles during high-speed overground running in a homogenous group of highly trained athletes. The results suggest that healthy uninjured hamstrings were activated somewhat uniformly during sprinting, with the exception of BFSH. However, previously injured BFLH muscles were activated ~50% less than homonymous contralateral muscles with no history of injury. These differences were present after rehabilitation, the return to pre-injury levels of training and competition and in the absence of pain. Evidence for altered activation patterns in previously injured hamstring muscles is in line with previous findings (Bourne, et al, 2015; Daly, McCarthy Persson, Twycross-Lewis, Woledge, & Morrissey, 2015; Opar, Williams, et al., 2013a, 2013b; Sole, et al., 2011a). A recent fmri investigation showed that previously injured hamstrings were significantly less active than uninjured contralateral muscles during the Nordic hamstring exercise (Bourne, et al, 2015). As in the current study, these activation deficits were present long after a return to sport (~10 months post-injury) and were present without discrepancies in muscle CSA (Bourne, et al, 2015). Earlier work employing semg and isokinetic dynamometry reported inhibition during eccentric actions in previously injured BF muscles (Opar, Williams, et al., 2013a, 2013b) many months after a return to sport, which suggests that this is likely a robust phenomenon. Most recently, reduced BF electromyographic activity relative to other hip and trunk muscles has been reported during treadmill running at 20km.h -1 in athletes with a unilateral history of HSI (Daly, et al., 2015). However, only the current study provides insights into the activation patterns of all the hamstring muscles in previously injured and uninjured limbs during overground running. Furthermore, the observation of activation deficits in BFLH muscles during high-speed running has direct implications for clinical 100

116 practice because this is the activity most often associated with injury (Askling, et al., 2007; Koulouris, et al., 2007; Schache, et al., 2009) and this muscle is the most common site of injury (Connell, et al., 2004; Koulouris, et al., 2007), It is also noteworthy that pain-free highspeed running is often used as a clinical marker for return to play (Heiderscheit, et al., 2010), although the activation deficits observed in this cohort persisted in the absence of pain and despite a complete return to pre-injury levels of training and competition. In the current study, a moderate effect size (0.6) was observed for the comparison between ST muscles in previously injured and uninjured limbs which raises the prospect that this study was insufficiently powered to identify this unexpected difference. Future work should determine whether or not BFLH injury is associated with reduced ipsilateral ST use. The current results do diverge from one recent fmri investigation (Schuermans, et al., 2014). Schuermans and colleagues (Schuermans, et al., 2014) reported an increased percentage change in T2 relaxation and more symmetrical muscle use in previously injured hamstrings relative to uninjured hamstrings following ~255s of exhaustive leg curl exercise. The authors proposed that the higher T2 shift represented a greater metabolic demand on the active tissue during sub-maximal exercise and by extension, a reduced strength endurance capacity of previously injured hamstrings. However, the prone leg curl exercise in this study (Schuermans, et al., 2014) was performed with submaximal loads (5kg) and only involved movement at the knee, which may not reflect the high-intensity and high-velocity multi-joint demands of sprinting. It is possible that previously injured hamstrings may respond very differently to submaximal loading, given that prior observations of reduced hamstring activation have been noted during high-velocity (Daly, et al., 2015; Sole, et al., 2011a) or maximal eccentric efforts (Bourne, et al, 2015; Opar, Williams, et al., 2013a, 2013b). However, in the work by Schuermans and colleagues (Schuermans, et al., 2014) injuries were 101

117 self-reported and heterogeneous in terms of location and severity which may have confounded the results. Further, the analysis of muscle activation was limited to a 35mm region in the distal thigh, which may not reflect whole-muscle use as accurately as the current investigation. It has been proposed that the acute pain and discomfort associated with HSI results in centrally-mediated neural changes that chronically reduce activation of previously injured muscles (Fyfe, et al., 2013b; Opar, et al., 2012). While acute inhibition is a normal response to pain and injury that may protect the injured structures from further damage (Lund, et al., 1991), it may also result in a permanent redistribution of motor activity within and between muscles (Hodges & Tucker, 2011) if not adequately addressed in rehabilitation. Activation deficits in previously injured hamstrings during high-speed running may reduce the ability of these muscles to generate high levels of force (Daly, et al., 2015; Opar, Williams, et al., 2013a, 2013b), particularly during the seemingly injurious (Schache, et al., 2009; Thelen, Chumanov, Hoerth, et al., 2005) terminal-swing phase of running where hamstring stresses are greatest (Chumanov, et al., 2011; Nagano, Higashihara, Takahashi, & Fukubayashi, 2014; Schache, Dorn, Blanch, Brown, & Pandy, 2012). Prior hamstring injury has also been reported to result in deficits in horizontal ground reaction forces during high speed running (Brughelli, Cronin, Mendiguchia, Kinsella, & Nosaka, 2010; Mendiguchia et al., 2014). Whether our athletes had similar deficits is unknown, although the observation of reduced muscle activation offers a potential explanation for lasting deficits in force production. The current study found no evidence of atrophy in previously injured hamstring muscles as has been reported previously (Bourne, et al, 2015). Sanfilippo and colleagues (Sanfilippo, Silder, Sherry, Tuite, & Heiderscheit, 2013) also found no evidence of atrophy in previously 102

118 injured BFLH muscles six months following the completion of a standardised hamstring rehabilitation program. However, chronically reduced BFLH volume and an apparently compensatory hypertrophy of the ipsilateral BFSH has been reported after BFLH strains (Silder, Heiderscheit, Thelen, Enright, & Tuite, 2008). These divergent findings may reflect differences in training practices and competition level. It is also possible that alterations to muscle architecture following HSI may reduce the sensitivity of CSA measures to muscle size. Timmins and colleagues (Timmins, Shield, et al., 2014) recently reported that previously injured BFLH muscles exhibited significantly shorter fascicles and greater pennation angles than homonymous muscles in the uninjured contralateral limb. This increase in pennation angle would tend to counter any effects of muscle atrophy on measures of muscle thickness, so measures of CSA or thickness may not be as sensitive to atrophy as are measures of muscle volume. Further work is required to understand the effect of prior HSI on hamstring morphology, particularly in highly trained athletes. This study and previous work (Bourne, et al, 2015; Sanfilippo, et al., 2013) suggest that structured, progressive intensity rehabilitation programs may be effective at maintaining muscle cross-sections after HSI while not adequately addressing activation deficits. The current results suggest that in limbs with no history of HSI, the two-joint hamstring muscles are activated rather uniformly during sprint running while BFSH is activated significantly less than the BFLH and ST. Sloniger and colleagues (Sloniger, et al., 1997) have previously reported very similar levels of BF, ST and SM muscle use according to T2 fmri changes after exhaustive treadmill running in recreationally active females, which is largely in line with findings of the current investigation. However, Sloniger and colleagues (Sloniger, et al., 1997) did not differentiate between the long and short heads of BF, which appear to display distinct activation magnitudes during running. 103

119 With regard to limitations of the current study, it is acknowledged that the retrospective nature of this experiment means that it is not possible to determine whether activation deficits in previously injured muscles were the cause or result of injury. Future prospective studies should seek to clarify whether poor hamstring activation is associated with an increased risk of HSI, although the expense of fmri renders this methodology impractical for such study designs. Also, given the absence of a control group without a history of HSI in either limb, it is impossible to know whether participants had normal patterns of muscle activation in their uninjured limbs. It is also important to consider that the T2 response to an exercise stimulus is highly dynamic and can be influenced by a range of factors such as the metabolic capacity and vascular dynamics of the active tissue (Patten, et al., 2003). We attempted to minimise these factors through strict inclusion criteria (ie. recruiting male athletes of a similar age with homogenous injury location and comparable training status), however, it was assumed that these factors would not differ significantly between previously injured and uninjured muscles. This study provides novel insights into hamstring muscle use during high-speed running in athletes with a unilateral history of BFLH injury. Future work should aim to clarify whether the inhibition observed here causes or results from HSI. Identifying the methods by which this deficit can be ameliorated should also be prioritised. 104

120 Chapter 6: STUDY 3 IMPACT OF EXERCISE SELECTION ON HAMSTRING MUSCLE ACTIVATION Publication statement This chapter is comprised of the following paper which has been accepted for publication at the British Journal of Sports Medicine: Bourne, MN., Williams, MD., Opar, DA., Al Najjar, A., & Shield, AJ. (2016). Impact of exercise selection on hamstring muscle activation. Br J Sports Med, Accepted. 105

121 Statement of Contribution of Co-Authors for Thesis by Published Paper The authors listed below have certified* that: 1. They meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. They take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. There are no other authors of the publication according to these criteria; 4. Potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit 5. They agree to the use of the publication in the student s thesis and its publication on the Australasian Research Online database consistent with any limitations set by publisher requirements. Contributor Matthew Bourne 08/03/2016 David Opar Aiman Al Najjar Morgan Williams Anthony Shield Statement of contribution* Experimental design, ethical approval, data collection and analysis, statistical analysis, manuscript preparation Aided in experimental design and manuscript preparation Aided in data collection Assisted with statistical analysis and manuscript preparation Aided in experimental design and manuscript preparation Principal Supervisor Confirmation I have sighted from all co-authors confirming their certifying authorship. Dr Anthony Shield 106

122 6.1 LINKING PARAGRAPH Fyfe et al. (2013) have recently proposed that high rates of HSI recurrence might be partly explained by chronic neuromuscular inhibition which results in a reduced capacity to voluntarily activate the BFLH muscle. Indeed, Chapter 5 demonstrated that previously injured BFLH muscles were activated ~50% less than homonymous contralateral muscles with no history of injury during the presumably injurious (Thelen, Chumanov, Hoerth, et al., 2005) task of high-speed running. These observations are consistent with findings from this student s Honours project showing that previously injured hamstrings were significantly less active than uninjured contralateral muscles during the NHE (Bourne, et al, 2015). Earlier work employing semg and isokinetic dynamometry also reported inhibition during eccentric actions in previously injured BF muscles (Opar, Williams, et al., 2013a, 2013b) which suggests that this is a robust phenomenon. Persistent neuromuscular inhibition of BFLH muscles, many months after rehabilitation and a full return to training and competition, may help to explain observations of persistent atrophy (Silder, et al., 2008) and altered architecture (Timmins, Shield, et al., 2014) in this muscle following injury. These data suggest the possibility that conventional rehabilitation practices may not be adequately targeting the previously injured BFLH. Heavy resistance training offers a practical and potent stimulus for improving voluntary activation (Akima et al., 1999) and evoking hypertrophy (Kraemer, et al., 2002) of skeletal muscle. However, there is an emerging body of evidence (Mendiguchia, Arcos, et al., 2013b; Mendiguchia, Garrues, et al., 2012; Ono, et al., 2011; Zebis et al., 2013) to suggest that different exercises target different portions of the hamstring muscle group and it is possible that some exercises employed in rehabilitation do not optimally target the injured muscle. An improved understanding of the spatial patterns of hamstring muscle activation during 107

123 different exercises may help practitioners to better tailor rehabilitation programs to the site of injury. 108

124 6.2 ABSTRACT To determine the extent to which different strength training exercises selectively activate the commonly injured biceps femoris long head (BFLH) muscle. METHODS: This two-part observational study recruited 24 recreationally active males. Part 1 explored the amplitudes and the ratios of lateral to medial hamstring (BF/MH) normalised electromyography (nemg) during the concentric and eccentric phases of 10 common strength training exercises. Part 2 used functional magnetic resonance imaging (fmri) to determine hamstring T2 relaxation time changes during two exercises which i) most selectively, and ii) least selectively activated the BF in part 1. RESULTS: Eccentrically, the largest BF/MH nemg ratio was observed in the 45 hip extension exercise and the lowest was observed in the Nordic hamstring (NHE) and bent-knee bridge exercises. Concentrically, the highest BF/MH nemg ratio was observed during the lunge and 45 hip extension and the lowest was observed for the leg curl and bent-knee bridge. fmri revealed a greater BFLH to semitendinosus activation ratio in the 45 hip extension than the NHE (p < 0.001). The T2 increase after hip extension for BFLH, semitendinosus and semimembranosus muscles were greater than that for BFSH (p < 0.001). During the NHE, the T2 increase was greater for the semitendinosus than for the other hamstrings (p 0.002). CONCLUSION: This investigation highlights the non-uniformity of hamstring activation patterns in different tasks and suggests that hip extension exercise more selectively activates the BFLH while the NHE preferentially recruits the semitendinosus. These findings have implications for strength training interventions aimed at preventing hamstring injury. 109

125 6.3 INTRODUCTION Hamstring tears are commonly experienced by athletes involved in running-based sports. They are the most prevalent injury in track and field (Opar, Drezner, et al., 2013), Australian Rules football (Orchard & Seward, 2002; Orchard, et al., 2013), and soccer (Ekstrand, et al., 2011b) and up to 30% recur within 12 months (Orchard & Best, 2002). Upwards of 80% of HSIs involve BFLH muscle (Koulouris, et al., 2007; Opar, et al., 2014; Timmins, Bourne, et al., 2015) and most injuries are thought to occur during the late swing phase of high-speed running (Schache, et al, 2009). During this phase of the gait cycle, the BFLH reaches its peak length and develops maximal force while undergoing a forceful eccentric contraction to decelerate the shank for foot strike (Chumanov, et al, 2007), and it is thought that these conditions may at least partly explain its propensity for injury. It has also been reported that prior BFLH injury is associated with a degree of neuromuscular inhibition (Opar, et al, 2013a; Bourne, et al, 2015) and prolonged atrophy (Silder, et al, 2008), which suggests that current rehabilitation practices do not adequately restore function to this muscle. While the aetiology of HSI is multifactorial, it has been proposed that hamstring weakness is a risk factor for future strain injury (Croisier, et al., 2008a; Opar, et al., 2014; Thorborg, 2014) and interventions aimed at increasing strength, particularly eccentric knee flexor strength, have been effective in reducing HSI rates in several sports (Arnason, et al., 2008; Askling, Tengvar, Tarassova, & Thorstensson, 2014; Askling, et al., 2013; Petersen, et al., 2011; van der Horst, Smits, Petersen, Goedhart, & Backx, 2015). However, despite an increased focus on hamstring strength in prophylactic programs (Heiderscheit, et al., 2010), exercise selection is often implemented on the basis of clinical recommendations and assumptions rather than empirical evidence (Guex & Millet, 2013; Malliaropoulos et al., 2012). There is currently a small body of work on the activation patterns of the hamstrings 110

126 during commonly employed exercises. Studies using functional magnetic resonance imaging (fmri) have shown that activation differs within and between hamstring muscles during different tasks (Bourne, et al, 2015; Mendiguchia, Arcos, et al., 2013a; Mendiguchia, Garrues, et al., 2012; Ono, et al., 2011; Ono, et al., 2010). For example, the semitendinosus (ST) appears to be selectively activated during the Nordic hamstring exercise (NHE) (Bourne, et al, 2015) and the eccentric prone leg curl (Ono, et al., 2010), while the semimembranosus (SM) is preferentially recruited during the stiff leg deadlift (Ono, et al., 2011). Surface electromyography (semg) has also been used in the analysis of hamstring exercises (Ditroilo, De Vito, & Delahunt, 2013; Ono, et al., 2011; Ono, et al., 2010; Schoenfeld et al., 2015; Zebis, et al., 2013). However, these studies are sometimes contradictory and are often inconsistent with the results from fmri (Bourne, et al, 2015; Mendiguchia, Arcos, et al., 2013a; Mendiguchia, Garrues, et al., 2012; Ono, et al., 2011; Ono, et al., 2010; Zebis, et al., 2013). The lack of complete agreement between fmri and semg might reflect the different physiological basis of each technique (Cagnie, et al., 2011). Surface EMG amplitude is sensitive to the electrical activity generated by active motor units and is detected by electrodes overlying the skin (Farina, et al., 2004). This provides valuable information on the neural strategies involved during muscle activation with high temporal resolution, but is prone to cross talk (Farina, et al., 2004) and cannot discriminate between closely approximated segments of muscles (Adams, et al., 1992) such as the medial hamstrings (semimembranosus and semitendinosus). By contrast, fmri reflects the metabolic activity associated with exercise (Cagnie, et al., 2011). Muscle activation is associated with a transient increase in the transverse (T2) relaxation time of tissue water, which can be interpreted from signal intensity changes in fmr images. These T2 shifts, which increase in proportion to exercise intensity (Fisher, Meyer, Adams, et al., 1990; Fleckenstein, et al., 111

127 1988), can be mapped in cross-sectional images of muscles and therefore provide significantly greater spatial clarity than semg (Adams, et al., 1992; Cagnie, et al., 2011). An improved understanding of the patterns of hamstring muscle activation during common strength training exercises may enable practitioners to make better informed decisions regarding exercise selection in injury prevention and rehabilitation programs. The purpose of this two-part study was to determine which exercises most selectively activate the BFLH. Part 1 used semg to determine the amplitude and ratio of lateral to medial hamstring activation during 10 commonly employed exercises. Based on these findings, part 2 employed fmri to map muscle activation during two exercises that appeared to a) most selectively; and b) least selectively activate the BF according to semg. We hypothesised that the patterns of hamstring muscle activation would be non-uniform between exercises and that higher levels of BF activity would be observed during hip-extension exercise (Ono, et al., 2011). 6.4 METHODS Participants Twenty-four recreationally active male athletes (age, 24.4 ± 3.3 years, height, ± 6.1 cm, weight, 85.2 ± 13.4 kg) participated in this study. Eighteen athletes (age, 23.9 ± 3.1, height, ± 5.9, weight, 86.0 ± 14.8) participated in part 1 and ten athletes (age, 24.6 ± 4.0, height, ± 7.0, weight, 83.5 ± 8.7) participated in part 2. A priori sample size estimates were based on 1) the capacity to detect a 10% difference in the ratio of BF to MH (BF/MH) semg amplitude between exercises (Zebis, et al., 2013); and 2) an effect size of 1.0 in T2 relaxation time between muscles (Bourne, et al, 2015), at a power of 0.80 and with p<0.05. Participants were free from soft tissue and orthopaedic injuries to the trunk, hips and lower 112

128 limbs at the time of testing and had no known history of cardiovascular, metabolic or neurological disorders. Participants had not suffered an HSI in the previous 12 months and had no history of anterior cruciate ligament injury. Prior to testing, all participants completed a cardiovascular screening questionnaire (Appendix B) to make sure it was safe for them to perform intense exercise and those who were involved in part 2 also completed a standard MRI screening questionnaire (Appendix C) to ensure it was safe for them to enter the magnetic field. All participants provided written, informed consent for this study, which was approved by the Queensland University of Technology Human Research Ethics Committee and the University of Queensland Medical Research Ethics Committee. Study Design This cross-sectional study involved two parts. In the first we explored the semg amplitudes and ratios of BF to medial hamstring (MH) semg activity during ten commonly employed strength training exercises. Based on these findings, part 2 involved an fmri investigation of two exercises which appeared to a) most selectively, and b) least selectively activate the BF muscle during eccentric contractions. PART 1 Prior to testing participants were familiarised with the exercises used in this investigation. All were shown a demonstration of each exercise (Figure 1) and performed several practice repetitions while receiving verbal feedback from the investigators. Once the participant could complete the exercise with appropriate technique, the loads were progressively increased until an approximate 12RM load was determined (unless the exercise was already supramaximal, ie. NHE and glute-ham-raise). On the day of testing, participants reported to the laboratory and were prepared for semg measurement. The testing session began with two 113

129 maximal voluntary isometric contractions (MVICs) for the hamstrings (see below). Subsequently, participants completed a single set of six repetitions of each exercise, each with the predetermined 12RM load, in randomised order. All data were sampled from a random limb, which was the exercised limb during all unilateral movements and all testing sessions were supervised by the same investigator (MNB) to ensure consistency of procedures. Electromyography Bipolar pre-gelled Ag/AgCl semg electrodes (10mm diameter, 15mm interelectrode distance) were used to record electromyographical activity from the BF and MH. The skin of the participants was shaved, lightly abraded and cleaned with alcohol before electrodes were placed on the posterior thigh, midway between the ischial tuberosity and tibial epicondyles. Electrodes were oriented parallel to the line between these two landmarks, as per SENIAM guidelines (Hermens, Freriks, Disselhorst-Klug, & Rau, 2000), and secured with tape to minimise motion artefact. The reference electrode was placed on the ipsilateral head of the fibula. Muscle bellies of the BF and MH were identified via palpation and correct placement was confirmed by observing active external and internal rotation of the knee in 90 of flexion (Opar, Williams, et al., 2013a; Timmins et al., 2014). During all exercise trials, hip and knee joint angles were measured simultaneously with semg data using two digital goniometers. The hip sensor s axis of rotation was aligned with the greater trochanter of the femur and the knee sensor was positioned superficial to the lateral femoral epicondyle. Maximal voluntary contraction Surface EMG activity was recorded during MVICs of the hamstrings using a custom-made device which was fitted with two uniaxial load cells (Opar, Piatkowski, et al., 2013a), 114

130 Participants lay prone with their hips in 0 of flexion and knees fully extended (180 ), with their ankles secured in immoveable yokes and were asked to perform forceful knee flexion while investigators provided strong verbal encouragement. After 1-2 warm-up contractions, participants completed two 3-4sec MVICs, with 30-sec of rest separating each attempt. The contraction that elicited the highest average amplitude for the BF and MH was used to represent the maximal EMG amplitude. Data analysis All semg and joint angle data were sampled at 1 khz through a 16-bit PowerLab 26T AD unit (ADInstruments, New South Wales, Australia) (amplification = 1000; common mode rejection ratio = 10dB) and analysed using LabChart 8.0 (AD Instruments, New South Wales, Australia). Raw semg data were filtered using a Bessel filter (frequency bandwidth = Hz) and then full wave rectified. Joint angle data were used to determine the concentric (lifting) and eccentric (lowering) phases of each repetition for each exercise. For each phase, the filtered semg signal was normalised to values obtained during MVIC and these normalised semg (nemg) values were averaged across the six repetitions. Exercise Protocol The 10 exercises were chosen based on a review of the scientific literature (Mendiguchia, Garrues, et al., 2012; Ono, et al., 2011; Schoenfeld, et al., 2015; Zebis, et al., 2013). They included the bilateral and unilateral stiff-leg deadlift, hip hinge, lunge, unilateral bent and straight knee bridges, leg curl, 45 hip extension, glute-ham-raise and the NHE (Figure 6-1). Unless the exercise was explosive (hip hinge) or supramaximal and eccentric-only (NHE and glute-ham raise) participants completed both the concentric and eccentric phases of each exercise using a 12-RM load at a constant pace (2 s up and 2 s down). 115

131 Figure 6-1. The 10 examined exercises. (a) bilateral stiff-leg deadlift, (b) hip hinge, (c) unilateral stiff-leg deadlift, (d) lunge, (e) unilateral bent knee bridge, (f) unilateral straight knee bridge, (g) leg curl, (h) 45 hip extension, (i) glute-ham-raise, (j) Nordic hamstring exercise (NHE). 116

132 Statistical analysis Data were analysed using JMP version (SAS Institute Inc, 2012). Descriptive statistics were calculated for mean nemg amplitudes of BF and MH for the concentric and eccentric phases of each exercise and an activation ratio was determined by dividing the average BF nemg amplitude by the average MH nemg amplitude (BF/MH); ratios >1.0 indicated that the BF was more active than the MH muscles. For both the concentric and eccentric phases, repeated measures linear mixed models fitted with the restricted maximum likelihood method were used to determine differences between exercises. For this analysis, exercise was the fixed factor and participant identity the random factor. When a significant main effect was observed for exercise, post hoc t-tests with Bonferroni corrections were used to identify the source and reported as mean differences with 95% CIs. For these analyses, the Bonferroni adjusted p value was set at < PART 2 A cross-sectional design was used to map the spatial patterns of hamstring muscle activation during the 45 hip extension and NHE. These exercises were chosen because they a) most selectively (45 hip extension) and b) least selectively (NHE) activated the BF muscle during eccentric contractions according to semg. Participants completed two separate exercise sessions, separated by at least six days (14 ± 5 days), with each session involving one of the aforementioned exercises. Functional MRI scans of both thighs were acquired before and immediately after each exercise bout. All testing sessions were supervised by the same investigator (MNB). 117

133 Exercise Protocol A depiction of the 45 hip extension and NHE can be found in Figure 6-1. All exercise was completed using the same equipment as that used in part 1. Participants completed five sets of 10 repetitions of each exercise with one-minute rest intervals between sets (Ono, et al., 2011), The higher volume of exercise (compared to part 1) was necessary because transient T2 changes reflect fluid shifts associated with glycolysis and have a higher detection threshold than semg (Cagnie, et al., 2011), All subjects completed 50 repetitions successfully. During the rest periods, participants remained in a seated position (for the hip extension exercise) or lay prone (NHE) to minimise activation of the hamstrings. The 45 hip extension exercise was performed unilaterally with a starting load corresponding to each participant s approximate 12-RM (median = 10 kg; range = 10 to 20 kg). However if the participant could no longer complete the exercise with the allocated load, the weight was gradually reduced by increments of 5kg until it could be completed at the desired speed (2 s up and 2 s down), which was controlled by an electronic metronome. The NHE was performed bilaterally with body weight only. Participants received verbal support from the investigators throughout all exercise sessions to promote maximal effort. All participants were returned to the scanner immediately following the cessation of exercise and post-exercise scans began within ± 24 s (mean ± SD). Functional muscle magnetic resonance imaging (fmri) All fmri scans were performed using a 3-Tesla (Siemens TrioTim, Germany) imaging system with a spinal coil. The participant was positioned supine in the magnet bore with their knees fully extended and hips in neutral and straps were secured around both limbs to prevent any undesired movement. Consecutive T2-weighted axial images were acquired of both limbs beginning at the level of the iliac crest and finishing distal to the tibial plateau using a 180 x 118

134 256 image matrix. Images were acquired before and immediately after exercise using a Car- Purcel-Meiboom-Gill (CPMG) spin-echo pulse sequence and the following parameters: transverse relaxation time (TR) = 2540 ms; echo time (TE) = 8, 16, 24, 32, 40, 48 and 56 ms; number of excitations = 1; slice thickness = 10 mm; interslice gap = 10 mm; field of view = 400 x mm). The total acquisition time for each scan was 6 min 24 s. A localiser adjustment (20 s) was applied prior to the first sequence of each scan to standardise the field of view and to align collected images between the pre- and post- exercise scans (Bourne, et al, 2015). To minimise any inhomogeneity in MR images caused by dielectric resonances at 3T, a post-processing (B1) filter was applied to all scans (de Sousa, et al., 2011); this is a post-processing image filter that improves the image signal intensity profile without affecting the image contrast. In addition, to ensure that the signal intensity profile of T2-weighted images was not disrupted by anomalous fluid shifts, participants were seated for a minimum of 15 min (Ono, et al., 2011) before data acquisition. For each exercise session, the T2 relaxation times of each hamstring muscle were measured in T2-weighted images acquired before and after exercise to evaluate the degree of muscle activation during exercise. All fmri scans were transferred to a Windows computer in the digital imaging and communications in medicine (DICOM) file format. The T2 relaxation times of each hamstring muscle (BFLH, BFSH, ST and SM) were measured in five axial slices, corresponding to 30, 40, 50, 60 and 70% of thigh length; these values were determined relative to the distance between the inferior margin of the ischial tuberosity (0%) and the superior border of the tibial plateau (100%) (Bourne, et al, 2015; Ono, et al., 2011). Image analysis software (Sante Dicom Viewer and Editor, Cornell University) was used to measure the signal intensity of each hamstring muscle in the exercised limb in both the pre- and postexercise scans. The signal intensity was measured in each slice using a circular region of 119

135 interest (ROI) (Mendiguchia, Garrues, et al., 2012) which was placed in a homogenous region of contractile tissue in each muscle belly (avoiding fat, aponeurosis, tendon, bone and blood vessels). The size of each ROI varied (0.2 to 5.6 cm 2 ) based on the cross-sectional area and the amount of homogeneous tissue available in each slice. The signal intensity reflected the mean value of all pixels within the ROI and was measured across seven echo times (8, 16, 24, 32, 40, 48, 56ms). To calculate the T2 relaxation time for each ROI, the signal intensity value at each echo time was fitted to a mono-exponential decay model using a least squares algorithm: [(SI= M exp(echo time / T2)] (Ono, et al., 2011) where SI is the signal intensity at a specific echo time, and M represents the pre-exercise fmri signal intensity. To assess the extent to which each ROI was activated during exercise, the mean percentage change in T2 was calculated as: [(mean post-exercise T2 / mean pre-exercise T2) x 100]. To provide a meaningful measure of whole-muscle activation, the percentage change in T2 relaxation time for each hamstring muscle was evaluated using the ROIs (at all five thigh levels). Previous studies have demonstrated excellent reliability of T2 relaxation time measures with intra-class correlation coefficients ranging from 0.87 to 0.94 (Cagnie, et al., 2008; Cagnie, et al., 2011). Statistical analysis Repeated measures linear mixed models fitted with the restricted maximum likelihood (REML) method were used to determine the spatial activation patterns of the hamstring muscles during the 45 hip extension and NHE. The percentage change in T2 relaxation time was compared between each hamstring muscle (BFLH, BFSH, ST and SM) for both exercises. For this analysis, muscle was the fixed factor and both participant identity and participant 120

136 identity x muscle the random factors. When a significant main effect was detected for muscle, post hoc t tests with Bonferroni corrections were used to determine the source; the adjusted p value was set at Given that the two examined exercises differed in intensity and contraction mode(s), it was not appropriate to directly compare the magnitude of the T2 shifts between exercises.(patten, et al., 2003) Instead, repeated measures linear mixed models fitted with the REML method were used to determine differences in the ratio of BF to ST (BFLH/ST and BFSH/ST) and SM to ST (SM/ST) percentage change in T2 relaxation time between exercises. For these analyses exercise was the fixed factor and participant identity the random factor. When a main effect was found for exercise, post hoc t tests were again used to determine the source and reported as mean difference (and 95% CI). Alpha was set at p<0.05 for these analyses. 6.5 RESULTS Levels of hamstring muscle activation Means and standard errors for the average nemg amplitudes of the BF and MH muscles during the 10 exercises can be found in Table 6-1. Average BF muscle activity ranged from 21.4% (lunge) to 99.3% (unilateral straight knee bridge) MVIC during the concentric phase and 10.7% (hip hinge) to 71.9% (NHE) during the eccentric phase. Average MH muscle activity ranged from 18.1% (lunge) to 120.7% (leg curl) during the concentric phase and 11.6% (hip hinge) to 101.8% (NHE) during the eccentric phase. 121

137 Table 6-1. Mean normalised EMG (nemg) amplitudes for the biceps femoris (BF) and medial hamstring (MH) muscles during the concentric and eccentric phases of 10 hamstring strengthening exercises. Data are expressed as means (standard error). Normalised semg Exercise Muscle Concentric Eccentric Bilateral stiff leg deadlift BF 54.6 (7.6) 21.4 (4.9) MH 49.3 (8.0) 18.2 (5.5) Hip hinge BF 44.9 (7.4) 10.7 (4.8) MH 45.7 (7.8) 11.6 (5.4) Unilateral stiff leg deadlift BF 51.9 (7.6) 26.7 (4.9) MH 57.5 (8.0) 27.7 (5.5) Lunge BF 21.4 (7.4) 13.9 (4.8) MH 18.1 (7.8) 16.9 (5.4) Unilateral bent-knee bridge BF 41.9 (7.4) 23.1 (4.8) MH 57.4 (7.8) 24.5 (4.6) Unilateral straight-knee bridge BF 99.3 (7.4) 55.8 (4.8) MH 90.6 (7.8) 54.9 (5.4) Unilateral leg curl BF 87.6 (7.4) 43.7 (4.8) MH (7.8) 54.6 (5.4) 45 hip extension BF 75.6 (7.4) 48.5 (4.8) MH 61.2 (7.8) 37.1 (5.4) Glute-ham-raise BF (5.1) MH (5.7) Nordic hamstring exercise BF (4.8) MH (5.4) The concentric BF/MH activation ratio for each exercise can be found in Figure 6-2a. A significant main effect was detected between exercises (p < 0.001) with post hoc t tests showing that the BF/MH ratio was greater during the lunge than the leg curl (mean difference = 0.81, 95% CI = 0.48 to 1.14, p < 0.001) and bent-knee bridge (mean difference = 0.74, 95% CI = 0.41 to 1.06, p < 0.001). Similarly, the BF/MH ratio was greater in the 45 hip extension exercise than the leg curl (mean difference = 0.62, 95% CI = 0.30 to 0.95, p < 0.001) and bent-knee bridge (mean difference = 0.55, 95% CI = 0.22 to 0.87, p = 0.001). 122

138 Eccentric biceps femoris to medial hamstring (BF:MH) activation ratio The eccentric BF/MH activation ratio for each exercise can be found in Figure 6-2b. A significant main effect was observed for exercise (p < 0.001) with post hoc analyses revealing that the BF/MH ratio was significantly greater in the 45 hip extension than the NHE (mean difference = 0.69, 95% CI = 0.40 to 0.99, p < 0.001), bent-knee bridge (mean difference = 0.69, 95% CI = 0.40 to 0.99, p<0.001), leg curl (mean difference = 0.60, 95% CI = 0.30 to 0.89, p < 0.001) and the glute-ham raise (mean difference = 0.56, 95% CI = 0.25 to 0.87, p < 0.001). No other between-exercise differences were observed once adjusted for multiple comparisons (p > 0.002). 123

139 Figure 6-2. Biceps femoris (BF) to medial hamstring (MH) normalised EMG (nemg) relationship for the (a) concentric and (b) eccentric phases of each exercise. (SDL) Bilateral stiff-leg deadlift, (HH) hip hinge, (USDL) unilateral stiff-leg deadlift, (L) lunge, (bkb) unilateral bent knee bridge, (SKB) unilateral straight knee bridge, (LC) leg curl, (HE) 45 hip extension, (GHR) glute-ham-raise, (NHE) Nordic hamstring exercise. Exercises to the left of the 45 line exhibited higher levels of BF than MH nemg and exercises to the right displayed higher levels of MH than BF nemg. 124

140 Percentage change in T2 relaxation time following the 45 hip extension exercise A significant main effect was observed for muscle (p < 0.001) with post hoc t tests revealing that the exercise-induced T2 changes in the BFSH were significantly lower than those observed for the BFLH (mean difference = 60.65%, 95% CI = to 80.06%, p < 0.001), ST (mean difference = 77.99%, 95% CI = to 97.59%, p < 0.001) and SM muscles (mean difference = 49.81%, 95% CI = to 69.52%, p < 0.001) (Figure 6-3). The T2 change for ST was significantly greater than SM (mean difference = 28.17%, 95% CI = 9.24 to 47.11%, p = 0.005) however, no difference was observed between the BFLH and SM (p = 0.245) or between the BFLH and ST muscles (p = 0.067). Figure 6-3. Percentage change in fmri T2 relaxation times of each hamstring muscle following the 45 hip extension exercise. Values are expressed as mean percentage change compared to values at rest. ** indicates significantly different from ST, BFLH and SM (p<0.001). * indicates significantly different from ST (p=0.005). Error bars depict standard 125

141 error. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. Percentage change in T2 relaxation time following the Nordic hamstring exercise A main effect was detected for muscle (p < 0.001). Post hoc analyses showed that the T2 changes induced by exercise within the ST were significantly larger than those observed for the BFLH (mean difference = 29.84%, 95% CI = to 39.16%, p < 0.001), BFSH (mean difference = 16.19%, 95% CI = 6.39 to 25.99%, p = 0.002) and SM (mean difference = 29.94%, 95% CI = to 39.44%, p < 0.001) muscles (Figure 6-4). In addition, the T2 increase observed for BFSH was significantly greater than for the BFLH (mean difference = 13.65%, 95% CI = 3.88 to 23.40%, p = 0.008) and SM (mean difference = 13.75, 95% CI = 3.81 to 23.68, p = 0.008) muscles. No difference was observed between the BFLH and SM muscles (p = 0.982). 126

142 Figure 6-4. Percentage change in fmri T2 relaxation times of each hamstring muscle following the Nordic hamstring exercise. Values are expressed as mean percentage change compared to values at rest. ** indicates significantly different from BFLH, BFSH and SM (p 0.002). * indicates significantly different from BFLH and SM (p=0.008) Error bars depict standard error. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. Comparison of hamstring activation ratios between exercises When comparing the BFLH/ST ratio a significant main effect was observed for exercise (p < 0.001) with post hoc analyses revealing a significantly greater ratio during 45 hip extension exercise than during the NHE (mean difference = 0.73, 95% CI = 0.59 to 0.87, p < 0.001) (Figure 6-5). 127

143 Figure 6-5. Ratio of biceps femoris long head (BFLH) to semitendinosus (ST) (BFLH/ST) percentage change in fmri T2 relaxation times following the 45 hip extension and the Nordic hamstring exercise (NHE). * indicates a significant difference between exercises (p<0.001). Error bars depict standard error. A significant main effect was also detected for exercise when comparing the BFSH/ST ratio (p < 0.001). Post hoc t tests demonstrated that this ratio was significantly greater during the NHE than during the 45 hip extension exercise (mean difference = 0.42, 95% CI = 0.24 to 0.62, p < 0.001).When comparing the SM/ST ratio a significant main effect was detected for exercise (p < 0.001) with post hoc t tests showing relatively higher ratios during the 45 hip extension than during the NHE (mean difference = 0.51, 95% CI = 0.39 to 0.64, p < 0.001). 128

144 6.6 DISCUSSION The primary aim of this study was to determine movements that most selectively activate the commonly injured BFLH. The results support the hypothesis that hamstring activation patterns differ markedly between exercises and provide evidence to suggest that hip extension exercise more selectively targets the BFLH than the NHE. The NHE has been shown, in a number of studies, (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015), to reduce HSIs in soccer players as long as compliance is adequate (Goode et al., 2015). However, we (Bourne, et al, 2015) and others (Mendiguchia, Arcos, et al., 2013a) have previously reported that the NHE preferentially activates the ST and this might be interpreted as evidence that the exercise is sub-optimal to protect against running-related strain injury. In this study, we have provided EMG evidence which shows, despite the relatively selective activation of the ST, that the lateral hamstrings were still strongly activated during the NHE. Indeed, BF nemg was higher during the NHE than during the eccentric phase of any other exercise and the evidence for this exercise s protective effects (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015) suggests that eccentric actions alone in a training program are sufficient to make the hamstrings more resistant to strain injury. High levels of BF nemg during the NHE are consistent with previous investigations (Zebis, et al., 2013) and are the result of the supramaximal intensity of the exercise, which potentially explains why high levels of BF nemg were also observed in the eccentric glute-ham raise. High levels of BF nemg in concentric actions were observed in several other exercises including the straight-knee bridge, leg curl and the 45 hip extension which corroborates previous observations (Zebis, et 129

145 al., 2013). However, the importance of hamstring activation patterns during concentric actions remains unclear from the perspective of injury prevention. While high levels of nemg are an important stimulus for improving strength and voluntary activation (Kraemer, et al., 2002), exercise selectivity may still have important implications for rehabilitation. For example, inhibition of previously injured BF muscles during eccentric actions has been reported many months after rehabilitation (Opar, et al, 2013a, 2013b; Bourne, et al, 2015), and it has been proposed (Fyfe, et al, 2013) that these deficits might partly explain observations of persistent eccentric knee flexor weakness (Opar, et al, 2013a), BFLH atrophy (Silder, et al, 2008) and a chronic shortening of BFLH fascicles (Timmins, et al, 2015). These data (Bourne, et al, 2015; Timmins, et al, 2015; Opar, et al, 2013a, 2013b; Silder, et al, 2008) are consistent with the possibility that conventional rehabilitation strategies may not adequately target the commonly injured BFLH. Previous studies have shown that the ratio of lateral to medial hamstring (BF/MH) semg varies with foot rotation (Lynn & Costigan, 2009) and differs between exercises (Zebis, et al., 2013). In the current study, the eccentric phase of the 45 hip extension exercise exhibited the greatest BF/MH nemg ratio (1.5 ± 0.1) while the NHE (0.8 ± 0.1) and bent-knee bridge exercises (0.8 ± 0.1) displayed the lowest ratios. These observations were confirmed in the subsequent fmri analysis whereby the ratio of BFLH to ST in the 45 hip extension exercise (0.96 ± 0.09) was markedly higher than that observed for the NHE (0.23 ± 0.08). It is also noteworthy that the eccentric phase of other hip-oriented exercises (straight-knee bridge, unilateral and bilateral stiff-leg deadlift and hip hinge) displayed BF/MH nemg ratios >1.0. In contrast, the eccentric phase of exercises that involved significant movement at the knee (leg curl, gluteham-raise, bent-knee bridge and NHE) had higher levels of medial hamstring nemg (BF/MH ratio <1.0). These data suggest the possibility that hamstring activation strategies during 130

146 eccentric efforts might be partly dependent on the joints involved in each movement. During concentric contractions, the most selective BF activation was observed in the lunge exercise which corroborates a previous fmri investigation (Mendiguchia, Garrues, et al., 2012). However, it is important to consider that the lunge also exhibited the lowest BF nemg amplitude (21.4 ± 7.4%) of any exercise which likely renders it an inadequate stimulus for improving strength or hypertrophy in this muscle (Kraemer, et al., 2002). Interestingly, the exercise that least selectively activated the BF during concentric contractions was the leg curl, which mimics the joint positions and hamstring muscle-tendon lengths experienced in the NHE. The mechanism for higher levels of BFLH activity during hip extension-oriented movements remains unclear, however, it is possible that hamstring muscle moment arms play a role. For example, the BFLH exhibits a larger moment arm at the hip than at the knee (Thelan, et al, 2005) and therefore possesses a greater mechanical advantage at this joint. As a result, the BFLH undergoes significantly more shortening during hip extension than knee flexion. By comparison, the ST displays a larger sagittal plane moment arm at the knee than both BFLH and SM (Thelan, et al, 2005), which may explain its preferential recruitment during movements at this joint, such as the NHE and leg curl exercises. It is also noteworthy that the ST is a fusiform muscle with long fibre lengths and many sarcomeres in series, which potentially makes it well-suited to forceful eccentric contractions (Lieber, et al, 2000) such as those experienced in the NHE. Further work is needed to clarify the mechanisms underpinning these unique strategies of hamstring activation during hip and knee movements. The current findings are different to some others that have investigated hamstring activation patterns during different tasks. For example, Zebis and colleagues (2013) recently reported 131

147 that both the NHE and the prone isokinetic leg curl were performed with very similar levels of ST and BFLH nemg. However, in the current investigation, the NHE and leg curl exercises resulted in more selective activation of the medial hamstrings and, in the case of the NHE, the fmri results also suggest selective use of the ST muscle. Differences between these studies may conceivably be related to participant sex (females (Zebis, et al., 2013) versus males in the current study) and electrode placement. However, it is also important to consider that semg does not have the spatial resolution of fmri and cannot reliably distinguish between neighbouring muscles (Adams, et al., 1992), such as the long and short heads of BF or the ST and SM, which appear to display distinct activation magnitudes (Bourne, et al, 2015; Ono, et al., 2011; Ono, et al., 2010). These data highlight the limitations of relying exclusively on semg to infer strategies of hamstring muscle activation during exercise and suggest the need for more spatially robust methods in future work. In interpreting the results of this study, it is important to consider that semg and fmri techniques measure different aspects of muscle activity. The absence of T2 relaxation time changes in people with McCardle s disease (Fleckenstein et al., 1991) suggests that fmri is sensitive to glycolysis (Meyer & Prior, 2000) and it is thought that the osmotic fluid shifts which persist after exercise and give rise to T2 changes are a consequence of the accumulation of glycolytic metabolites (Patten, et al., 2003). Fortunately, the proportion of Type II glycolytic fibres does not appear to vary across the hamstring muscles (Garrett, Califf, & Bassett, 1984) so this is unlikely to be a confounding factor. However, exercise induced changes in T2 will be influenced by contraction mode because concentric work is markedly less efficient than eccentric work against the same loads (Shellock, Fukunaga, Mink, & Edgerton, 1991) As a consequence, the differences in T2 relaxation time changes after the 45 hip extension exercise which involved concentric and eccentric actions and the 132

148 almost entirely eccentric NHE do not reflect only the levels of voluntary muscle activation. Instead, fmri can offer insights into the relative metabolic activity and reliance upon different hamstring muscles in each exercise. According to fmri, the NHE involves preferential ST use with modest use of the other hamstrings, while the 45 hip extension exercise appears to heavily recruit both the BFLH and ST muscles. These observations are largely consistent with the semg component of this study, which also suggested higher activation of the medial than lateral hamstrings in the NHE and more even activation of the medial and lateral hamstrings in the 45 hip extension. Characterising the activation patterns of the hamstrings during different tasks is an important first step in identifying exercises worthy of further investigation however, electrical or metabolic activity of muscles should not be the only factors considered in exercise selection. Further work is required to understand how the hamstrings adapt to these exercises and adaptation is influenced by a range of factors, such as contraction mode (Roig et al., 2009; Timmins, Ruddy, et al., 2015) and range of motion (Lieber & Friden, 2000), which were not a part of the current investigation. For example, there is little reason to believe that concentric or concentrically-biased exercise is effective in HSI prevention or rehabilitation programs (Askling, et al., 2014; Askling, et al., 2013; Mjolsnes, Arnason, Osthagen, Raastad, & Bahr, 2004). Indeed, there is evidence that concentric training may shorten BFLH fascicles (Timmins, Ruddy, et al., 2015) and shift knee flexor torque-joint angle relationships towards shorter muscle lengths (Kilgallon, Donnelly, & Shafat, 2007) and neither of these adaptations are considered beneficial for HSI prevention (Brockett, et al., 2004; Timmins, Bourne, et al., 2015). Because eccentric and concentric training programs appear to have opposing effects on fascicle lengths (Timmins, Ruddy, et al., 2015), it is possible that exercises combining contraction modes may have minimal or at least blunted effects on muscle architecture. 133

149 Future studies are needed to assess the impact of certain exercises on known or proposed risk factors for HSI such as eccentric strength (Opar, et al., 2014) and fascicle lengths (Timmins, Bourne, et al., 2015), and only then will there be sufficient evidence to justify use of those exercises in intervention studies aimed at reducing the risk of injury. Given the high cost of fmri, it was not possible to include all participants in both parts of the experiment. Therefore, comparing the results of part 1 and 2 should be done with caution. Furthermore, all of our participants were recreationally active men so it remains to be seen whether these findings can be applied to more highly trained athletes. We have previously shown that recreationally active young men with a history of unilateral hamstring strain exhibited less T2 change in previously injured muscles than in their uninjured homologous muscles from the contralateral limb after performing the Nordic exercise (Bourne, et al., 2015). Therefore, more research will be needed to establish whether the patterns of selective muscle activation observed in the current study are also evident in athletes with a history of strain injury. Lastly, it should be acknowledged that the T2 response to an exercise stimulus is highly dynamic and can be influenced by a range of factors such as the metabolic capacity and vascular dynamics of the active tissue (Patten, et al., 2003). We attempted to minimise this by recruiting only male participants with a similar age and training status. The current study suggests that the patterns of hamstring muscle activation are heterogeneous between different strength exercises. We have provided semg evidence to suggest that, during eccentric contractions, hip extension exercise more selectively activates the lateral hamstrings while knee flexion-oriented exercises may preferentially recruit the medial hamstrings. However, despite being the least selective activator of the BF, the NHE still elicited higher levels of BF nemg than any other exercise which may help to explain how it 134

150 confers HSI-preventive benefits (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015). The results of the fmri investigation largely confirm our initial semg observations, however, they also show that the BFLH, BFSH, ST and SM all display distinct patterns of muscle use during different tasks. Collectively, the results of this study highlight the limitations of relying on a single method to infer strategies of muscle activation and suggest that the hip extension exercise may be useful for improving strength and voluntary activation of the commonly injured BFLH. Future work is needed to determine the effect of this and other exercises on hamstring architecture and morphology. 135

151

152 Chapter 7: STUDY 4 ADAPTABILITY OF HAMSTRING ARCHITECTURE AND MORPHOLOGY TO TARGETED RESISTANCE TRAINING Publication statement This chapter is comprised of the following paper which is currently under review at the British Journal of Sports Medicine: Bourne, MN., Duhig, SJ., Timmins, RG., Williams, MD., Opar, DA., Al Najjar, A., Kerr, G., & Shield, AJ. (2016). Impact of Nordic hamstring and hip extension training on hamstring muscle architecture and morphology. Br J Sports Med, Under Review. 137

153 Statement of Contribution of Co-Authors for Thesis by Published Paper The authors listed below have certified* that: 1. They meet the criteria for authorship in that they have participated in the conception, execution, or interpretation, of at least that part of the publication in their field of expertise; 2. They take public responsibility for their part of the publication, except for the responsible author who accepts overall responsibility for the publication; 3. There are no other authors of the publication according to these criteria; 4. Potential conflicts of interest have been disclosed to (a) granting bodies, (b) the editor or publisher of journals or other publications, and (c) the head of the responsible academic unit 5. They agree to the use of the publication in the student s thesis and its publication on the Australasian Research Online database consistent with any limitations set by publisher requirements. Contributor Matthew Bourne 08/03/2016 Steven Duhig Ryan Timmins David Opar Aiman Al Najjar Morgan Williams Graham Kerr Anthony Shield Statement of contribution* Experimental design, ethical approval, data collection and analysis, statistical analysis, manuscript preparation Aided in experimental design and data collection Assisted in data collection Aided in experimental design and manuscript preparation Assisted with data collection Assisted with statistical analysis and manuscript preparation Assisted with experimental design and manuscript preparation Aided in experimental design and manuscript preparation Principal Supervisor Confirmation I have sighted from all co-authors confirming their certifying authorship. Dr Anthony Shield 138

154 7.1 LINKING PARAGRAPH The result of Chapter 6 demonstrated that hip extension exercise more selectively activates the BFLH while the NHE preferentially recruits the semitendinosus. These findings have implications for strength training interventions aimed at preventing hamstring injury. However, it remains to be seen how hamstring muscle architecture and morphology adapts to these exercises after a period of training. Indeed, adaptation is influenced by a range of factors, such as contraction mode (Roig, et al., 2009; Timmins, Ruddy, et al., 2015) and range of motion (Lieber & Friden, 2000), which were not explored in Chapter

155 7.2 ABSTRACT The architectural and morphological adaptations of the hamstrings in response to training with different exercises have not been explored. PURPOSE: To evaluate changes in hamstring muscle volume, anatomical cross-sectional area (ACSA) and biceps femoris long head (BFLH) fascicle length following 10-weeks of Nordic hamstring exercise (NHE) or hip extension (HE) training. METHODS: Thirty recreationally active male athletes (age, 22.0 ± 3.6 years, height, ± 7 cm, weight, 80.8 ± 11.1 kg) were randomly allocated to one of three groups: 1) HE training (n=10), NHE training (n=10), or no training (CON) (n=10). BFLH fascicle length was assessed before, during (5 weeks) and after the intervention with two-dimensional ultrasound. Muscle volumes and ACSAs of the hamstring muscles were determined before and after training via magnetic resonance imaging. RESULTS: Compared to baseline, BFLH fascicles were longer in the NHE and HE groups at mid- (p < 0.001) and post-training (p < 0.001) but remained unchanged for the CON group (p > 0.05). The percentage change in BFLH volume was greater for the HE than the NHE (p<0.037) and CON (p < 0.001) groups. Similarly, BFLH ACSA increased more in the HE group than the NHE (p = 0.047) and CON groups (p < 0.001). Both exercises induced similar (p > 0.05) increases in semitendinosus volume and ACSA which were greater than those observed for the CON group (all p 0.002). However, only the NHE group exhibited increased BF short head ACSA, and only the HE group displayed increased semimembranosus volume (p = 0.007) and ACSA (p = 0.015), compared to the CON group. CONCLUSION: NHE and HE training both stimulate significant increases in BFLH fascicle length however, HE training may be more effective for promoting hypertrophy in the BFLH and semimembranosus than the NHE, which appears to selectively develop the semitendinosus and BF short head. Future studies should seek to clarify whether HE exercise is effective in reducing hamstring strain injury in sport. 140

156 7.3 INTRODUCTION Hamstring tears are endemic in sports involving high-speed running and upwards of 80% of these injuries involve the BFLH (Bourne, Opar DA, Williams, & Shield, In Press; Koulouris, et al., 2007; Opar, et al., 2014; Verrall, et al., 2003). Hamstring strains represent the most common injury in athletics (Opar, Drezner, et al., 2013), Australian Rules football (Orchard & Seward, 2002; Orchard, et al., 2011), and soccer (Ekstrand, et al., 2011b) and as many as 30% reoccur within 12 months (Orchard & Best, 2002). These findings highlight the need for improved hamstring injury prevention programs while also suggesting the possibility that these programs should specifically target the BFLH. There has been significant interest in exploring the patterns of muscle activity in hamstring exercises (Bourne, et al, 2015; Bourne et al., In review; Mendiguchia, Arcos, et al., 2013a; Ono, et al., 2011; Ono, et al., 2010; Zebis, et al., 2013), but there is little research examining the architectural and morphological adaptations of these muscles to different exercise interventions. The Nordic hamstring exercise (NHE) has proven effective in increasing eccentric knee flexor strength (Mjolsnes, et al., 2004) and reducing injuries (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015) in soccer, although there is disagreement in the literature as to which hamstring muscles are most active during this exercise (Bourne, et al, 2015; Bourne, In review; Ditroilo, et al., 2013; Mendiguchia, Arcos, et al., 2013a). We have previously reported that the NHE preferentially activates the ST (Bourne, et al, 2015; Bourne, In review), however, we have also observed high levels of BFLH activity in this exercise (Bourne, In review) which suggests that it may nevertheless provide a powerful stimulus for adaptation within this most commonly injured muscle (Bourne, et al., In press; Koulouris, et al., 2007; Opar, et al., 2014; Verrall, et al., 2003). Eccentric exercise has been proposed to increase muscle fascicle lengths via sarcomerogenesis (Brockett, et al., 141

157 2001; Lynn, Talbot, & Morgan, 1998) and Timmins and colleagues (2015) have recently observed such an adaptation after eccentric knee flexor training on an isokinetic dynamometer while also noting that concentric training caused fascicle shortening despite occurring at long muscle lengths. Furthermore, we have recently reported that soccer players with shorter BFLH fascicles (<10.56cm) were at fourfold greater risk of hamstring strain injury than players with longer fascicles (Timmins, Bourne, et al., 2015). Given the effectiveness of the predominantly eccentric NHE in hamstring injury prevention and rehabilitation (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015), it is of interest to examine the impact of this and alternative exercises on BFLH fascicle lengths and morphology. We have recently observed that the 45 hip extension (HE) exercise resulted in more uniform activation of the two-joint hamstrings and greater BFLH activity than the NHE (Bourne, In review). HE exercises are also performed at longer hamstring lengths than the NHE and it has been suggested that this may make them more effective in hamstring injury prevention than the NHE (Guex & Millet, 2013). However, HE and most other hamstring exercises are typically performed with both eccentric and concentric phases and it remains to be seen how the combination of contraction modes will affect fascicle length by comparison with an almost purely eccentric exercise like the NHE. Nevertheless, the greater activation of BFLH during HE (Bourne, et al, 2015; Bourne, In review) may provide a greater stimulus for muscle hypertrophy, which might have implications for rehabilitation practices given observations of persistent atrophy in this muscle following injury (Silder, et al., 2008). The purpose of this study was to evaluate changes in BFLH architecture and hamstring muscle volume and anatomical cross-sectional area (ACSA) following 10-week resistance training 142

158 programs consisting exclusively of HE or the NHE. We tested the hypothesis that the HE exercise would elicit greater increases in BFLH fascicle length than the NHE. Furthermore, given recent observations (Bourne, In review), we hypothesised that HE training would stimulate more BFLH hypertrophy than the NHE, while those training with the NHE would experience more selective hypertrophy of the ST muscle. 7.4 METHODS Participants Thirty recreationally active male athletes (age, 22.0 ± 3.6 years, height, ± 7 cm, weight, 80.8 ± 11.1 kg) provided written informed consent to participate in this study. Participants were free from soft tissue and orthopaedic injuries to the trunk, hips and lower limbs and had no known history of hamstring strain, anterior cruciate ligament or other traumatic knee injury. Before enrolment in the study, all participants completed a cardiovascular screening questionnaire (Appendix B) and a standard MRI questionnaire (Appendix C) to ensure it was safe for them to enter the magnetic field. This study was approved by the Queensland University of Technology Human Research Ethics Committee and the University of Queensland Medical Research Ethics Committee. Study design This randomised controlled study was conducted between April and June, Approximately one week before the intervention commenced, participants underwent MR and 2D ultrasound imaging of their posterior thighs to determine hamstring muscle size and BFLH architecture, respectively. After scanning, all participants were familiarised with the NHE and 45 HE exercise and subsequently underwent strength assessments on each 143

159 exercise. After all of the pre-training assessments had been completed, participants were allocated to one of three groups: NHE, HE or control (CON). Allocation of participants to groups was performed using stratified pseudo-random methods on the basis of baseline BFLH fascicle lengths because this ensured that this parameter did not differ significantly between groups before the intervention. The NHE and HE groups completed a 10-week progressive strength training program consisting exclusively of their allocated exercise (Table 1). The CON group were advised to continue their regular physical activity levels but not to engage in any resistance training for the lower body. At the beginning of every training session, participants in both training groups reported their level of perceived soreness in the posterior thigh using a 1-10 numeric pain rating scale. All CON participants were required to report to the laboratory at least once per week. For all participants, BFLH architecture was re-assessed five weeks into the intervention and within 3-5 days of the final training session. MRI scans were acquired for all participants <7 days after the final training session. Strength testing was conducted after all imaging had been completed. Training intervention Nordic hamstring exercise (NHE) An illustration of the NHE can be found in Figure 1 (see also video supplement). Participants knelt on a padded board, with the ankles secured immediately superior to the lateral malleolus by individual ankle braces which were attached to uniaxial load cells (Figure 1). The ankle braces and load cells were secured to a pivot which allowed the force generated by the knee flexors to be measured through the long axis of the load cells. From the initial kneeling position with their ankles secured in padded yokes, arms on the chest and hips extended, participants lowered their bodies as slowly as possible to a prone position (Bourne, et al, 2015). Participants performed only the lowering (eccentric) portion of the exercise and 144

160 were instructed to use their arms and flex at the hips and knees to push back into the starting position so as to minimise concentric knee flexor activity. When participants developed sufficient strength to completely stop the movement in the final 10-20⁰ of the range of motion, they were instructed to hold a weight plate (range = 2.5 kg to 20 kg) to their chest (centred to the xiphoid process) to ensure the exercise was still of supramaximal intensity. Hip extension exercise (HE) Participants were positioned in a 45 hip extension machine (BodySolid, IL, USA) with their trunk erect and hip joints extended and superior to the level of support pad (Figure 2; see also video supplement). The ankle of the exercised limb was hooked under an ankle pad and the unexercised limb was allowed to rest above its ankle restraint. Participants held one or more circular weight plate(s) to the chest (centred to the xiphoid process) and were instructed to flex their hip until they reached a point approximately 90 from the starting position. Once participants had reached this position they were instructed to return to the starting position by extending their hip, while keeping their trunk in a rigid neutral position throughout. Both limbs were trained in alternating fashion with 30s of rest provided between each set. The load held to the chest was progressively increased throughout the training period whenever the prescribed repetitions and sets could be completed with appropriate technique. 145

161 Figure 7-1. (a) The 45 0 hip extension (HE) exercise and (b) the Nordic hamstring exercise (NHE), progressive from left to right. Hamstring training program Participants in both intervention groups completed a progressive intensity training program consisting of 20 supervised exercise sessions (two per week) over the 10 week period (Table 1). Each session was followed by at least 48 hours of recovery and participants were prohibited from engaging in any other resistance training for the lower body. The training program was based on the approximate loads, repetitions and sets employed in previous interventions using the NHE (Mjolsnes, et al., 2004; Petersen, et al., 2011; van der Horst, et al., 2015), although the volume (number of repetitions) was reduced in the final two weeks to accommodate increases in exercise intensity. All sessions were conducted in the same 146

162 laboratory, employed the same exercise equipment and were supervised by the same investigators (MNB and SJD) to ensure consistency of procedures. Table 7-1. Training program variables Week Frequency Sets Repetitions Strength assessments Before and <7 days after the intervention, all participants underwent an assessment of their maximal eccentric knee-flexor strength during three repetitions of the NHE, and their 3- repetition maximum (3-RM) strength on the 45 hip extension machine. All strength tests were conducted by the same investigators (MNB, SJD and AJS) with tests completed at approximately the same time of day before and after the intervention. Nordic eccentric strength test The assessment of eccentric knee flexor force using the NHE has been reported previously (Bourne, et al., In press; Opar, Piatkowski, et al., 2013a; Opar, et al., 2014; Timmins, Bourne, et al., 2015). Participants completed a single warm-up set of five repetitions followed, one minute later, by a set of three maximal repetitions of the bilateral NHE. A repetition was deemed acceptable when the force output reached a distinct peak (indicative of maximal eccentric strength), followed by a rapid decline when the athlete was no longer able to resist 147

163 the effects of gravity. Eccentric strength was determined for each leg from the peak force produced during the three repetitions of the NHE and was reported in absolute terms (N). Hip extension strength test All strength assessments on the 45 hip extension machine were conducted unilaterally. Participants initially warmed up by performing 8-10 repetitions on each leg using body weight only. Subsequently, the loads were progressively increased until investigators determined the maximal load that could be lifted 3 times. BF LH architecture assessment BFLH fascicle length was determined from ultrasound images taken along the longitudinal axis of the muscle belly utilising a two-dimensional, B-mode ultrasound (frequency, 12Mhz; depth, 8 cm; field of view, 14 x 47 mm) (GE Healthcare Vivid-i, Wauwatosa, U.S.A). Participants were positioned prone on a plinth with their hips in neutral and knees fully extended, while images were acquired from a point midway between the ischial tuberosity and the knee joint fold, parallel to the presumed orientation of BFLH fascicles. After the scanning site was determined, the distance of the site from various anatomical landmarks were recorded to ensure its reproducibility for future testing sessions. These landmarks included the ischial tuberosity, head of the fibula and the posterior knee joint fold at the midpoint between BF and ST tendon. On subsequent visits the scanning site was determined and marked on the skin and then confirmed by replicated landmark distance measures. Images were obtained from both limbs following at least five minutes of inactivity. To gather ultrasound images, the linear array ultrasound probe, with a layer of conductive gel was placed on the skin over the scanning site, aligned longitudinally and perpendicular to the posterior thigh. Care was taken to ensure minimal pressure was placed on the skin by the 148

164 probe as this may influence the accuracy of the measures (Klimstra, Dowling, Durkin, & MacDonald, 2007). The orientation of the probe was manipulated slightly by the sonographer (RGT) if the superficial and intermediate aponeuroses were not parallel. Ultrasound images were analysed using MicroDicom software (Version 0.7.8, Bulgaria). For each image, 6 points were digitised as described by Blazevich and colleagues.(blazevich, Gill, & Zhou, 2006) Following the digitising process, muscle thickness was defined as the distance between the superficial and intermediate aponeuroses of the BFLH. A fascicle of interest was outlined and marked on the image. Fascicle length was determined as the length of the outlined fascicle between aponeuroses and was reported in absolute terms (cm). As the entire fascicles were not visible in the probe s field of view, their lengths were estimated using the following equation: FL=sin (AA+90 ) x MT/sin(180 -(AA+180 -PA)). Where FL=fascicle length, AA=aponeurosis angle, MT=muscle thickness and PA=pennation angle. (Blazevich, et al., 2006; Kellis, Galanis, Natsis, & Kapetanos, 2009) All images were collected and analysed by the same investigator (RGT) who was blinded to participant identity and training group allocation. The assessment of BFLH architecture using the aforementioned procedures by this investigator (RGT) is highly reliable (intraclass correlations >0.90) (Timmins, Shield, et al., 2014). 149

165 Magnetic resonance imaging All MRI scans were performed using a 3-Tesla (Siemens TrioTim, Germany) imaging system with a spinal coil. The participant was positioned supine in the magnet bore with the knees fully extended and hips in neutral, and straps were placed around both limbs to prevent any undesired movement. Contiguous T1-weighted axial MR images (transverse relaxation time: 750ms; echo time: 12ms; field of view: 400mm; slice thickness: 10mm; interslice distance: 0mm) were taken of both limbs beginning at the iliac crest and finishing distal to the tibial condyles. A localiser adjustment (20s) was applied prior to the acquisition of T1-weighted images to standardise the field of view. In addition, to minimise any inhomogeneity in MR images caused by dielectric resonances at 3T, a post-processing (B1) filter was applied to all scans (de Sousa, et al., 2011). The total scan duration was 3min 39sec. Muscle volumes and anatomical cross-sectional areas (ACSAs) of the BFLH and short head (BFSH), semitendinosus (ST) and semimembranosus (SM) muscles were determined for both limbs using manual segmentation. Muscle boundaries were identified and traced on each image in which the desired structure was present using image analysis software (Sante Dicom Viewer and Editor, Cornell University) (Figure 2). Volumes were determined for each muscle by multiplying the summed CSAs (from all the slices containing the muscle of interest) by the interslice distance (Silder, et al., 2008). ACSA was determined by locating the 10mm slice with the greatest CSA and averaging this along with the two slices immediately cranial and caudal (five slices). All traces (pre- and post-training) were completed by the same investigator (MNB) who was blinded to participant identity and training group in all posttesting. 150

166 Figure 7-2. T1-weighted image (transverse relaxation time = 750ms; echo time = 12ms, slice thickness = 10mm), depicting the regions of interest for each hamstring muscle. The right side of the image corresponds to the participant s left side as per radiology convention. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. Statistical analysis All statistical analyses were performed using SPSS version (IBM Corporation, Chicago, IL). Where appropriate, data were screened for normal distribution using the Shapiro-Wilk test and homoscedasticity using Levene s test. Repeated measures split plot ANOVAs were used to determine training-induced changes in BFLH architecture, hamstring muscle volumes and ACSA, and strength, for each group. For the analysis of BFLH architecture, the within-subject variable was time (baseline, mid-training, post-training) and the between-subject variable was group (HE, NHE, CON). Because BFLH architecture did not 151

167 differ between limbs (dominant vs non-dominant) at any time point (p>0.05), the left and right limbs were averaged to provide a single value for each participant. To determine differences in the percentage change in hamstring muscle volume and ACSA between groups, the within-subject variable was muscle (BFLH, BFSH, ST, SM) and the between-subject variable was group (HE, NHE, CON). To explore changes in Nordic and 45 hip extension strength the within-subject variable was (baseline and post-training) and the between-subject variable was group (HE, NHE, CON). For all analyses, when a significant main effect was detected, post hoc independent t tests with Bonferroni corrections were used to determine which comparisons differed. The mean differences were reported with their 95% confidence intervals (CIs). Ratings of perceived soreness (NPRS) throughout the training period were analysed and reported descriptively (mean ± SD) for each group. Sample size A priori sample size estimates were based on anticipated differences in BFLH fascicle length following the training intervention. A sample size of 10 in each group was calculated to provide sufficient statistical power (80%) to detect an effect size of 1.0 with p < Effect size estimates were based on two previous studies (Potier, Alexander, & Seynnes, 2009; Timmins, Ruddy, et al., 2015) which reported a 13-34% increase in BFLH fascicle length following eccentric hamstring training, with an effect size range of

168 7.5 RESULTS No significant differences were observed in age, height or body mass between the three groups (p>0.05) (Table 2). Compliance rates were excellent for both training groups (HE: 100%; NHE: 99.5%). Table 7-2. Participant characteristics Group Age (years) Height (cm) Mass (kg) HE 23.1± ± ±9.7 NHE 21.6± ± ±10.9 CON 21.3± ± ±11.8 Biceps femoris long head (BF LH) architecture A significant group x time interaction effect was observed for fascicle length during the training period (p < 0.001). Post hoc analyses revealed BFLH fascicle length was significantly longer in the NHE group at mid- (mean difference = 1.23cm, 95% CI = 0.84 to 1.63cm, p < 0.001) and post-training (mean difference = 2.22cm, 95% CI = 1.74 to 2.69cm, p < 0.001) compared to baseline (Figure 7-3). The HE group also displayed significantly longer fascicles at mid- (mean difference = 0.75cm, 95% CI = 0.35 to 1.15cm, p < 0.001) and post-training (mean difference = 1.33cm, 95% CI = to 1.80cm, p < 0.001) than baseline. The CON group remained unchanged relative to baseline values at all time points (p > 0.05). The NHE group displayed significantly longer fascicles than the CON group at mid- (mean difference = 1.50cm, 95% CI = 0.58 to 2.41cm, p = 0.001) and post-training (mean difference = 2.40cm, 95% CI = 1.28 to 3.53cm, p < 0.001). The HE group exhibited significantly longer fascicles than the CON group at mid- (mean difference = 1.14cm, 95% CI = 0.22 to 2.05cm, p = 0.011) 153

169 and post-training (mean difference = 1.63cm, 95% CI = 0.51 to 2.76cm, p = 0.003). No significant differences were observed between training groups at either baseline, mid- or post-training points (p>0.05). Figure 7-3. Biceps femoris long head (BFLH) fascicle lengths before (baseline), during (midtraining) and after (post-training) the intervention period for the hip extension (HE), Nordic hamstring exercise (NHE) and control (CON) groups. Fascicle length is expressed in absolute terms (cm) with error bars depicting standard error (SE). * indicates p<0.05 compared to baseline (week 0). ** signifies p<0.001 compared to baseline. # indicates p<0.05 compared to the control group. Hamstring muscle volumes A significant main effect was detected for the muscle x group interaction for hamstring muscle volume changes (p < 0.001). HE training stimulated a greater increase in volume for the ST than the BFSH (mean difference = 5.61%, 95% CI = 1.12% to 10.10%, p = 0.009). No other significant between-muscle differences were noted for volume changes after HE training (p > 0.05 for all pairwise comparisons) or in the CON group (p > 0.05). After NHE 154

170 training, the percentage change in volume was greater for the BFSH than the BFLH (mean difference = 9.56%, 95% CI = 4.30 to 14.80%, p < 0.001) and SM (mean difference = 10.33%, 95% CI = 5.33 to 15.34%, p < 0.001). Similarly, ST volume increased more than BFLH (mean difference = 15.28%, 95% CI = to 19.87%, p < 0.001) and SM (mean difference = 16.06%, 95% CI = to 21.16%, p < 0.001) (Figure 7-4a). BFLH volume increased significantly more in the HE than the NHE (mean difference = 6.72%, 95% CI = 0.32 to 13.11%, p = 0.037) and CON groups (mean difference = 12.10%, 95% CI = 5.71 to 18.50%, p < 0.001), and no significant difference was observed between the NHE and CON groups (mean difference = 5.39%, 95% CI = to %, p = 0.122) (Figure 7-4b). BFSH volume increased more in the HE (mean difference = 8.51%, 95% CI = 0.17 to 16.85%, p = 0.044) and NHE groups (mean difference = 15.29%, 95% CI = 6.95 to 23.63%, p < 0.001) than in the CON group. Both the NHE (mean difference = 21.21%, 95% CI = to 30.88%, p < 0.001) and HE (mean difference = 14.32%, 95% CI = 4.65 to 23.98%, p = 0.002) training groups exhibited a greater increase in ST volume than the CON group. No significant difference in ST volume change was noted between NHE and HE groups (mean difference = 6.90%, 95% CI = to 16.56%, p = 0.239). The percentage change in volume for the SM was significantly greater for the HE group than for CON (mean difference = 8.95%, 95% CI = 2.21 to %, p = 0.007), while no difference was observed between the NHE and CON group changes (mean difference = 3.38%, 95% CI = to 10.12%, p = 0.636) for this muscle. 155

171 Figure 7-4. Percentage change in volume (cm 3 ) for each hamstring muscle after the intervention. Values are expressed as a mean percentage change compared to the values at baseline with error bars representing standard error (SE). (a) Depicts pairwise comparisons for each muscle and (b) shows pairwise comparisons for each group. For all comparisons, * indicates p<0.05 and ** signifies that p< BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. 156

172 Peak hamstring muscle cross-sectional areas (CSA) A significant main effect was detected for the muscle x group interaction (p < 0.001). After HE training, the change in ACSA observed for the ST was significantly greater than the BFLH (mean difference = 6.46, 95% CI = 0.84 to 12.10%, p = 0.017), BFSH (mean difference = 9.98%, 95% CI = 4.25 to 15.71%, p < 0.001) and SM (mean difference = 6.73%, 95% CI = 1.54 to 11.92%, p = 0.006) (Figure 7-5a). No other pairwise between-muscle differences in ACSA change were noted after HE training (all p>0.05). After NHE training, the change in ACSA was greater for BFSH than BFLH (mean difference = 9.30%, 95% CI = 3.47 to 15.12%, p = 0.001) and SM (mean difference = 9.50%, 95% CI = 4.92 to 14.08, p < 0.001), while ST ACSA increased more than BFLH (mean difference = 14.14%, 95% CI = 8.52 to 19.76%, p < 0.001) and SM (mean difference = 14.35%, 95% CI = 9.15 to 19.54%, p < 0.001) (Figure 7-5a). The percentage change in BFLH ACSA was greater in the HE training group than in the NHE (mean difference = 5.24%, 95% CI = to 10.41, p = 0.047) and CON groups (mean difference = 8.90%, 95% CI = 3.73 to 14.07%, p < 0.001), while no difference was observed between the NHE and CON groups (mean difference = 3.67%, 95% CI = to 8.84%, p = 0.245) (Figure 7-5b). BFSH ACSA increased significantly more in the NHE than the CON group (mean difference = 13.26%, 95% CI = 4.98 to 21.54%, p = 0.001), while no difference was observed between changes exhibited by the HE and CON groups (mean difference = 5.69%, 95% CI = to 0.70%, p = 0.273). The percentage change in ST ACSA was significantly greater in the NHE (mean difference = 17.60%, 95% CI = 7.60 to 27.61%, p < 0.001) and HE (mean difference = 15.16%, 95% CI = 5.15 to 25.17%, p = 0.002) groups than the CON group, however no significant difference was noted between changes in the NHE and HE groups (mean difference = 2.4%, 95% CI = to 12.45%, p = 1.000). The percentage increase in SM ACSA was greater in the HE than the CON group (mean 157

173 difference = 7.19%, 95% CI = 1.21 to 13.18%, p = 0.015), but was not significantly greater in NHE than CON (mean difference = 2.02%, 95% CI = to 8.01%, p = 1.000). No significant difference in SM ACSA change was noted between the HE and NHE groups (main difference = 5.17%, 95% CI = -8.2 to 11.16%, p = 0.109) (Figure 7-5b). 158

174 Figure 7-5. Percentage change in anatomical cross sectional area (ACSA) (cm 2 ) for each hamstring muscle after the intervention. Values are expressed as a mean percentage change compared to the values at baseline with error bars representing standard error (SE). (a) Depicts pairwise comparisons for each muscle and (b) shows pairwise comparisons for each group. For all comparisons, * indicates p<0.05 and ** signifies that p< BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. 159

175 Strength A significant group x time interaction effect was observed for the Nordic eccentric strength test (p < 0.001) (Figure 7-6). Post hoc t tests demonstrated that the NHE (mean difference = 97.38N, 95% CI = to N, p < 0.001) and HE (mean difference = N, 95% CI = to N, p < 0.001) groups were significantly stronger at post-training compared to baseline while the CON group did not change (mean difference = 8.91N, 95% CI = to 24.69N, p = 0.590). No groups differed at baseline (p > 0.461) however at posttraining, the NHE (mean difference = N, 95% CI = to N, p = 0.003) and HE (mean difference = 94.27N, 95% CI = 8.60 to N, p = 0.028) groups were significantly stronger than the CON group. No significant difference was observed between training groups at post-training (mean difference = 29.16N, 95% CI = to N, p = 1.000). 160

176 Figure 7-6. Eccentric knee flexor force measured during the Nordic strength test before (baseline) and after (post-training) the intervention period for the hip extension (HE), Nordic hamstring exercise (NHE) and control (CON) groups. Force is reported in absolute terms (N) with error bars depicting standard error (SE). * indicates p<0.001 compared to baseline (week 0). # signifies p<0.05 compared to the control group. A significant group x time interaction effect was also observed for 3-RM strength as assessed during the 45⁰ HE strength test (p < 0.001) (Figure 7-7). Post hoc analyses demonstrated that the HE (mean difference = 41.00kg, 95% CI = to 46.03kg, p < 0.001) and NHE groups (mean difference = 26.00kg, 95% CI = to 31.03kg, p < 0.001) improved significantly from baseline whereas the CON group did not change (mean difference = 3.50kg, 95% CI = to 8.53kg, p = 0.165). No groups differed significantly at baseline (p > 0.091) however at post-training, both the HE (mean difference = 43.50kg, 95% CI = to 56.07kg, p < 0.001) and NHE groups (mean difference = 32.0kg, 95% CI = to 44.57kg, p < 0.001) were significantly stronger than CON. No difference was observed between training groups (mean difference = 11.50kg, 95% CI = to 24.07kg, p = 0.082). 161

177 Figure 7-7. Hip extension three-repetition maximum (3RM) before (baseline) and after (post training) the intervention period for the hip extension (HE), Nordic hamstring exercise (NHE) and control (CON) groups. Force is reported in absolute terms (kg) with error bars depicting standard error (SE). ** indicates p<0.001 compared to baseline (week 0). # signifies p<0.001 compared to the control group. 162

178 7.6 DISCUSSION This study is the first to explore the architectural and morphological adaptations of the hamstrings in response to different strength training exercises. These data suggest that both the HE and NHE stimulate significant increases in BFLH fascicle length. However, HE training appears to elicit more hypertrophy in the BFLH than does the NHE, which preferentially develops the ST and BFSH muscles. Both exercises resulted in significant strength increases which were equally evident in the NHE and HE strength tests. Fascicle lengthening is one possible mechanism by which the NHE (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015) and other eccentric or long length hamstring exercises (Askling, et al., 2014) protect muscles from injury. We have recently shown, prospectively, that professional soccer players with fascicles <10.56cm were ~4 times more likely to suffer a hamstring strain than athletes with longer fascicles and that the probability of injury was reduced by ~74% for every 0.5cm increase in fascicle length (Timmins, Bourne, et al., 2015). In the current study, participants increased their fascicle lengths from ~10.6cm prior to training, to 12.8 and 12.0cm in the NHE and HE groups, respectively, which would likely result in significant reductions in hamstring injury risk. Despite its success in reducing hamstring strain injuries, the adoption of the NHE in elite European soccer has been reported to be poor with only ~11% of Norwegian premier league and UEFA teams deemed to have adequately implemented the NHE programs that have proven effective in randomised controlled trials (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015). Some conditioning coaches and researchers (Guex & Millet, 2013) believe that the exercise does not challenge the hamstrings at sufficient lengths to optimise injury prevention benefits. However, this study shows, for the first time, that the 163

179 limited excursion of the hamstrings during the NHE does not prevent the exercise from increasing BFLH fascicle length. Indeed, the exercise resulted in greater fascicle lengthening than the HE, although the current study lacked the statistical power to distinguish between the two. Together with observations that long length concentric hamstring training can shorten muscle fascicles (Timmins, Ruddy, et al., 2015), the current findings are consistent with the possibility that the combination of concentric and eccentric contractions somewhat dampens the elongation of BFLH fascicles. The advantage of the NHE may be its almost purely eccentric or eccentrically-biased nature. Further work is needed to clarify whether eccentrically-biased or purely eccentric HE exercise may yield greater improvements in BFLH fascicle length than the combined concentric and eccentric contraction modes used in this investigation. Observations of increased fascicle length following eccentric hamstring exercise are largely consistent with previous literature. For example, Potier and colleagues (Potier, et al., 2009) reported a 34% increase in BFLH fascicle length following eight weeks of eccentric leg curl exercise, while Timmins and colleagues (Timmins, Ruddy, et al., 2015) reported a 16% increase in BFLH fascicle length after six weeks of eccentrically-biased training on an isokinetic dynamometer (Timmins, Ruddy, et al., 2015), These adaptations most likely result from the addition of in-series sarcomeres, as has been shown to occur within the rat vastus intermedius muscle after five days of downhill running (Lynn & Morgan, 1994). It has been proposed that this increase in serial sarcomeres accounts for both a rightward shift in a muscle s force-length relationship (Reeves, et al., 2004), while also reducing its susceptibility to damage (Brockett, et al., 2001; Lynn, et al., 1998). However, theoretically, it is also possible that fascicles lengthen as a product of increased tendon or aponeurotic 164

180 stiffness. Further research is needed to clarify the precise mechanism(s) responsible for these architectural changes. To the authors knowledge, this is the first study to explore the morphological adaptations of the hamstrings to different strengthening exercises. These data suggest that the NHE and HE exercises induce different patterns of hamstring muscle hypertrophy, with the former preferentially stimulating ST and BFSH growth and the latter resulting in significantly more hypertrophy of the BFLH and more homogenous growth of all two-joint hamstring muscles. We have previously noted transient T2 relaxation time changes after 50 repetitions of each of these exercises that almost exactly fitted this pattern (Bourne, et al., 2016), so it appears that the acute changes observed via functional MRI match quite well the hypertrophic effects observed after 10 weeks of training. However, neither muscle volume nor ACSA have been identified as risk factors for hamstring strain injury, so the exact significance of these findings is unknown. Indeed, we have previously reported that BFLH muscle thickness measured via ultrasound is not a risk factor for hamstring injury in elite soccer (Timmins, Bourne, et al., 2015). Nevertheless, BFLH muscle atrophy has been noted as long as 5-23 months after injury in recreational athletes (Silder, et al., 2008), so unilateral HE exercises may prove more beneficial than the NHE at redressing this deficit in rehabilitation. Interestingly, reduced muscle volumes of the ST have been observed months after anterior cruciate ligament injury (Nomura, Kuramochi, & Fukubayashi, 2015) and the results of the current investigation suggest that the NHE may be valuable in rehabilitation of this injury. Future intervention studies analogous to those employing the NHE previously (Arnason, et al., 2008; Gabbe, Branson, & Bennell, 2006; Petersen, et al., 2011; van der Horst, et al., 2015), should seek to clarify whether HE training is effective in reducing hamstring strain injuries. 165

181 Hamstring strengthening is an important component of injury prevention strategies (Guex & Millet, 2013; Malliaropoulos, et al., 2012; Opar, et al., 2012). Indeed, several large scale interventions employing the NHE have shown ~65% reductions in hamstring strain injury rates in soccer (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015) and recent prospective findings in elite Australian football (Opar, et al., 2014) and soccer (Timmins, Bourne, et al., 2015) suggest that eccentric strength improvements like those reported here and previously (Mjolsnes, et al., 2004) are at least partly responsible for these protective benefits. For example, elite athletes in these sports who generated less than < 279 N (Australian football) and < 337 N (soccer) of knee flexor force at the ankles during the NHE strength test were ~4 times more likely to sustain hamstring injuries than stronger counterparts (Opar, et al., 2014; Timmins, Bourne, et al., 2015). In this study, our recreational level athletes were able to generate, on average, 460N and 431N after NHE and HE training, respectively, after 10 weeks of training, making them substantially stronger than these elite Australian football (Opar, et al., 2014) and soccer players (Timmins, Bourne, et al., 2015). Significant improvements in 3-RM HE strength were also observed for both training groups, which suggests that hamstring strengthening, at least in recreationally trained athletes, is not highly specific to the chosen exercise. While the benefits of high levels of HE strength remain unclear from the perspective of injury prevention, the observed effects of HE training on BFLH fascicle lengths and eccentric knee flexor strength suggest the potential for this exercise to reduce hamstring injury risk and this possibility should be explored in future investigations. The authors acknowledge that there are some limitations associated with the current study. Firstly, muscle architecture was only assessed in the BFLH and it may not be appropriate to generalise these findings to other knee flexors, given that each hamstring muscle displays 166

182 unique architectural characteristics (Woodley & Mercer, 2005). Further, the assessment of fascicle length using two-dimensional ultrasound requires some degree of estimation, because the entire length of the BFLH is not always visible. While the estimation equation used in this study has been validated against cadaveric samples (Kellis, et al., 2009), there is still the potential for error, and future studies employing extended field of view ultrasound methods may be needed to completely eliminate this. Lastly, all of the athletes in this study were recreational level males of a similar age, and it remains to be seen if these results are applicable to other populations. However, our participants were, on average, as strong as elite Australian football players (Opar, et al., 2014) and stronger than professional soccer players (Timmins, Bourne, et al., 2015) at the start of the study. Furthermore, our cohort displayed average fascicle lengths before training that were within one standard deviation of the values reported in elite soccer players previously (Timmins, Bourne, et al., 2015), so it is unlikely that they were unrepresentative of higher-level athletes, in these parameters at least. This is the first study to demonstrate that training with different exercises elicits unique architectural and morphological adaptations within the hamstring muscle group. We have provided evidence to suggest that both NHE and HE training are effective in lengthening BFLH fascicles. However, HE training appears to be more effective for promoting hypertrophy in the commonly injured BFLH than the NHE, which preferentially develops the ST and BFSH muscles. These data may help to explain the mechanism(s) by which the NHE confers injury preventive benefits and also provide compelling evidence to warrant the further exploration of HE-oriented exercises in hamstring strain injury prevention protocols. Future prospective studies are needed to ascertain whether HE training interventions are effective in reducing the incidence of hamstring strain injury in sport and whether or not the combination of HE and NHE is more effective than the NHE alone. 167

183

184 Chapter 8: GENERAL DISCUSSION, LIMITATIONS & CONCLUSION This program of research aimed to 1) further examine the role of eccentric knee flexor strength and between-limb imbalances in eccentric strength, in hamstring strain injury (HSI) occurrence; 2) explore the neuromuscular maladaptations that may manifest following HSI and underpin high rates of injury recurrence; and 3) characterise the activation patterns and the architectural and morphological adaptations of the hamstrings to difference strength training exercises. Study 1 demonstrated that between-limb imbalances in eccentric knee flexor strength and prior HSI both increased the risk of future strain injury in rugby union players. Moreover, for those athletes who had been injured previously, their risk of re-injury was augmented if they had returned to competition with strength imbalances. Study 2 provided novel evidence to suggest that elite athletes with a history of strain injury to the BFLH have a reduced capacity to activate the previously injured muscle during high-speed running, for many months following a return to sport. In light of these findings, it was important to understand how training practices can be improved to redress strength imbalances and to better target the commonly injured BFLH in prophylactic programs. Study 3 demonstrated that the hamstrings are activated non-uniformly during various strength training exercises and that hip extension tasks more selectively activate the BFLH than the Nordic hamstring exercise (NHE), which preferentially recruits the semitendinosus. Study 4 built upon these findings by exploring the architectural and morphological adaptations of the hamstrings to 10 weeks of hip extension or NHE training. This study provided evidence to suggest that both NHE and hip extension training stimulate significant increases in BFLH fascicle length however, hip extension exercise may be more effective for promoting hypertrophy in BFLH than the NHE, which selectively develops the semitendinosus muscle. 169

185 The prospective findings from Study 1 highlight the importance of ameliorating betweenlimb imbalances in eccentric strength, particularly following HSI, to reduce the risk of future injury. It might be expected that these imbalances would be addressed in the rehabilitation process however, the retrospective findings from Study 2 ((and previous observations (Bourne, et al, 2015; Opar, Williams, et al., 2013a)), suggest the possibility that conventional HSI rehabilitation strategies may not be effectively restoring strength and voluntary activation to the commonly injured BFLH. Indeed, it has been proposed (Fyfe, et al., 2013b) that these activation deficits may mediate the preferentially eccentric weakness (Opar, Williams, et al., 2013a), reduced rates of knee flexor torque development (Opar, Williams, et al., 2013b), BFLH atrophy (Silder, et al., 2008) and reduced fascicle lengths (Timmins, Shield, et al., 2014) that have been reported in the literature. Further work is required to determine whether neuromuscular inhibition is the cause or result of prior HSI, and if indeed it represents a risk factor for injury recurrence. An improved understanding of the patterns of hamstring muscle activation during common strength training exercises may enable practitioners to make better informed decisions regarding exercise selection in injury prevention and rehabilitation programs. Using semg and fmri, study 3 demonstrated that the commonly employed NHE only moderately activates the BFLH relative to the semitendinosus muscle. In contrast, hip extension exercise appeared to most selectively activate the BFLH which suggests that it might provide a more effective stimulus for stimulating adaptations in this muscle. Study 4 corroborated this hypothesis by demonstrating that 10 weeks of hip extension training elicited significantly more hypertrophy to the BFLH than a volume-matched program using the NHE. However, both exercises promoted significant lengthening of BFLH fascicles, which may help to explain how the NHE (Arnason, et al., 2008; Petersen, et al., 2011; van der Horst, et al., 2015) and 170

186 other long-length hamstring exercises (Askling, et al., 2014; Askling, et al., 2013) protect against hamstring injury. These data also provide compelling evidence to warrant the further exploration of hip extension exercise in hamstring strain injury prevention protocols. There are some limitations associated with program of research. With respect to Study 1, the assessment of eccentric knee-flexor strength and between-limb imbalance was only performed at a single time point in the pre-season period and it is important to consider that strength may change throughout the pre-season and in-season periods. Further, given that we only employed rugby union players, the results may not be generalizable to other sports or populations. Study 2 was retrospective in nature and it remains to be seen if the observed activation deficits were the cause or result of the athletes prior HSI. Moreover, given the absence of an uninjured control group, it is difficult to determine whether participants had normal patterns of muscle activation in their uninjured limbs. The techniques used to assess voluntary hamstring activation in Study 2 and 3 also have limitations. For example, semg is prone to cross talk (Farina, et al., 2004) and cannot discriminate between closely approximated segments of muscles (Adams, et al., 1992). Functional MRI largely overcomes this spatial limitation however the T2 response to an exercise stimulus is highly dynamic and can be influenced by a range of factors such as the metabolic capacity and vascular dynamics of the active tissue. With respect to Study 4, the assessment of muscle architecture was only performed on the BFLH. This was justified on the basis that the BFLH was the most frequently injured muscle in Study 1, however, the adaptability of muscle architecture in other hamstring muscles is worthy of future investigation. Further, the assessment of fascicle length using two-dimensional ultrasound requires some degree of estimation, because the entire length of the BFLH is not always visible. While the estimation equation used in this study has been validated against cadaveric samples, (Kellis, et al., 2009) there is still the 171

187 potential for error, and future studies employing extended field of view ultrasound methods may be needed to completely eliminate this. In conclusion, this program of research has provided prospective data on risk factors for HSI and has retrospectively explored maladaptations which may manifest as a result of injury. In addition, it has provided novel data on the activation patterns and the architectural and morphological adaptations of the hamstrings to different strength training exercises. The findings from studies 1 & 2 highlight the importance of ameliorating strength and voluntary activation deficits, particularly following HSI, while data from studies 3 & 4 provide an evidence base from which to form decisions regarding exercise selection in prophylactic programs. 172

188 Bibliography Aagaard, P., Simonsen, E. B., Andersen, J. L., Magnusson, S. P., Halkjaer-Kristensen, J., & Dyhre-Poulsen, P. (2000). Neural inhibition during maximal eccentric and concentric quadriceps contraction: effects of resistance training. J Appl Physiol, 89(6), Aagaard, P., Simonsen, E. B., Magnusson, S. P., Larsson, B., & Dyhre-Poulsen, P. (1998). A new concept for isokinetic hamstring: quadriceps muscle strength ratio. Am J Sports Med, 26(2), Adams, G. R., Duvoisin, M. R., & Dudley, G. A. (1992). Magnetic resonance imaging and electromyography as indexes of muscle function. J Appl Physiol (1985), 73(4), Adams, G. R., Harris, R. T., Woodard, D., & Dudley, G. A. (1993). Mapping of electrical muscle stimulation using MRI. J Appl Physiol, 74(2), Agre, J. C. (1985). Hamstring injuries. Proposed aetiological factors, prevention, and treatment. Sports Med, 2(1), Akima, H., Takahashi, H., Kuno, S. Y., Masuda, K., Masuda, T., Shimojo, H.,... Katsuta, S. (1999). Early phase adaptations of muscle use and strength to isokinetic training. Med Sci Sports Exerc, 31(4), Allen, T. J., Leung, M., & Proske, U. (2010). The effect of fatigue from exercise on human limb position sense. J Physiol, 588(8), Amiridis, I., Martin, A., Morlon, B., Martin, L., Cometti, G., Pousson, M., & Van Hoecke, J. (1996). Co-activation and tension-regulating phenomena during isokinetic knee extension in sedentary and highly skilled humans. European J Appl Physiol and occupational physiology, 73(1), Arendt-Nielsen, L., & Zwarts, M. (1989). Measurement of muscle fiber conduction velocity in humans: techniques and applications. J Clin Neurophysiol, 6(2), 173. Arnason, A., Andersen, T. E., Holme, I., Engebretsen, L., & Bahr, R. (2008). Prevention of hamstring strains in elite soccer: an intervention study. Scand J Med Sci Sports, 18(1), Bibliography 173

189 Arnason, A., Sigurdsson, S. B., Gudmundsson, A., Holme, I., Engebretsen, L., & Bahr, R. (2004). Risk factors for injuries in football. Am J Sports Med, 32(1 Suppl), 5S-16S. Askling, C., Karlsson, J., & Thorstensson, A. (2003). Hamstring injury occurrence in elite soccer players after preseason strength training with eccentric overload. Scand J Med Sci Sports, 13(4), Askling, C., Saartok, T., & Thorstensson, A. (2006). Type of acute hamstring strain affects flexibility, strength, and time to return to pre-injury level. Br J Sports Med, 40(1), Askling, C. M., Tengvar, M., Saartok, T., & Thorstensson, A. (2007). Acute First-Time Hamstring Strains During High-Speed Running A Longitudinal Study Including Clinical and Magnetic Resonance Imaging Findings. Am J Sports Med, 35(2), Askling, C. M., Tengvar, M., Tarassova, O., & Thorstensson, A. (2014). Acute hamstring injuries in Swedish elite sprinters and jumpers: a prospective randomised controlled clinical trial comparing two rehabilitation protocols. Br J Sports Med, 48(7), Askling, C. M., Tengvar, M., & Thorstensson, A. (2013). Acute hamstring injuries in Swedish elite football: a prospective randomised controlled clinical trial comparing two rehabilitation protocols. Br J Sports Med, 47(15), Babault, N., Pousson, M., Michaut, A., Ballay, Y., & Hoecke, J. (2002). EMG activity and voluntary activation during knee-extensor concentric torque generation. European J Appl Physiol, 86(6), Bahr, R., & Holme, I. (2003). Risk factors for sports injuries--a methodological approach. Br J Sports Med, 37(5), Behm, D. G., & St-Pierre, D. M. M. (1997). Fatigue characteristics following ankle fractures. Med Sci Sports Exerc, 29(9), Belanger, A., & McComas, A. (1981). Extent of motor unit activation during effort. J Appl Physiol: respiratory, environmental and exercise physiology, 51(5), Beltman, J. G. M., Sargeant, A. J., Mechelen, W. v., & Haan, A. d. (2004). Voluntary activation level and muscle fibre recruitment of human quadriceps during lengthening contractions. J Appl Physiol, 97(2), Bibliography 174

190 Bennell, K., & Crossley, K. (1996). Musculoskeletal injuries in track and field: incidence, distribution and risk factors. Aust J Sci Med Sport, 28(3), 69. Bennell, K., Tully, E., & Harvey, N. (1999). Does the toe-touch test predict hamstring injury in Australian Rules footballers? Aust J Physiother, 45(2), Bennell, K., Wajswelner, H., Lew, P., Schall-Riaucour, A., Leslie, S., Plant, D., & Cirone, J. (1998). Isokinetic strength testing does not predict hamstring injury in Australian Rules footballers. Br J Sports Med, 32(4), Bilsborough, J. C., Greenway, K., Opar, D., Livingstone, S., Cordy, J. T., Bird, S., & Coutts, A. J. (2014). Comparison of anthropometry, upper body strength and lower body power characteristics in different levels of Australian Football players. J Strength Cond Res, 29(3): Blazevich, A. J., Gill, N. D., & Zhou, S. (2006). Intra- and intermuscular variation in human quadriceps femoris architecture assessed in vivo. J Anat, 209(3), Bourne, M., Opar DA, Williams, M., & Shield, A. (In press). Eccentric Knee-flexor Strength and Hamstring Injury Risk in Rugby Union: A Prospective Study. Am J Sports Med, 43(11): Bourne, M., Opar,DA, Williams,MD, Al Najjar, A, Shield, AJ (2015). Muscle activation patterns in the Nordic hamstring exercise: Impact of prior strain injury. Scand J Med Sci Sports. [Epub ahead of print] Bourne MN, W. M., Opar DA, Al Najjar A, Shield AJ. (In review). Impact of exercise selection on hamstring muscle activation. Br J Sports Med. Bradley, P. S., & Portas, M. D. (2007). The relationship between preseason range of motion and muscle strain injury in elite soccer players. J Strength Cond Res, 21(4), Brockett, C., Morgan, D., & Proske, U. (2001). Human hamstring muscles adapt to eccentric exercise by changing optimum length. Med Sci Sports Exerc, 33(5), Brockett, C. L., Morgan, D. L., & Proske, U. (2004). Predicting hamstring strain injury in elite athletes. Med Sci Sports Exerc, 36(3), Brooks, J. H., Fuller, C. W., Kemp, S. P., & Reddin, D. B. (2005a). Epidemiology of injuries in English professional rugby union: part 1 match injuries. Br J Sports Med, 39(10), Bibliography 175

191 Brooks, J. H., Fuller, C. W., Kemp, S. P., & Reddin, D. B. (2005b). Epidemiology of injuries in English professional rugby union: part 2 training Injuries. Br J Sports Med, 39(10), Brooks, J. H. M., Fuller, C. W., Kemp, S. P. T., & Reddin, D. B. (2005c). Epidemiology of injuries in English professional rugby union: part 1 match injuries. Br J Sports Med, 39, Brooks, J. H. M., Fuller, C. W., Kemp, S. P. T., & Reddin, D. B. (2006). Incidence, risk, and prevention of hamstring muscle Injuries in professional rugby union. Am J Sports Med, 34(8), Brown, S. J., Child, R. B., Donnelly, A. E., Saxton, J., & Day, S. H. (1996). Changes in human skeletal muscle contractile function following stimulated eccentric exercise. European J Appl Physiol, 72(5), Brughelli, M., Cronin, J., Mendiguchia, J., Kinsella, D., & Nosaka, K. (2010). Contralateral leg deficits in kinetic and kinematic variables during running in Australian rules football players with previous hamstring injuries. J Strength Cond Res, 24(9), Burkett, L. N. (1970). Causative factors in hamstring strains. Med Sci Sports Exerc, 2(1), Cagnie, B., Dickx, N., Peeters, I., Tuytens, J., Achten, E., Cambier, D., & Danneels, L. (2008). The use of functional MRI to evaluate cervical flexor activity during different cervical flexion exercises. J Appl Physiol (1985), 104(1), Cagnie, B., Elliott, J. M., O'Leary, S., D'Hooge, R., Dickx, N., & Danneels, L. A. (2011). Muscle functional MRI as an imaging tool to evaluate muscle activity. J Orthop Sports Phys Ther, 41(11), Carolan, B., & Cafarelli, E. (1992). Adaptations in coactivation after isometric resistance training. J Appl Physiol, 73(3), Chumanov, E. S., Heiderscheit, B. C., & Thelen, D. G. (2007). The effect of speed and influence of individual muscles on hamstring mechanics during the swing phase of sprinting. J Biomech, 40(16), Chumanov, E. S., Heiderscheit, B. C., & Thelen, D. G. (2011). Hamstring musculotendon dynamics during stance and swing phases of high-speed running. Med Sci Sports Exerc, 43(3), Bibliography 176

192 Ciccotti, M. G., Kerlan, R. K., Perry, J., & Pink, M. (1994). An Electromyographic Analysis of the Knee During Functional Activities I. The Normal Profile. The Am J Sports Med, 22(5), Connell, D. A., Schneider-Kolsky, M. E., Hoving, J. L., Malara, F., Buchbinder, R., Koulouris, G.,... Bass, C. (2004). Longitudinal study comparing sonographic and MRI assessments of acute and healing hamstring injuries. AJR Am J Roentgenol, 183(4), Craig, T. T. (1973). American Medical Association Comments in Sports Medicine. American Medical Association, Croisier, J. L. (2004). Factors associated with recurrent hamstring injuries. Sports Med, 34(10), Croisier, J. L., & Crielaard, J. M. (2000). Hamstring muscle tear with recurrent complaints: an isokinetic profile. Isokinet Exerc Sci, 8(3), Croisier, J. L., & Crielaard, J. M. (2001). [Isokinetic exercise and sports injuries]. Rev Med Liege, 56(5), Croisier, J. L., Forthomme, B., Namurois, M. H., Vanderthommen, M., & Crielaard, J. M. (2002). Hamstring muscle strain recurrence and strength performance disorders. Am J Sports Med, 30(2), Croisier, J. L., Ganteaume, S., Binet, J., Genty, M., & Ferret, J. M. (2008a). Strength imbalances and prevention of hamstring injury in professional soccer players. Am J Sports Med, 36(8), Croisier, J. L., Ganteaume, S., Binet, J., Genty, M., & Ferret, J. M. (2008). Strength imbalances and prevention of hamstring injury in professional soccer players: a prospective study. Am J Sports Med, 36(8), D'Souza, D. (1994). Track and field athletics injuries--a one-year survey. Br J Sports Med, 28(3), Daly, C., McCarthy Persson, U., Twycross-Lewis, R., Woledge, R. C., & Morrissey, D. (2015). The biomechanics of running in athletes with previous hamstring injury: A case-control study. Scand J Med Sci Sports. Bibliography 177

193 Dauty, M., Potiron-Josse, M., & Rochcongar, P. (2003). Consequences and prediction of hamstring muscle injury with concentric and eccentric isokinetic parameters in elite soccer players. Ann Readapt Med Phys, 46(9), De Smet, A. A., & Best, T. M. (2000). MR imaging of the distribution and location of acute hamstring injuries in athletes. AJR Am J Roentgenol, 174(2), de Sousa, P. L., Vignaud, A., Fleury, S., & Carlier, P. G. (2011). Fast monitoring of T(1), T(2), and relative proton density (M(0) ) changes in skeletal muscles using an IR- TrueFISP sequence. J Magn Reson Imaging, 33(4), Deschenes, M. R., & Kraemer, W. J. (2002). Performance and physiologic adaptations to resistance training. Am J Phys Med, 81(11), S3. Devlin, L. (2000). Recurrent posterior thigh symptoms detrimental to performance in rugby union: predisposing factors. Sports Med, 29(4), Diederichsen, L. P., Nørregaard, J., Dyhre-Poulsen, P., Winther, A., Tufekovic, G., Bandholm, T.,... Krogsgaard, M. (2009). The activity pattern of shoulder muscles in subjects with and without subacromial impingement. J Electromyogr Kinesiol, 19(5), Ditroilo, M., De Vito, G., & Delahunt, E. (2013). Kinematic and electromyographic analysis of the Nordic Hamstring Exercise. J Electromyogr Kinesiol, 23(5), Drezner, J., Ulager, J., & Sennet, M. (2005). Hamstring muscle injuries in track and field athletes: A 3-year study at the Penn Relay Carnival [abstract]. Clin J Sport Med, 15(5), 386. Dudley, G. A., Tesch, P., Miller, B., & Buchanan, P. (1991). Importance of eccentric actions in performance adaptations to resistance training. Aviat Space Environ Med, 62(6), 543. Duthie, G., Pyne, D., & Hooper, S. (2003). Applied physiology and game analysis of rugby union. Sports Med, 33(13), Dyhre-Poulsen, P., & Krogsgaard, M. R. (2000). Muscular reflexes elicited by electrical stimulation of the anterior cruciate ligament in humans. J Appl Physiol, 89(6), Bibliography 178

194 Edman, K., Elzinga, G., & Noble, M. (1978). Enhancement of mechanical performance by stretch during tetanic contractions of vertebrate skeletal muscle fibres. Physiol, 281(1), 139. Ekstrand, J., & Gillquist, J. (1983). Soccer injuries and their mechanisms: a prospective study. Med Sci Sports Exerc, 15(3), Ekstrand, J., Hagglund, M., & Walden, M. (2010). Injury incidence and injury patterns in professional football: the UEFA injury study. Br J Sports Med, 45(7): Ekstrand, J., Hagglund, M., & Walden, M. (2011a). Epidemiology of Muscle Injuries in Professional Football (Soccer). Am J Sports Med, 39(6): Ekstrand, J., Hagglund, M., & Walden, M. (2011b). Injury incidence and injury patterns in professional football: the UEFA injury study. Br J Sports Med, 45(7), Elliott, M. C., Zarins, B., Powell, J. W., & Kenyon, C. D. (2011). Hamstring Muscle Strains in Professional Football Players: A 10-Year Review. Am J Sports Med, 39(4): Farina, D., Cescon, C., & Merletti, R. (2002). Influence of anatomical, physical, and detection-system parameters on surface EMG. Biol Cybern, 86(6), Farina, D., Merletti, R., & Enoka, R. M. (2004). The extraction of neural strategies from the surface EMG. J Appl Physiol, 96(4), Feeley, B. T., Kennelly, S., Barnes, R. P., Muller, M. S., Kelly, B. T., Rodeo, S. A., & Warren, R. F. (2008). Epidemiology of National Football League training camp injuries from 1998 to Am J Sports Med, 36(8), Fisher, M.J., Meyer, R. A., Adams, G. R., Foley, J. M., & Potchen, E. J. (1990). Direct relationship between proton T2 and exercise intensity in skeletal muscle MR images. Invest Radiol, 25(5), Fleckenstein, J. L., Canby, R. C., Parkey, R. W., & Peshock, R. M. (1988). Acute effects of exercise on MR imaging of skeletal muscle in normal volunteers. AJR Am J Roentgenol, 151(2), Fleckenstein, J. L., Haller, R. G., Lewis, S. F., Archer, B. T., Barker, B. R., Payne, J.,... Peshock, R. M. (1991). Absence of exercise-induced MRI enhancement of skeletal muscle in McArdle's disease. J Appl Physiol (1985), 71(3), Bibliography 179

195 Foley, J. M., Jayaraman, R. C., Prior, B. M., Pivarnik, J. M., & Meyer, R. A. (1999). MR measurements of muscle damage and adaptation after eccentric exercise. J Appl Physiol, 87(6), Fousekis, K., Tsepis, E., Poulmedis, P., Athanasopoulos, S., & Vagenas, G. (2011). Intrinsic risk factors of non-contact quadriceps and hamstring strains in soccer: a prospective study of 100 professional players. Br J Sports Med, 45(9), Freckleton, G., Cook, J., & Pizzari, T. (2014). The predictive validity of a single leg bridge test for hamstring injuries in Australian Rules Football Players. Br J Sports Med, 48(8), Freckleton, G., & Pizzari, T. (2013). Risk factors for hamstring muscle strain injury in sport: a systematic review and meta-analysis. Br J Sports Med, 47(6), Friden, J., & Lieber, R. L. (1992). Structural and mechanical basis of exercise-induced muscle injury. Med Sci Sports Exerc, 24(5), Fuller, C. W., Sheerin, K., & Targett, S. (2013). Rugby World Cup 2011: International Rugby Board injury surveillance study. Br J Sports Med, 47(18), Fyfe, J. J., Opar, D. A., Williams, M. D., & Shield, A. J. (2013a). The role of neuromuscular inhibition in hamstring strain injury recurrence. J Electromyogr Kinesiol, 23(3), Gabbe, B., Bennell, K., Finch, C., Wajswelner, H., & Orchard, J. (2006). Predictors of hamstring injury at elite level of Australian football. Scand J Med Sci Sports, 16(1), Gabbe, B., Finch, C., Wajswelner, H., & Bennell, K. (2002). Australian Football: Injury profile at the community level. J Sci Med Sport, 5(2), Gabbe, B. J., Bennell, K. L., & Finch, C. F. (2006). Why are older Australian football players at greater risk of hamstring injury? J Sports Sci Med, 9(4), Gabbe, B. J., Branson, R., & Bennell, K. L. (2006). A pilot randomised controlled trial of eccentric exercise to prevent hamstring injuries in community-level Australian Football. J Sci Med Sport, 9(1-2), Gabbe, B. J., Finch, C., Bennell, K., & Wajswelner, H. (2005). Risk factors for hamstring injuries in community level Australian football. Br J Sports Med, 39(2), Bibliography 180

196 Gandevia, S. (1998). Neural control in human muscle fatigue: changes in muscle afferents, moto neurones and moto cortical drive. Acta physiol Scand, 162(3), Garrett, W., Safran, M., Seaber, A. V., Glisson, R. R., & Ribbeck, B. (1987). Biomechanical comparison of stimulated and nonstimulated skeletal muscle pulled to failure. Am J Sports Med, 15(6), Garrett, W. E., Jr. (1990). Muscle strain injuries: clinical and basic aspects. Med Sci Sports Exerc, 22(4), Garrett, W. E., Jr., Califf, J. C., & Bassett, F. H. (1984). Histochemical correlates of hamstring injuries. Am J Sports Med, 12(2), Goode, A. P., Reiman, M. P., Harris, L., DeLisa, L., Kauffman, A., Beltramo, D.,... Taylor, A. B. (2015). Eccentric training for prevention of hamstring injuries may depend on intervention compliance: a systematic review and meta-analysis. Br J Sports Med, 49(6), Gordon, J. (1991). Spinal mechanisms of motor coordination. Kandel ER, Schwartz JH, Jessell TM. Principles of neural science. 3ª ed. New York (NY): Elsevier Science, Graven-Nielsen, T., Svensson, P., & Arendt-Nielsen, L. (1997). Effects of experimental muscle pain on muscle activity and co-ordination during static and dynamic motor function. Electroencephalogr Clin Neurophysiol, 105(2), Graven Nielsen, T., Lund, H., Arendt Nielsen, L., Danneskiold Samsøe, B., & Bliddal, H. (2002). Inhibition of maximal voluntary contraction force by experimental muscle pain: a centrally mediated mechanism. Muscle Nerve, 26(5), Greig, M. (2008). The influence of soccer-specific fatigue on peak isokinetic torque production of the knee flexors and extensors. Am J Sports Med, 36(7), Guex, K., & Millet, G. P. (2013). Conceptual framework for strengthening exercises to prevent hamstring strains. Sports Med, 43(12), Hagglund, M., Walden, M., & Ekstrand, J. (2006). Previous injury as a risk factor for injury in elite football: a prospective study over two consecutive seasons. Br J Sports Med, 40(9), Bibliography 181

197 Hagglund, M., Walden, M., & Ekstrand, J. (2009). UEFA injury study--an injury audit of European Championships 2006 to Br J Sports Med, 43(7), Hakkinen, K., Alen, M., & Komi, P. V. (1985). Changes in isometric force and relaxtion time, electromyographic and muscle fibre characteristics of human skeletal muscle during strength training and detraining. Acta physiol Scand, 125(4), Häkkinen, K., Kallinen, M., Linnamo, V., Pastinen, U., Newton, R., & Kraemer, W. (1996). Neuromuscular adaptations during bilateral versus unilateral strength training in middle aged and elderly men and women. Acta physiol Scand, 158(1), Hakkinen, K., & Komi, P. V. (1986). Training-induced changes in neuromuscular performance under voluntary and reflex conditions. Eur J Appl Physiol, 55(2), Heiderscheit, B. C., Hoerth, D. M., Chumanov, E. S., Swanson, S. C., Thelen, B. J., & Thelen, D. G. (2005). Identifying the time of occurrence of a hamstring strain injury during treadmill running: a case study. Clin Biomech (Bristol, Avon), 20(10), Heiderscheit, B. C., Sherry, M. A., Silder, A., Chumanov, E. S., & Thelen, D. G. (2010). Hamstring strain injuries: Recommendations for diagnosis, rehabilitation and injury prevention. J Orthop Sports Phys Ther, 40(2), 67. Heiser, T. M., Weber, J., Sullivan, G., Clare, P., & Jacobs, R. R. (1984). Prophylaxis and management of hamstring muscle injuries in intercollegiate football players. Am J Sports Med, 12(5), Henderson, G., Barnes, C. A., & Portas, M. D. (2009). Factors associated with increased propensity for hamstring injury in English Premier League soccer players. J Sci Med Sport, 13(4), Hermens, H. J., Freriks, B., Disselhorst-Klug, C., & Rau, G. (2000). Development of recommendations for SEMG sensors and sensor placement procedures. J Electromyogr Kinesiol, 10, Hickey, J., Shield, A. J., Williams, M. D., & Opar, D. A. (2013). The financial cost of hamstring strain injuries in the Australian Football League. Br J Sports Med. Higbie, E. J., Cureton, K. J., Warren, G. L., & Prior, B. M. (1996). Effects of concentric and eccentric training on muscle strength, cross-sectional area, and neural activation. J Appl Physiol, 81(5), Bibliography 182

198 Hodges, P. W., Mellor, R., Crossley, K., & Bennell, K. (2009). Pain induced by injection of hypertonic saline into the infrapatellar fat pad and effect on coordination of the quadriceps muscles. Arthritis Rheum, 61(1), Hodges, P. W., & Tucker, K. (2011). Moving differently in pain: a new theory to explain the adaptation to pain. Pain, 152(3 Suppl), S Hortobagyi, T., Hill, J. P., Houmard, J. A., Fraser, D. D., Lambert, N. J., & Israel, R. G. (1996). Adaptive responses to muscle lengthening and shortening in humans. J Appl Physiol, 80(3), Houk, J., Crago, P., & Rymer, W. (1980). Functional properties of the Golgi tendon organs. Spinal and Supraspinal Mechanisms of Voluntary Motor Control and Locomotion, 8, Hurley, M. (1997). The effects of joint damage on muscle function, proprioception and rehabilitation. Man Ther, 2(1), Jayaraman, R., Reid, R., Foley, J., Prior, B., Dudley, G., Weingand, K., & Meyer, R. (2004). MRI evaluation of topical heat and static stretching as therapeutic modalities for the treatment of eccentric exercise-induced muscle damage. Eur J Appl Physiol, 93(1), Jonhagen, S., Nemeth, G., & Eriksson, E. (1994). Hamstring Injuries in Sprinters. The role of concentric and eccentric hamstring muscle strength and flexibility. Am J Sports Med, 22, 262. Katz, B. (1939). The relation between force and speed in muscular contraction. J Physiol, 96(1), 45. Kellis, E., & Baltzopoulos, V. (1998). Muscle activation differences between eccentric and concentric isokinetic exercise. Med Sci Sports Exerc, 30(11), Kellis, E., Galanis, N., Natsis, K., & Kapetanos, G. (2009). Validity of architectural properties of the hamstring muscles: correlation of ultrasound findings with cadaveric dissection. J Biomech, 42(15), Kilgallon, M., Donnelly, A. E., & Shafat, A. (2007). Progressive resistance training temporarily alters hamstring torque-angle relationship. Scand J Med Sci Sports, 17(1), Bibliography 183

199 Kinugasa, R., Kawakami, Y., & Fukunaga, T. (2006). Quantitative assessment of skeletal muscle activation using muscle functional MRI. Magn Reson Imaging, 24(5), Klimstra, M., Dowling, J., Durkin, J. L., & MacDonald, M. (2007). The effect of ultrasound probe orientation on muscle architecture measurement. J Electromyogr Kinesiol, 17(4), Komi, P. V., & Buskirk, E. R. (1972). Effect of eccentric and concentric muscle conditioning on tension and electrical activity of human muscle. Ergonomics, 15, Koulouris, G., Connell, D. A., Brukner, P., & Schneider-Kolsky, M. (2007). Magnetic Resonance Imaging Parameters for Assessing Risk of Recurrent Hamstring Injuries in Elite Athletes. Am J Sports Med, 35(9), Kraemer, W. J., Adams, K., Cafarelli, E., Dudley, G. A., Dooly, C., Feigenbaum, M. S.,... American College of Sports, M. (2002). American College of Sports Medicine position stand. Progression models in resistance training for healthy adults. Med Sci Sports Exerc, 34(2), Lee, M., Reid, S., Elliott, B., & Lloyd, D. (2009). Running biomechanics and lower limb strength associated with prior hamstring injury. Med Sci Sports Exerc, 41(10), Lieber, R. L., & Friden, J. (2000). Functional and clinical significance of skeletal muscle architecture. Muscle Nerve, 23(11), Lund, J. P., Donga, R., Widmer, C. G., & Stohler, C. S. (1991). The pain-adaptation model: a discussion of the relationship between chronic musculoskeletal pain and motor activity. Can J Physiol Pharmacol, 69(5), Lynn, R., & Morgan, D. (1994). Decline running produces more sarcomeres in rat vastus intermedius muscle fibers than does incline running. J Appl Physiol, 77(3), Lynn, R., Talbot, J., & Morgan, D. (1998). Differences in rat skeletal muscles after incline and decline running. J Appl Physiol, 85(1), 98. Lynn, S. K., & Costigan, P. A. (2009). Changes in the medial-lateral hamstring activation ratio with foot rotation during lower limb exercise. J Electromyogr Kinesiol, 19(3), e Bibliography 184

200 Mair, S. D., Seaber, A. V., Glisson, R. R., & Garrett, W. E., Jr. (1996). The role of fatigue in susceptibility to acute muscle strain injury. Am J Sports Med, 24(2), Malliaropoulos, N., Mendiguchia, J., Pehlivanidis, H., Papadopoulou, S., Valle, X., Malliaras, P., & Maffulli, N. (2012). Hamstring exercises for track and field athletes: injury and exercise biomechanics, and possible implications for exercise selection and primary prevention. Br J Sports Med, 46(12), Mann, R. A., Moran, G. T., & Dougherty, S. E. (1986). Comparative electromyography of the lower extremity in jogging, running, and sprinting. Am J Sports Med, 14(6), Marieb, E. N., & Hoehn, K. (2007). Human anatomy & physiology: Pearson Education. Markee, J. E., Logue, J. T., Jr., Williams, M., Stanton, W. B., Wrenn, R. N., & Walker, L. B. (1955). Two-joint muscles of the thigh. J Bone Joint Surg Am, 37-A(1), McHugh, M. P., Connolly, D. A., Eston, R. G., Kremenic, I. J., Nicholas, S. J., & Gleim, G. W. (1999). The role of passive muscle stiffness in symptoms of exercise-induced muscle damage. Am J Sports Med, 27(5), Mendiguchia, J., Alentorn-Geli, E., & Brughelli, M. (2012). Hamstring strain injuries: are we heading in the right direction? Br J Sports Med, 46(2), Mendiguchia, J., Arcos, A. L., Garrues, M. A., Myer, G. D., Yanci, J., & Idoate, F. (2013). The use of MRI to evaluate posterior thigh muscle activity and damage during nordic hamstring exercise. J Strength Cond Res, 27(12), Mendiguchia, J., Garrues, M. A., Cronin, J. B., Contreras, B., Los Arcos, A., Malliaropoulos, N.,... Idoate, F. (2013). Nonuniform changes in MRI measurements of the thigh muscles after two hamstring strengthening exercises. J Strength Cond Res, 27(3), Mendiguchia, J., Samozino, P., Martinez-Ruiz, E., Brughelli, M., Schmikli, S., Morin, J. B., & Mendez-Villanueva, A. (2014). Progression of mechanical properties during onfield sprint running after returning to sports from a hamstring muscle injury in soccer players. Int J Sports Med, 35(8), Meyer, R. A., & Prior, B. M. (2000). Functional magnetic resonance imaging of muscle. Exerc Sport Sci Rev, 28(2), Bibliography 185

201 Milner-Brown, H., & Lee, R. (1975). Synchronization of human motor units: possible roles of exercise and supraspinal reflexes. Electroencephalogr Clin Neurophysiol, 38(3), Mjolsnes, R., Arnason, A., Osthagen, T., Raastad, T., & Bahr, R. (2004). A 10-week randomized trial comparing eccentric vs. concentric hamstring strength training in well-trained soccer players. Scand J Med Sci Sports, 14(5), Montgomery, W. H., Pink, M., & Perry, J. (1994). Electromyographic analysis of hip and knee musculature during running. Am J Sports Med, 22, Morgan, D. L. (1990). New insights into the behavior of muscle during active lengthening. Biophysical Journal, 57(2), Moritani, T., & DeVries, H. (1979). Neural factors versus hypertrophy in the time course of muscle strength gain. Am J Phys Med, 58(3), 115. Nagano, Y., Higashihara, A., Takahashi, K., & Fukubayashi, T. (2014). Mechanics of the muscles crossing the hip joint during sprint running. J Sports Sci, 32(18), Narici, M., Roi, G., Landoni, L., Minetti, A., & Cerretelli, P. (1989). Changes in force, crosssectional area and neural activation during strength training and detraining of the human quadriceps. Eur J Appl Physiol, 59(4), Nomura, Y., Kuramochi, R., & Fukubayashi, T. (2015). Evaluation of hamstring muscle strength and morphology after anterior cruciate ligament reconstruction. Scand J Med Sci Sports, 25(3), Onishi, H., Yagi, R., Oyama, M., Akasaka, K., Ihashi, K., & Handa, Y. (2002). EMG-angle relationship of the hamstring muscles during maximum knee flexion. J Electromyogr Kinesiol, 12(5), Ono, T., Higashihara, A., & Fukubayashi, T. (2011). Hamstring functions during hipextension exercise assessed with electromyography and magnetic resonance imaging. Res Sport Med, 19(1), Ono, T., Okuwaki, T., & Fukubayashi, T. (2010). Differences in activation patterns of knee flexor muscles during concentric and eccentric exercises. Res Sport Med, 18(3), Opar, D., Williams, M., & Shield, A. (2012). Hamstring strain injuries: Factors that lead to injury and re-injury. Sport Med, 1;42(3), Bibliography 186

202 Opar, D. A., Drezner, J., Shield, A., Williams, M., Webner, D., Sennett, B.,... Cronholm, P. F. (2013). Acute hamstring strain injury in track-and-field athletes: A 3-year observational study at the Penn Relay Carnival. Scand J Med Sci Sports. 24(4), e Opar, D. A., Piatkowski, T., Williams, M. D., & Shield, A. J. (2013a). A novel device using the Nordic hamstring exercise to assess eccentric knee flexor strength: a reliability and retrospective injury study. J Orthop Sports Phys Ther, 43(9), Opar, D. A., Williams, M. D., Timmins, R. G., Dear, N. M., & Shield, A. J. (2013a). Knee flexor strength and bicep femoris electromyographical activity is lower in previously strained hamstrings. J Electromyogr Kinesiol, 23(3), Opar, D. A., Williams, M. D., Timmins, R. G., Dear, N. M., & Shield, A. J. (2013b). Rate of torque and electromyographic development during anticipated eccentric contraction is lower in previously strained hamstrings. The Am J Sports Med, 41(1), Opar, D. A., Williams, M. D., Timmins, R. G., Hickey, J., Duhig, S. J., & Shield, A. J. (2014). Eccentric Hamstring Strength and Hamstring Injury Risk in Australian Footballers. Med Sci Sports Exerc, 47(4), Orchard, J., & Best, T. M. (2002). The management of muscle strain injuries: an early return versus the risk of recurrence. Clin J Sport Med, 12(1), 3-5. Orchard, J., Marsden, J., Lord, S., & Garlick, D. (1997). Preseason hamstring muscle weakness associated with hamstring muscle injury in Australian footballers. Am J Sports Med, 25(1), Orchard, J., & Seward, H. (2002). Epidemiology of injuries in the Australian Football League; seasons Br J Sports Med, 36(1), Orchard, J., & Seward, H. (2010). Injury Report 2009: Australian Football League. Sport Health, 28(2), Orchard, J., & Seward, H. (2011). Injury report 2010: Australian football league. Sport Health, 29(2), 15. Orchard, J. W. (2001). Intrinsic and extrinsic risk factors for muscle strains in Australian football. Am J Sports Med, 29(3), Bibliography 187

203 Orchard, J. W., James, T., & Portus, M. R. (2006). Injuries to elite male cricketers in Australia over a 10-year period. J Sci Med Sport, 9, Orchard, J. W., & Seward, H. (2009b). Decreased incidence of knee posterior cruciate ligament injury in Australian Football League after ruck rule change. Br J Sports Med, 43(13), Orchard, J. W., Seward, H., & Orchard, J. J. (2013). Results of 2 decades of injury surveillance and public release of data in the Australian Football League. Am J Sports Med, 41(4), Patten, C., Meyer, R. A., & Fleckenstein, J. L. (2003). T2 mapping of muscle. Semin Musculoskelet Radiol, 7(4), Petersen, J., Thorborg, K., Nielsen, M. B., Budtz-Jørgensen, E., & Hölmich, P. (2011). Preventive Effect of Eccentric Training on Acute Hamstring Injuries in Men s Soccer A Cluster-Randomized Controlled Trial. Am J Sports Med, 39(11), Pinniger, G. J., Steele, J. R., & Groeller, H. (2000). Does fatigue induced by repeated dynamic efforts affect hamstring muscle function? Med Sci Sports Exerc, 32(3), Potier, T. G., Alexander, C. M., & Seynnes, O. R. (2009). Effects of eccentric strength training on biceps femoris muscle architecture and knee joint range of movement. Eur J Appl Physiol, 105(6), Pratt, C. A. (1995). Evidence of positive force feedback among hindlimb extensors in the intact standing cat. J Neurophysiol, 73(6), Reeves, N. D., Narici, M. V., & Maganaris, C. N. (2004). In vivo human muscle structure and function: adaptations to resistance training in old age. Exp Physiol, 89(6), Rio, R., Kidgell, D., Purdam, C., Gaida, J., Moseley, G. L., Pearce, A. J., Cook, J. (2015) Isometric exercise induces analgesia and reduces inhibition in patella tendinopathy. Br J Sports Med, 49: Roig, M., O Brien, K., Kirk, G., Murray, R., McKinnon, P., Shadgan, B., & Reid, W. (2009). The effects of eccentric versus concentric resistance training on muscle strength and mass in healthy adults: a systematic review with meta-analysis. Br J Sports Med, 43(8), Bibliography 188

204 Rymer, W., Houk, J., & Crago, P. (1979). Mechanisms of the clasp-knife reflex studied in an animal model. Exp Brain Res, 37(1), Sale, D., Upton, A., McComas, A., & MacDougall, J. (1983). Neuromuscular function in weight-trainers. Exp Neurol, 82(3), Sanfilippo, J. L., Silder, A., Sherry, M. A., Tuite, M. J., & Heiderscheit, B. C. (2013). Hamstring strength and morphology progression after return to sport from injury. Med Sci Sports Exerc, 45(3), Schache, A. G., Dorn, T. W., Blanch, P. D., Brown, N. A., & Pandy, M. G. (2012). Mechanics of the human hamstring muscles during sprinting. Med Sci Sports Exerc, 44(4), Schache, A. G., Wrigley, T. V., Baker, R., & Pandy, M. G. (2009). Biomechanical response to hamstring muscle strain injury. Gait Posture, 29(2), Schoenfeld, B. J., Contreras, B., Tiryaki-Sonmez, G., Wilson, J. M., Kolber, M. J., & Peterson, M. D. (2015). Regional differences in muscle activation during hamstrings exercise. J Strength Cond Res, 29(1), Schuermans, J., Van Tiggelen, D., Danneels, L., & Witvrouw, E. (2014). Biceps femoris and semitendinosus-teammates or competitors? New insights into hamstring injury mechanisms in male football players: a muscle functional MRI study. Br J Sports Med, 48(22), Seward, H., Orchard, J., Hazard, H., & Collinson, D. (1993). Football injuries in Australia at the elite level. Med J Aust, 159, Shellock, F. G., Fukunaga, T., Mink, J. H., & Edgerton, V. R. (1991). Acute effects of exercise on MR imaging of skeletal muscle: concentric vs eccentric actions. AJR Am J Roentgenol, 156(4), Sherry, M. A., & Best, T. M. (2004). A comparison of 2 rehabilitation programs in the treatment of acute hamstring strains. J Orthop Sports Phys Ther, 34(3), Shield, A., & Zhou, S. (2004). Assessing voluntary muscle activation with the twitch interpolation technique. Sport Med, 34(4), Silder, A., Heiderscheit, B. C., Thelen, D. G., Enright, T., & Tuite, M. J. (2008). MR observations of long-term musculotendon remodeling following a hamstring strain injury. Skeletal Radiol, 37(12), Bibliography 189

205 Silder, A., Reeder, S. B., & Thelen, D. G. (2010). The influence of prior hamstring injury on lengthening muscle tissue mechanics. J Biomech, 43(12), Skinner, H., Wyatt, M., Hodgdon, J., Conard, D., & Barrack, R. (1986). Effect of fatigue on joint position sense of the knee. J Orthop Res, 4(1), Sloniger, M. A., Cureton, K. J., Prior, B. M., & Evans, E. M. (1997). Lower extremity muscle activation during horizontal and uphill running. J Appl Physiol (1985), 83(6), Sohn, M. K., Graven Nielsen, T., Arendt Nielsen, L., & Svensson, P. (2000). Inhibition of motor unit firing during experimental muscle pain in humans. Muscle Nerve, 23(8), Sole, G., Milosavljevic, S., Nicholson, H., & Sullivan, S. J. (2011a). Altered muscle activation following hamstring injuries. Br J Sports Med. 46(2): Sole, G., Milosavljevic, S., Nicholson, H. D., & Sullivan, S. J. (2011b). Selective strength loss and decreased muscle activity in hamstring injury. J Orthop Sports Phys Ther, 41(5), Subotnick, S. (1985). The biomechanics of running. Implications for the prevention of foot injuries. Sports Med, 2(2), 144. Suetta, C., Aagaard, P., Magnusson, S., Andersen, L. L., Sipilä, S., Rosted, A.,... Kjaer, M. (2007). Muscle size, neuromuscular activation, and rapid force characteristics in elderly men and women: effects of unilateral long-term disuse due to hiposteoarthritis. J Appl Physiol, 102(3), Sugiura, Y., Saito, T., Sakuraba, K., Sakuma, K., & Suzuki, E. (2008). Strength deficits identified with concentric action of the hip extensors and eccentric action of the hamstrings predispose to hamstring injury in elite sprinters. J Orthop Sports Phys Ther, 38(8), Suter, E., Herzog, W., & Bray, R. (1998). Quadriceps inhibition following arthroscopy in patients with anterior knee pain. Clin Biomech, 13(4-5), Suter, E., & Lindsay, D. (2001). Back muscle fatigability is associated with knee extensor inhibition in subjects with low back pain. Spine, 26(16), E361. Bibliography 190

206 Svensson, P., Houe, L., & Arendt-Nielsen, L. (1997). Bilateral experimental muscle pain changes electromyographic activity of human jaw-closing muscles during mastication. Exp Brain Res, 116(1), Thelen, D. G., Chumanov, E. S., Best, T. M., Swanson, S. C., & Heiderscheit, B. C. (2005). Simulation of biceps femoris musculotendon mechanics during the swing phase of sprinting. Med Sci Sports Exerc, 37(11), Thelen, D. G., Chumanov, E. S., Hoerth, D. M., Best, T. M., Swanson, S. C., Li, L.,... Heiderscheit, B. C. (2005). Hamstring muscle kinematics during treadmill sprinting. Med Sci Sports Exerc, 37(1), Thorborg, K. (2014). What are the most important risk factors for hamstring muscle injury? Clin J Sport Med, 24(2), Thorstensson, A., Grimby, G., & Karlsson, J. (1976). Force-velocity relations and fiber composition in human knee extensor muscles. J Appl Physiol, 40(1), Timmins, R., Bourne, M., Shield, A., Williams, M., Lorenzon, C., & Opar, D. (2015). Short biceps femoris fascicles and eccentric knee flexor weakness increase the risk of hamstring injury in elite football (soccer): a prospective cohort study. Br J Sports Med, [Epub ahead of print]. Timmins, R. G., Opar, D. A., Williams, M. D., Schache, A. G., Dear, N. M., & Shield, A. J. (2014). Reduced biceps femoris myoelectrical activity influences eccentric knee flexor weakness after repeat sprint running. Scand J Med Sci Sports. 24(4), e Timmins, R. G., Ruddy, J. D., Presland, J., Maniar, N., Shield, A. J., Williams, M. D., & Opar, D. A. (2015). Architectural changes of the biceps femoris after concentric or eccentric training. Med Sci Sports Exerc, [Epub ahead of print]. Timmins, R. G., Shield, A. J., Williams, M. D., Lorenzen, C., & Opar, D. A. (2014). Biceps Femoris Long-Head Architecture: A Reliability and Retrospective Injury Study. Med Sci Sports Exerc, 47(5), Upton, P. A., Noakes, T. D., & Juritz, J. M. (1996). Thermal pants may reduce the risk of recurrent hamstring injuries in rugby players. Br J Sports Med, 30(1), Urbach, D., Nebelung, W., Becker, R., & Awiszus, F. (2001). Effects of reconstruction of the anterior cruciate ligament on voluntary activation of quadriceps femoris a prospective twitch interpolation study. J Bone Joint Surg Br, 83(8), Bibliography 191

207 Van Cutsem, M., Duchateau, J., & Hainaut, K. (1998). Changes in single motor unit behaviour contribute to the increase in contraction speed after dynamic training in humans. J Physiol, 513(1), 295. van der Horst, N., Smits, D. W., Petersen, J., Goedhart, E. A., & Backx, F. J. (2015). The Preventive Effect of the Nordic Hamstring Exercise on Hamstring Injuries in Amateur Soccer Players: A Randomized Controlled Trial. Am J Sports Med, 43(6), van Mechelen, W., Hlobil, H., & Kemper, H. C. (1992). Incidence, severity, aetiology and prevention of sports injuries. A review of concepts. Sports Med, 14(2), Veiersted, K. (1991). The reproducibility of test contractions for calibration of electromyographic measurements. Eur J Appl Physiol, 62(2), Verrall, G. M., Slavotinek, J. P., Barnes, P. G., & Fon, G. T. (2003). Diagnostic and prognostic value of clinical findings in 83 athletes with posterior thigh injury: comparison of clinical findings with magnetic resonance imaging documentation of hamstring muscle strain. Am J Sports Med, 31(6), Verrall, G. M., Slavotinek, J. P., Barnes, P. G., Fon, G. T., & Spriggins, A. J. (2001). Clinical risk factors for hamstring muscle strain injury: a prospective study with correlation of injury by magnetic resonance imaging. Br J Sports Med, 35(6), Warren, P., Gabbe, B. J., Schneider-Kolsky, M., & Bennell, K. L. (2010). Clinical predictors of time to return to competition and of recurrence following hamstring strain in elite Australian footballers. Br J Sports Med, 44(6), Weir, J., Housh, D., Housh, T., & Weir, L. (1995). The effect of unilateral eccentric weight training and detraining on joint angle specificity, cross-training, and the bilateral deficit. J Orthop Sports Phys Ther, 22(5), 207. Westing, S., Cresswell, A., & Thorstensson, A. (1991). Muscle activation during maximal voluntary eccentric and concentric knee extension. Eur J Appl Physiol, 62(2), Westing, S. H., Seger, J. Y., Karlson, E., & Ekblom, B. (1988). Eccentric and concentric torque-velocity characteristics of the quadriceps femoris in man. Eur J Appl Physiol, 58(1), Williams, S., Trewartha, G., Kemp, S., & Stokes, K. (2013). A meta-analysis of injuries in senior men's professional Rugby Union. Sports Med, 43(10), Bibliography 192

208 Witvrouw, E., Danneels, L., Asselman, P., D'Have, T., & Cambier, D. (2003). Muscle flexibility as a risk factor for developing muscle injuries in male professional soccer players. A prospective study.. Am J Sports Med, 31(1), Woodley, S. J., & Mercer, S. R. (2005). Hamstring muscles: architecture and innervation. Cells Tissues Organs, 179(3), Woods, C., Hawkins, R., Hulse, M., & Hodson, A. (2002). The Football Association Medical Research Programme: an audit of injuries in professional football-analysis of preseason injuries. Br J Sports Med, 36(6), Woods, C., Hawkins, R. D., Maltby, S., Hulse, M., Thomas, A., & Hodson, A. (2004). The Football Association Medical Research Programme: an audit of injuries in professional football--analysis of hamstring injuries. Br J Sports Med, 38(1), Worrell, T. W., & Perrin, D. (1992). Hamstring muscle injury: the influence of strength, flexibility, warm-up and fatigue. J Orthop Sports Phys Ther, 16(1), Worrell, T. W., Smith, T. L., & Winegardner, J. (1994). Effect of hamstring stretching on hamstring muscle performance. J Orthop Sports Phys Ther, 20(3), Yamamoto, T. (1993). Relationship between hamstring strains and leg muscle strength. A follow-up study of collegiate track and field athletes. J Sports Med Phys Fitness, 33(2), Yao, W., Fuglevand, R. J., & Enoka, R. M. (2000). Motor-unit synchronization increases EMG amplitude and decreases force steadiness of simulated contractions. J Neurophysiol, 83(1), Yeung, S., Suen, A., & Yeung, E. (2009). A prospective cohort study of hamstring injuries in competitive sprinters preseason muscle imbalance as a possible risk factor. Br J Sports Med, 43, Yu, B., Queen, R., Abbey, A., Liu, Y., Moorman, C., & Garrett, W. (2008). Hamstring muscle kinematics and activation during overground sprinting. J Biomech, 41, Zebis, M. K., Skotte, J., Andersen, C. H., Mortensen, P., Petersen, H. H., Viskaer, T. C.,... Andersen, L. L. (2013). Kettlebell swing targets semitendinosus and supine leg curl targets biceps femoris: an EMG study with rehabilitation implications. Br J Sports Med, 47(18), Bibliography 193

209 Zemski, A. J., Slater, G. J., & Broad, E. M. (2015). Body composition characteristics of elite Australian rugby union athletes according to playing position and ethnicity. J Sports Sci, 1-9. Zvijac, J. E., Toriscelli, T. A., Merrick, S., & Kiebzak, G. M. (2013). Isokinetic concentric quadriceps and hamstring strength variables from the NFL Scouting Combine are not predictive of hamstring injury in first-year professional football players. Am J Sports Med, 41(7), Bibliography 194

210 Appendices Appendix A. Muscle activation patterns in the Nordic hamstring exercise: Impact of prior strain injury Publication statement This appendix is comprised of the following paper which was published in the Scandinavian Journal of Medicine and Science in Sports during this student s candidature: Bourne, M., Opar,DA, Williams,MD, Al Najjar, A, Shield, AJ (2015). Muscle activation patterns in the Nordic hamstring exercise: Impact of prior strain injury. Scand J Med Sci Sports. [Epub ahead of print] Appendices 195

211 ABSTRACT This study aimed to determine: 1) the spatial patterns of hamstring activation during the Nordic hamstring exercise (NHE); 2) whether previously injured hamstrings display activation deficits during the NHE; and, 3) whether previously injured hamstrings exhibit altered cross-sectional area. Ten healthy, recreationally active males with a history of unilateral hamstring strain injury underwent functional magnetic resonance imaging (fmri) of their thighs before and after 6 sets of 10 repetitions of the NHE. Transverse (T2) relaxation times of all hamstring muscles (biceps femoris long head, (BFLH); biceps femoris short head (BFSH); semitendinosus (ST); semimembranosus (SM)), were measured at rest and immediately after the NHE and cross-sectional area (CSA) was measured at rest. For the uninjured limb, the ST s percentage increase in T2 with exercise was 16.8, 15.8 and 20.2% greater than the increases exhibited by the BFLH, BFSH and SM, respectively (p<0.002 for all). Previously injured hamstring muscles (n=10) displayed significantly smaller increases in T2 post-exercise than the homonymous muscles in the uninjured contralateral limb (mean difference -7.2%, p=0.001). No muscles displayed significant between limb differences in CSA. During the NHE, the ST is preferentially activated and previously injured hamstring muscles display chronic activation deficits compared to uninjured contralateral muscles. Appendices 196

212 INTRODUCTION Hamstring strains are the most prevalent of all injuries in sports that involve high speed running (Woods et al., 2004; Drezner et al., 2005; Orchard et al., 2006; Brooks et al., 2006a; Brooks et al., 2006b; Ekstrand et al., 2011) and 80% or more of these insults involve the biceps femoris muscle (BF) (Verrall et al., 2003; Askling et al., 2007; Koulouris et al., 2007; Silder et al., 2008). High rates of hamstring muscle strain injury (HSI) recurrence (Heiser et al., 1984; Woods et al., 2004; Orchard et al., 2006; Brooks et al., 2006b) are also troublesome, particularly because re-injuries typically result in greater periods of convalescence than first-time occurrences (Brooks et al., 2006; Ekstrand et al., 2011). These observations highlight the need for improved hamstring prevention and rehabilitation practices while also suggesting that these exercise programs should specifically target (activate) the BF. The importance of eccentric conditioning in HSI prevention is reasonably well recognised (Stanton & Purdham., 1989; Brockett et al., 2001; Askling et al., 2013) and intuitively appealing in light of evidence that hamstring stresses are highest when actively lengthening in the presumably injurious (Thelen et al., 2005; Schache et al., 2009), terminal swing phase of sprinting (Schache et al., 2009; Chumanov et al., 2011). The Nordic hamstring exercise (NHE), the most widely investigated of these eccentric movements, has been reported to reduce first time (Arnason et al., 2008; Petersen et al., 2011) and recurrent (Petersen et al., 2011) HSIs in large scale interventions in soccer. Furthermore, rugby union teams employing the NHE appear to have significantly lower HSI rates than those that do not (Brooks et al., 2006b). Despite the observed benefits of the NHE in reducing injury risk, relatively little is known about the patterns of hamstring muscle activation during this task. One study has reported a non-uniform pattern of hamstring activation during the NHE in male soccer Appendices 197

213 referees (Mendiguchia et al., 2013). However, there is a need to extend these observations, particularly to athletes with a history of HSI, given the prominent role of the NHE in prevention and rehabilitation programs. Fyfe et al. (2013) have recently proposed that the high rates of HSI recurrence might be partly explained by chronic neuromuscular inhibition which results in a reduced capacity to voluntarily activate the BF muscle during eccentric but not concentric knee flexor efforts (Opar et al., 2013a; Opar et al., 2013b). These contraction mode-specific deficits in BF activation can persist despite rehabilitation and return to sport and may mediate preferentially eccentric hamstring weakness (Jonhagen et al., 1994; Croisier et al., 2000; Croisier et al., 2002), reduced rates of knee flexor torque development (Opar et al., 2013b) and persistent BF long head (BFLH) atrophy (Silder et al., 2008), all of which have been observed months to years after HSI. It has been proposed that reduced activation of the BF during active lengthening may diminish the stimuli that would otherwise promote adaptation to the demands of running and strength exercises employed in rehabilitation and training (Opar et al., 2012; Fyfe et al., 2013). However, the aforementioned activation deficits have only been noted during eccentric isokinetic tasks and it remains to be seen whether they also exist during the performance of exercises like the NHE. Further insight into muscle activation patterns during the NHE in uninjured and previously injured muscles will be critical in better understanding how this exercise confers HSIpreventative benefits. Functional magnetic resonance imaging (fmri) allows for assessment of muscle size and this technique is also increasingly employed to investigate muscle activation patterns during exercise (Akima et al., 1999; Mendiguchia et al., 2013; Ono et al., 2011). Functional MRI enables the measurement of T2 relaxation times of imaged skeletal Appendices 198

214 muscles and these values, increase in proportion with exercise intensity (Fleckenstein et al., 1988) and in parallel with electromyographic measures of muscle activation (Adams et al., 1992). Fortunately, the changes in T2 relaxation times last for minutes after intense physical activity (Patten et al., 2003) so post-exercise fmri scans can reveal the extent to which muscles have been activated even after exercise ceases. In addition, because T2 relaxation times are mapped out across cross-sectional images of muscles, fmri is able to determine differences in activation within and between muscles and this excellent spatial resolution overcomes several limitations of surface electromyography (semg) (Adams et al., 1992). The purpose of this study was to use fmri to determine: 1) the spatial patterns of hamstring activation during the NHE; 2) whether previously injured hamstrings display activation deficits compared to homonymous muscles in the uninjured limb during the NHE; and, 3) whether previously injured hamstrings exhibit reduced cross sectional areas (CSAs) compared to homonymous muscles in the uninjured limb. We hypothesised that the hamstrings of uninjured limbs would be activated non-uniformly during the NHE and that previously injured hamstring muscles would display reduced activation and reduced CSA, compared to homonymous muscles in the uninjured limb. METHODS Experimental Design This study used a cross-sectional design in which all participants visited the laboratory on two occasions. During the first, participants were familiarised with the NHE and had baseline anthropometric measures taken. Experimental testing, completed at least seven days later, involved the performance of a NHE session with pre- and post-exercise fmri scans to Appendices 199

215 compare the extent of hamstring muscle activation during the NHE and to assess hamstring muscle CSA between limbs. Participants Ten healthy and recreationally active males, aged (age, 21.6 ± 1.9 years; height, ± 7.4 cm; weight, 81.3 ± 6.5 kg) with a history of unilateral HSI within the previous 24 months were recruited. A sample size of 10 was calculated to provide sufficient statistical power ( 0.80) to avoid a type II error given a presumed effect size of 1.0 for the differences in exercise induced T2 relaxation time changes between muscles of the same limb and between homonymous muscles in opposite limbs when p<0.05. Since this investigation was the first to explore between limb differences in T2 relaxation times following a HSI, the effect size was estimated based on a previous fmri study (Ono et al., 2010) that reported an approximate change (mean ± standard deviation) in T2 of 42±4% in ST, 7±1% in SM and 11±6% in BFLH following eccentric knee flexor exercise using 120% of the 1-repetition maximum load. Participants completed an injury history questionnaire with reference to clinical notes provided by their physical therapist which detailed the location, grade and rehabilitation period of their most recent HSI as well as the total number of HSIs that they had sustained. Participants had all returned to full training and competition schedules, were free of orthopaedic abnormalities of the lower limbs and had no history of neurological or motor disorders. All completed a cardiovascular risk factor questionnaire prior to testing. Additionally, all participants completed a standardised MRI screening questionnaire provided by the imaging facility to ensure that it was safe for them to undergo scanning. Participants were instructed to avoid strength training of the lower body and to abstain from antiinflammatory medications for the week preceding experimental testing. This study was Appendices 200

216 approved by the Queensland University of Technology Human Research Ethics Committee and the University of Queensland Human Ethics Committee. Familiarisation Session A familiarisation session was conducted approximately 8 days (±1 day) before experimental testing. Upon arrival at the laboratory, the participant s height and mass were recorded before they received a demonstration and instructions on the performance of the NHE. From the initial kneeling position with their ankles secured in padded yokes, arms crossed on the chest and hips extended, participants were instructed to lower their bodies as slowly as possible to a prone position (Figure 1). Participants performed only the lowering (eccentric) portion of the exercise and after catching their fall, were instructed to use their arms to push back into the starting position so as to minimise concentric knee flexor activity. Verbal feedback was provided to correct any technique faults while participants completed several practice repetitions (typically three sets of six repetitions). Figure 1. The Nordic hamstring exercise, progressing from left to right. Experimental Session Nordic hamstring exercise protocol Appendices 201

217 Each participant completed 6 sets of 10 repetitions of the NHE with 1-minute rest intervals between sets. During the 1min rest, the participant lay in the prone position. Investigators verbally encouraged maximal effort throughout each repetition. Participants were returned to the scanner immediately (<15s) following the exercise protocol and post-exercise T2- weighted scans began within 90 ± 16s (mean ± SD) following localiser adjustments. Functional magnetic resonance imaging All fmri scans were performed using a Siemens 3-Tesla (3T) TrioTim imaging system with a spinal coil. The participant was positioned supine in the magnet bore with the knees fully extended and hips in neutral, while contiguous MR images were taken of both limbs, beginning immediately superior to the iliac crest and finishing immediately distal to the tibial plateau. Transaxial T2-weighted images were acquired before and immediately after the NHE protocol using a CPMG spin-echo pulse sequence (transverse relaxation time = 2000ms; echo time = 10, 20, 30, 40, 50 and 60ms; number of excitations = 1; slice thickness = 10mm; interslice gap = 10mm). All T2-weighted images were collected using a 180 x 256 image matrix and a 400 x 281.3mm field of view. T1-weighted axial spin-echo images were also obtained but only during the pre-exercise scan (transverse relaxation time = 1180ms; echo time = 12ms; field of view = 400 x mm; number of excitations = 1; slice thickness = 10mm; interslice gap = 10mm). The total acquisition time for pre-exercise images was 15min 10s and for post-exercise images, 10min. Given the high field strength of 3T, a B1 filter was applied to minimise any inhomogeneity in MR images caused by dielectric resonances (De Souza, 2011). Further, to minimise the effects of intramuscular fluid shifts before the preexercise scans, the participant was seated for a minimum of 15 minutes before data acquisition. Appendices 202

218 Data analysis All T1- and T2-weighted fmr images were transferred to a personal computer in the DICOM file format and image analysis software (Sante Dicom Viewer and Editor, Cornell University) was used for subsequent analysis. To evaluate the degree of muscle activation during the NHE protocol, the T2 relaxation times of each hamstring muscle were measured before and immediately after exercise for both the previously injured and uninjured contralateral limb. To quantify T2 relaxation times, the signal intensity of each hamstring muscle (BFLH, BFSH, SM and ST) was measured using a 5 mm² region of interest (ROI) in three slices corresponding to 40%, 50% and 60% respectively, of the distance between the inferior margin of the ischial tuberosity (0%) and the superior border of the tibial plateau (100%) (Ono et al., 2010). For BFSH, a single 5mm² ROI was selected at 50% of thigh length because it was not always possible to identify this muscle in more cranial or caudal slices. All ROIs were selected in the centre of the muscle belly with great care taken to avoid scar and connective tissue, fatty deposits, aponeurosis, tendon, bone and blood vessels. The signal intensity reflected the mean value of all pixels within the ROI and was determined for each ROI across six echo times (10, 20, 30, 40, 50 and 60ms). The signal intensity at each echo time was then graphed to a mono-exponential time curve using a least squares algorithm [(SI= M exp(echo time / T2), where SI is the signal intensity at a specific echo time, and M represents the pre-exercise fmri signal intensity] to extrapolate the T2 relaxation times for each ROI. The absolute T2 relaxation times at all three thigh levels (40%, 50% and 60%) were averaged to provide a mean T2 value for each muscle (BFLH, BFSH, ST, SM) before and after exercise. To assess muscle activation during the NHE protocol, the averaged postexercise T2 value for each muscle was expressed as a percentage change relative to the preexercise value (Fleckenstein et al., 1988; Ono et al., 2011). Muscle cross-sectional area obtained from pre-exercise T1-weighted images was analysed to determine differences in Appendices 203

219 hamstring muscle CSA in limbs with and without a history of HSI. The muscle boundaries of BFLH, SM and ST were identified and traced manually at slices 40%, 50% and 60% of the distance between the inferior margin of the ischial tuberosity (0%) and superior border of the tibial plateau (100%) (Ono et al., 2010) while BFSH was only traced at 50% of thigh length for reasons described previously. Muscle CSA was calculated as the total number of cm 2 within each trace and was averaged across the three slices to provide a mean value for each muscle. The averaged CSA of previously injured muscles was compared with homonymous muscles in the uninjured contralateral limb to evaluate between-limb differences following an HSI. Statistical Analysis To determine the spatial activation patterns in healthy (uninjured) limbs, a repeated measures design linear mixed model fitted with the restricted maximum likelihood (REML) method was used. Exercise-induced percentage changes in T2 relaxation times were compared for each hamstring muscle in the 10 limbs without prior HSI. Muscle (BFLH, BFSH, ST or SM) was the fixed factor with participant as a random factor. When a significant main effect was detected, Bonferroni corrections were used for post-hoc testing and reported as mean difference with 95% CIs. The between-limb analyses of muscle activation and CSA were carried out on all participants. Paired t-tests were used to compare exercise-induced percentage changes in T2 relaxation times and pre-exercise muscle CSA s of the 10 previously injured muscles (7 BFLH, 2 ST, 1 SM) to the homonymous muscles in the uninjured limbs. For these analyses, T2 relaxation times and CSA were reported as uninjured limb versus injured limb mean differences both with 95% CIs. Bonferroni corrections were again used for post-hoc testing and significance Appendices 204

220 was set at p<0.05. Finally, given the possibility that changes in activation patterns and CSA after injury may be muscle-specific, the between-limb analyses (injured v uninjured) were repeated using only the seven participants who had injured their biceps femoris muscles. RESULTS Participant injury histories All participants had a history of unilateral HSI within the previous 24 months, with an average time of 9.8 months (± 8.7 months) since the last insult. At the time of injury, all participants had their HSI diagnosis confirmed with MRI (n=7) or ultrasound (n=3). The details of all participants HSI histories can be found in Table 1. Table 1. Hamstring strain injury (HSI) details of all participants (n=10). Participant Injured Dominant Location Number Months Grade of Rehabilitation Limb Limb of HSIs since last last HSI period (wks) HSI (1-3) 1 Right Yes ST Left No BFlh Right Yes BFlh Right Yes BFlh Right Yes BFlh Left No ST Left No BFlh Right Yes BFlh Right Yes SM Right Yes BFlh Appendices 205

221 Spatial activation of the uninjured limb following the NHE In the uninjured limbs, there was a significant main effect for muscle with respect to exerciseinduced T2 changes following the NHE protocol (p<0.001). Post-hoc tests revealed that the T2 changes induced by exercise within the ST were significantly larger than those observed for the BFLH (ST vs. BFLH mean difference = 16.8%, 95% CI = 7.1 to 26.4%, p=0.001), BFSH (ST vs. BFSH mean difference = 15.8%, CI = 6.1 to 25.4%, p=0.002) and SM (ST vs. SM mean difference = 20.2%, 95% CI = 10.6 to 29.9%, p<0.001) (Figure 2). All other betweenmuscle comparisons in the percentage change of T2 relaxation times were small and nonsignificant (BFLH vs. BFSH, mean difference = 1.0%, 95% CI = -8.7 to 10.6%, p=0.834; BFLH vs. SM, mean difference = 3.4%, 95% CI = -6.2 to 13.1%, p=0.467; BFSH vs. SM, mean difference = 4.5%, 95% CI = -5.2 to 14.1%, p=0.351). Appendices 206

222 Figure 2. Percentage change in functional MRI T2 relaxation times of each hamstring muscle for all 10 uninjured limbs. Values are expressed as a percentage change compared to the values at rest. * indicates p<0.05 when compared to ST with the error bars displaying standard deviation. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. Between-limb comparisons of muscle activation in previously injured hamstring muscles The 10 previously injured hamstring muscles displayed a significantly lower percentage increase in T2 relaxation time (mean difference = -7.2%, 95% CI = -3.8 to -10.7%, p=0.001) (Figure 3) after the NHE than the uninjured homonymous muscles in the contralateral limbs. Appendices 207

223 Figure 3. Percentage change in fmri T2 relaxation times of each hamstring muscle for both the previously injured (inj) and uninjured (uninj) limbs. Values are expressed as a percentage change compared to the values at rest. * indicates p<0.05 when compared to ST with the error bars displaying standard deviation. BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus. Between-limb comparison of hamstring muscle cross-sectional area There were no statistically significant between-limb differences in CSA between the 10 homonymous muscles in the previously injured and uninjured limbs (mean difference = cm 2, CI = 1.21 to -1.80cm 2, p=0.670 (Figure 4). Figure 4. CSA (cm 2 ) of each hamstring muscle (BFLH, biceps femoris long head; BFSH, biceps femoris short head; ST, semitendinosus; SM, semimembranosus) for both the Appendices 208

224 previously injured (Inj) and uninjured (Uninj) contralateral limbs. Values are expressed as means and p<0.05 for all paired comparisons. Error bars depict standard deviation. When only BFLH injuries were considered (n=7), the previously injured BFLH s displayed a significantly lower percentage increase in T2 relaxation time (mean difference = -7.9%, 95% CI = -3.0 to -12.9%, p=0.008) after the NHE than the contralateral uninjured BFLH. However, no additional significant between-limb differences were observed for the other muscles (BFSH mean difference = -0.6%, 95% CI = -7.0 to 5.8, p=0.837; ST mean difference = 4.7%, 95% CI = to 15.6, p=0.382; SM mean difference = 2.7%, 95% CI = -3.7 to 9.1, p=0.400). Previously injured BFLH muscles did not display any significant deficits in CSA when compared to uninjured contralateral BFLH muscles (mean difference = -0.26cm 2, CI = to 1.99cm 2, p=0.785). DISCUSSION The results of this study suggest that in healthy, uninjured limbs, the ST is activated significantly more than other hamstring muscles during the NHE. Furthermore, previously injured hamstring muscles are activated less completely than the homonymous uninjured muscles in the opposite limbs, although these activation deficits are not associated with any significant differences in muscle CSA. Selective recruitment of ST during the NHE is an interesting finding. Maximum forcegenerating capacity of skeletal muscle is dependent on its physiological CSA (Lieber et al., 2000), and as such, pennate muscles are generally stronger than fusiform muscles. Nonetheless, the results of this study suggest that ST, which is long, thin and fusiform (Woodley & Mercer., 2005), is more active during the NHE than BFLH and SM, which are Appendices 209

225 bulkier pennate muscles. These findings are consistent with a recent fmri investigation of the NHE (Mendiguchia et al., 2013) which reported a greater percentage change in T2 for ST (14-20%) than for BFLH (6-7%) and non-significant changes in the SM. Interestingly, Mendiguchia and colleagues (2013) also reported a significant T2 increase in the distal BFSH which remained elevated for 72 hours however, it is important to consider that this delayed increase in T2 is representative of muscle damage and sites of preferential damage may not reflect sites of preferential activation (Bourne et al., 2013). In contrast to the current investigation, recent work employing semg in female athletes reported no significant difference in the extent to which BFLH and ST muscles were activated during the NHE (Zebis et al., 2013). However, semg is prone to cross-talk from neighbouring muscles (Adams et al., 1992) and this may account to some extent for the divergent results. While the mechanism for selective recruitment of ST during the NHE remains unclear, it is possible that differences between hamstring muscle moment arms play a role. At the knee, ST has a larger sagittal plane moment arm than BF and SM (Thelen et al., 2005) and it consequently possesses the greatest mechanical advantage which may explain its preferential recruitment during movements at this joint. Indeed, preferential ST recruitment has previously been observed during eccentric knee flexor exercise using a leg curl machine (Ono et al., 2010) so this strategy appears to be characteristic of hamstring recruitment associated with knee movements when the hip joint angle is fixed. These observations suggest the possibility that the NHE, with its modest activation of BFLH in comparison to ST, may not be the optimal exercise for the prevention of running related strain injury. However, despite this possibility, some large-scale intervention studies have shown that the NHE is effective in reducing first time and recurrent HSIs (Arnason et al., 2008; Petersen et al., 2011; Van der Horst et al., 2014). These benefits may be mediated via improvements in eccentric knee Appendices 210

226 flexor strength (Mjølsnes et al., 2004) and/or a shift of the hamstring torque-joint angle relationship to longer muscle lengths (Brockett et al., 2001). It is possible that even a relatively mild training stimulus is sufficient to protect the BFLH from strain injury or that activation of this muscle progressively increases with regular training and the progressive overload that has been employed in effective intervention programs (Arnason et al., 2008; Petersen et al., 2011). Increased muscle use has previously been observed, via fmri, after two weeks of knee extensor (Akima et al., 1999) and 12 weeks of neck extensor training (Conley et al., 1997), so it is reasonable to expect that several weeks of NHE training would result in athletes activating the BF and SM muscles more completely than we have observed here. Another possibility is that NHE interventions do preferentially stimulate ST adaptations and that the BFLH is effectively protected in running by an enhanced load bearing capacity of its agonist. Nevertheless, there is evidence that BFLH is more selectively activated in the stiff leg deadlift exercise (Ono et al., 2011) so further exploration of the injury prevention benefits of this and other hip-oriented hamstring exercises is warranted. Observations of reduced hamstring activation during the NHE after strain injury are consistent with other findings. Opar et al. (2013a) recently reported inhibition of previously injured BF muscles during eccentric knee flexor contractions using surface electromyography and isokinetic dynamometry. However, by assessing hamstring activation during the NHE, the present findings have more direct implications for conventional rehabilitation practices. Importantly, these activation deficits persist despite apparently successful rehabilitation and a return to pre-injury levels of training and match play, which corroborates previous work (Opar et al., 2013a). Appendices 211

227 The existence of reduced voluntary activation in previously injured hamstring muscles suggests the possibility of a muscle- and possibly contraction mode-specific tension-limiting mechanism(s) at one or more levels of the central nervous system. Neuromuscular inhibition, evident in the form of reduced strength and voluntary activation of surrounding skeletal muscles has been shown to occur after a range of musculoskeletal injuries including anterior cruciate ligament rupture (Urbach et al., 2001) and ankle fractures (Stevens et al., 2006). Recently, it has been suggested that the acute pain associated with a HSI may result in chronic neural inhibition that may compromise hamstring rehabilitation (Fyfe et al., 2013). Short-lasting inhibition constitutes a well-accepted protective strategy to minimise discomfort and preserve the injured structures from further damage (Hodges et al., 2010; Opar et al., 2012). However, activation deficits that persist throughout rehabilitation would reduce the injured muscle s loading, particularly during eccentric contractions and this may compromise hypertrophy and sarcomerogenesis (Timmins et al., 2014; Brockett et al., 2001), both of which are thought to be important in allowing muscles to adapt to the demands of sprinting. Evidence of persistent inhibition, many months after conventional rehabilitation and a full return to training and competition also suggests that inadequate attention has been paid to increasing voluntary activation of the previously injured muscle (Fyfe et al., 2013). Heavy resistance training offers a practical and potent stimulus for improving voluntary activation of skeletal muscle (Akima et al., 1999; Conley et al., 1997). However, in light of recent evidence (Mendiguchia et al., 2013; Ono et al., 2010; Zebis et al., 2013) that different exercises target different portions of the hamstring muscle group, it is possible that some exercises employed in rehabilitation do not optimally target the injured muscle. An improved understanding of the spatial patterns of hamstring muscle activation during different exercises may help practitioners to better tailor rehabilitation programs to the site of injury and should be a focus of future investigations. Appendices 212

228 Despite the presence of activation deficits, the current study found no evidence of atrophy in previously injured hamstring muscles. These findings differ from an earlier investigation that reported chronic atrophy of previously injured BFLH muscles and compensatory hypertrophy of the ipsilateral BFSH 5-23 months following an HSI in recreational athletes (Silder et al., 2008). However, subsequent work from the same group found no evidence of atrophy six months after completion of standardised hamstring rehabilitation (Sanfilippo et al., 2013) and this suggests that different rehabilitation and training practices might at least partially explain the disparate results. Methodological differences between the current study and that of previous work may also explain some of the discrepancies. The current investigation assessed hamstring muscle CSA at 40, 50 and 60% of thigh length, whereas previous investigations (Silder et al., 2008; Sanfilippo et al., 2013) assessed the volume of each hamstring muscletendon unit. Timmins and colleagues (2014) recently reported that ultrasound measures of biceps femoris muscle architecture revealed significantly shorter fascicles coupled with greater pennation angles and no significant differences in muscle thickness between previously injured muscles and uninjured homonymous muscles in the opposite limb. This increase in pennation angle would tend to counter any effects of muscle atrophy on measures of muscle thickness, so measures of cross-section or thickness may not be as sensitive to atrophy as are measures of muscle volume. Participants in this study had received their injuries in the 3 to 24 months prior to being tested so it might be argued that this group is not particularly homogenous in terms of stage of recovery. However, when the activation deficits on the injured limbs were plotted against time since injury, no relationship was observed (R 2 = 0.03) and all participants had resumed full training and competition schedules. Furthermore, there are numerous reports in the literature suggesting that the deficits in eccentric hamstring strength (Jonhagen et al., 1994; Appendices 213

229 Croisier et al., 2002; Lee et al., 2009) and muscle volume (Silder et al., 2008) persist long after strain injury. For example, Lee and colleagues (2009) reported deficits in eccentric knee flexor performance in a group of athletes with an average time since injury of 19 ± 12.5 months. Furthermore, Silder et al. (2008) provided evidence of BFLH atrophy 5-23 months following injury. These observations are consistent with an argument that some effects of hamstring strain are particularly persistent (Fyfe et al., 2013). It should be acknowledged that some limitations are present in the current study. Firstly, because of the retrospective design, we do not know whether activation deficits in previously injured hamstring muscles are the cause or the result of prior HSI. Furthermore, given the absence of a control group with no history of HSI in either limb, it is not possible to know with certainty whether the participants in this study have normal patterns of muscle activation in their uninjured legs. However, similar preferential recruitment of ST has been reported during the NHE (Mendiguchia et al., 2013) and during eccentric knee flexor exercise (Ono et al., 2010) so this pattern of activation is likely to be a robust phenomenon. Finally, it is important to consider that T2 changes are multifactorial and can be influenced by confounding factors such as the metabolic capacity and vascular dynamics of the active tissue (Patten et al., 2003). Such factors have been proposed to account for the high variability in exercise-induced T2 changes between individuals (Patten et al., 2003). To minimise this effect we recruited a homogenous male population with limited ranges in age and levels of physical activity. Furthermore, participants were instructed to avoid strength training of the lower body and to abstain from anti-inflammatory medications for the week preceding experimental testing. Appendices 214

230 Conclusion The current study provides novel insight into the spatial activation patterns of the hamstring muscles during the NHE and how these are altered by prior strain injury. We have provided evidence that ST is selectively activated during the NHE and that previously injured hamstring muscles are less active compared to uninjured homonymous muscles in the contralateral limb. However, these activation deficits are not associated with any significant between-limb differences in muscle CSA. The sub-optimal activation of the BFLH during the NHE may suggest the need to investigate the protective effects of alternative hamstring exercises for the prevention of running related HSI. Furthermore, the observation of persistent activation deficits in previously injured hamstring muscles suggests that conventional rehabilitation practices are not addressing the mechanism(s) underpinning neuromuscular inhibition following HSI (Fyfe et al., 2013). These findings provide evidence for altered muscle use during eccentric hamstring exercise which should be a focus of future investigations. Perspective This study demonstrated that during the performance of the NHE, the ST muscle is activated significantly more than the BF and SM. This may have implications for the use of this exercise in HSI prevention protocols given that the vast majority of HSIs involve the BF as the primary site of injury (Verrall et al., 2003; Askling et al., 2007; Koulouris et al., 2007; Silder et al., 2008). Furthermore, previously injured hamstring muscles were activated significantly less than uninjured contralateral muscles during the NHE, in the absence of diminished cross-sectional areas and despite apparently successful rehabilitation and a return to full training and competition. From a practical point of view, these activation deficits may compromise the rehabilitation process and would likely render the athlete weaker, Appendices 215

231 particularly during eccentric contractions, following a return to sport. Future work should seek to clarify whether these activation deficits are a risk factor for hamstring strain re-injury. Acknowledgements We thank the Queensland Academy of Sport s Centre of Excellence for Applied Sport Science Research, for funding this investigation. The authors also acknowledge the facilities, and the scientific and technical assistance of the National Imaging Facility at the Centre for Advanced Imaging, University of Queensland. Appendices 216

232 References 1. Adams GR, Duvoisin MR, Dudley GA. Magnetic resonance imaging and electromyography as indexes of muscle function. J Appl Physiol. 1992;73(4): Akima H, Takahashi H, Kuno SY, Masuda K, Masuda T, Shimojo H, Anno I, Itai Y, Katsuta S. Early phase adaptations of muscle use and strength to isokinetic training. Med Sci Sports Exerc. 1999;31(4): Arnason A, Andersen TE, Holme I, Engebretsen L, Bahr R. Prevention of hamstring strains in elite soccer: an intervention study. Scand J Med Sci Sports. 2008;18(1): Askling CM, Tengvar M, Saartok T, Thorstensson A. Acute First-Time Hamstring Strains During High-Speed Running A Longitudinal Study Including Clinical and Magnetic Resonance Imaging Findings. Am J Sports Med. 2007;35(2): Askling CM, Tengvar M, Thorstensson A. Acute hamstring injuries in Swedish elite football: a prospective randomised controlled clinical trial comparing two rehabilitation protocols. Br J Sports Med Mar 27; [Epub ahead of print] PMID: Bourne M, Opar DA, Shield, AJ. [Letter to the Editor]. J Strength Cond Res. 2013; 27(10):1. doi: /jsc.0b013e3182a8827c 7. Brockett C, Morgan D, Proske U. Human hamstring muscles adapt to eccentric exercise by changing optimum length. Med Sci Sports Exerc. 2001;33(5): Brooks JHM, Fuller CW, Kemp SPT, Reddin DB. Epidemiology of injuries in English professional rugby union: part 1 match injuries. Br J Sports Med. 2005;39: Brooks JHM, Fuller CW, Kemp SPT, Reddin DB. Incidence, risk, and prevention of hamstring muscle Injuries in professional rugby union. Am J Sports Med. 2006;34(8): Chumanov ES, Heiderscheit BC, Thelen DG. Hamstring musculotendon dynamics during stance and swing phases of high-speed running. Med Sci Sports Exerc. 2011;43(3): Conley MS, Stone MH, Nimmons M, Dudley GA. Resistance training and human cervical muscle recruitment plasticity. J Appl Physiol. 1997;83(6): Croisier JL, Crielaard JM. Hamstring muscle tear with recurrent complaints: an isokinetic profile. Isokin Exer Sci. 2000;8: Croisier JL, Forthomme B, Namurois MH, Vanderthommen M, Crielaard JM. Hamstring muscle strain recurrence and strength performance disorders. Am J Sports Med. 2002;30(2): De Souza P, Vignaud A, Fleury S, Carlier G. Fast monitoring of T1, T2 and relative proton density (MO) changes in skeletal muscles using an IR-TrueFISP sequence. J Magn Reson Imaging. 2011;33: Appendices 217

233 15. Drezner J, Ulager J, Sennet M. Hamstring muscle injuries in track and field athletes: A 3- year study at the Penn Relay Carnival [abstract]. Clin J Sport Med. 2005;15(5): Ekstrand J, Hagglund M, Walden M. Injury incidence and injury patterns in professional football: the UEFA injury study. Br J Sports Med. 2011;45(7): Fleckenstein JL, Canby RC, Parkey RW, Peshock RM. Acute effects of exercise on MR imaging of skeletal muscle in normal volunteers. AJR Am J Roentgenol. 1988;151(2): Fyfe JJ, Opar DA, Williams MD, Shield AJ. The role of neuromuscular inhibition in hamstring strain injury recurrence. J Electromyogr Kinesiol. 2013;23(3): Heiser TM, Weber J, Sullivan G, Clare P, Jacobs RR. Prophylaxis and management of hamstring muscle injuries in intercollegiate football players. Am J Sports Med. 1984;12(5): Hodges PW, Tucker K. Moving differently in pain: a new theory to explain the adaptation to pain. Pain. 2010;152(3): Jonhagen S, Nemeth G, Eriksson E. Hamstring injuries in sprinters. The role of concentric and eccentric hamstring muscle strength and flexibility. Am J Sports Med. 1994;22: Koulouris G, Connell DA, Brukner P, Schneider-Kolsky M. Magnetic resonance imaging parameters for assessing risk of recurrent hamstring injuries in elite athletes. Am J Sports Med. 2007;35(9): Lee MJ, Reid SL, Elliott BC, et al. Running biomechanics and lower limb strength associated with prior hamstring injury. Med Sci Sports Exerc. 2009;41(10): Lieber RL, Friden J. Functional and clinical significance of skeletal muscle architecture. Muscle Nerve. 2000;23(11): Mendiguchia J, Arcos AL, Garrues MA, Myer GD, Yanci J, Idoate F. The use of MRI to evaluate posterior thigh muscle activity and damage during the Nordic hamstring exercise. J Stength Cond Res. 2013;27(12): [Medline] 26. Mjølsnes R, Arnason A, Osthagen T, Raastad T, Bahr R. A 10-week randomized trial comparing eccentric vs. concentric hamstring strength training in well-trained soccer players. Scand J Med Sci Sports. 2004;14(5): Ono T, Higashihara A, Fukubayashi T. Hamstring functions during hip-extension exercise assessed with electromyography and magnetic resonance imaging. Res Sports Med. 2011;19(1): Ono T, Okuwaki T, Fukubayashi T. Differences in activation patterns of knee flexor muscles during concentric and eccentric exercises. Res Sports Med. 2010;18(3): Appendices 218

234 29. Opar DA, Williams MD, Shield AJ. Hamstring strain injuries: factors that lead to injury and re-injury. Sports Med. 2012;42(3): Opar DA, Williams MD, Timmins RG, Dear NM, Shield AJ. Knee flexor strength and bicep femoris electromyographical activity is lower in previously strained hamstrings. J Electromyogr Kinesiol. 2013a;23(3): Opar DA, Williams MD, Timmins RG, Dear NM, Shield AJ. Rate of torque and electromyographic development during anticipated eccentric contraction is lower in previously strained hamstrings. Am J Sports Med. 2013b;41(1): Orchard J, Seward H. Injury Report 2009: Australian Football League. Sport Health. 2010;28(2): Orchard JW, James T, Portus MR. Injuries to elite male cricketers in Australia over a 10- year period. J Sci Med Sport. 2006;9: Patten C, Meyer RA, Fleckenstein MD. T2 Mapping of Muscle. Seminars in Musculoskeletal Radiology. 2003;7(4): Petersen J, Thorborg K, Nielsen MB, Budtz-Jørgensen E, Hölmich P. Preventive Effect of Eccentric Training on Acute Hamstring Injuries in Men s Soccer: A Cluster-Randomized Controlled Trial. Am J Sports Med. 2011;39(11): Sanfilippo JL, Silder A, Sherry MA, Tuite MJ, Heiderscheit BC. Hamstring strength and morphology progression after return to sport from injury. Med Sci Sports Exerc. 2013;45(3): Schache AG, Dorn TW, Blanch PD, Brown NA, Pandy MG. Mechanics of the human hamstring muscles during sprinting. Med Sci Sports Exerc. 2012;44(4): Schache AG, Wrigley TV, Baker R, Pandy MG. Biomechanical response to hamstring muscle strain injury. Gait Posture. 2009;29(2): Silder A, Heiderscheit B, Thelen D, Enright T, Tuite M. MR observations of long-term musculotendon remodeling following a hamstring strain injury. Skeletal Radiol. 2008;37: Stanton P, Purdham C. Hamstring injuries in sprinting - the role of eccentric exercise. J Orthop Sports Phys Ther. 1989;10(9): Stevens JE, Pathare NC, Tillman SM, Scarborough MT, Gibbs CP, Shah P, Jayaraman A, Walter GA, Vandenborne K.. Relative contributions of muscle activation and muscle size to plantarflexor torque during rehabilitation after immobilization. J Orthop Res. 2006;24(8): Thelen DG, Chumanov ES, Hoerth DM, Best TM, Swanson SC, Li L, Young M, Heiderscheit BC. Hamstring muscle kinematics during treadmill sprinting. Med Sci Sports Exerc. 2005;37(1): Appendices 219

235 43. Timmins RG, Shield AJ, Williams MD, Lorenzen C, Opar DA. (2014). Biceps femoris long head architecture: a reliability and retrospective injury study. Med Sci Sports Exerc, accepted. 44. Urbach D, Nebelung W, Becker R, Awiszus F. Effects of reconstruction of the anterior cruciate ligament on voluntary activation of quadriceps femoris a prospective twitch interpolation study. J Bone Joint Surg Br. 2001;83(8): Van der Horst N, Wouter Smits D, Petersen J, Goedhart E, Backx F. The preventive effect of the Nordic Hamstring Exercise on hamstring injuries in amateur soccer players: a randomized controlled trial. Br J Sports Med. 2014;48: doi: /bjsports Verrall GM, Slavotinek JP, Barnes PG, Fon GT. Diagnostic and prognostic value of clinical findings in 83 athletes with posterior thigh injury: comparison of clinical findings with magnetic resonance imaging documentation of hamstring muscle strain. Am J Sports Med. 2003;31(6): Woodley SJ, Mercer SR. Hamstring muscles: architecture and innervation. Cells Tissues Organs. 2005;179(3): Woods C, Hawkins RD, Maltby S, Hulse M, Thomas A, Hodson A. The Football Association Medical Research Programme: an audit of injuries in professional footballanalysis of hamstring injuries. Br J Sports Med. 2004;38(1): Zebis MK, Skotte J, Andersen CH, Mortensen P, Petersen HH, Viskaer TC, Jensen TL, Bencke J, Andersen LL. Kettlebell swing targets semitendinosus and supine leg curl targets biceps femoris: an EMG study with rehabilitation implications. Br J Sports Med. 2013;47(18): Appendices 220

236 Appendix B. Cardiovascular & Injury History Questionnaires CARDIOVASCULAR RISK FACTOR QUESTIONNAIRE To be eligible to participate in the experiment you are required to complete the following questionnaire which is designed to assess the risk of you experiencing any harm during the course of the study. A full and honest disclosure of your medical history is vital for your own safety. Name: Date of Birth: Age: years Weight: kg Height: cm Give a brief description of your average weekly activity pattern: Please tick / answer the appropriate responses for the following questions: 1. Are you overweight? Yes No Don t Know 2. Do you smoke? Yes No Don t Know 3. Does your family have a history of premature (<70 years) cardiovascular problems (eg. heart attack, stroke)? Yes No Don t Know 4. Are you asthmatic? Yes No Don t Know 5. Are you diabetic? Yes No Don t Know 6. Do you have high blood cholesterol levels? Yes No Don t Know 7. Do you have high blood pressure? Yes No Don t Know Appendices 221

237 8. Do you have low blood pressure? Yes No Don t Know 9. Do you have a heart murmur Yes No Don t Know 10. Do you have, or have you ever had, any bloodclots in any of your blood vessels (eg. deepvein thrombosis)? Yes No Don t Know 11. Do you have, or have you ever had, any tendency to bleed for long periods after cutting yourself? Yes No Don t Know 12. Are you currently using any medication Yes No If so, what is the medication? 13. Have you ever experienced any of the following during exertion (exercise or physical labour) or at rest? Light headedness or dizziness Pain in the chest, neck, jaw or arm Numbness or pins-and-needles in any part of your body Loss of consciousness 14. Do you think you have any medical complaint or any other reason which you know of that may prevent you from safely participating in intense exercise? Yes No If yes, please elaborate. I,, believe that the answers to these questions are true and correct. Signed: Date: Appendices 222

238 INJURY QUESTIONNAIRE Section 1: Participant details Participant ID: Date of birth: Sport: Level: 1a) Which is your preferred foot (i.e. to kick with)? Left Right Both 1b) For how long have you played your chosen sport? years 1c) On average how many hours per week do you train: On your own: With the team (if appropriate): hours per week hours per week Section 2: Note: Hamstring injury particulars Please fill out a section for each hamstring injury suffered in chronological order, from the first injury suffered to the most recent. First Hamstring Injury 2a) What date did this hamstring injury occur? / / 2b) What were you doing when the injury occurred? Sprinting Kicking Picking up the ball Other If other, please specify: 2c) Were you: Training on your own Training with the team Competing 2d) On which leg did the injury occur? Left Right 2e) Who diagnosed this injury? Doctor Physiotherapist Other If other, please specify: 2f) Were any imaging investigations performed on this injury? Appendices 223

239 MRI Ultrasound Other If other, please specify: 2g) What was the severity of the strain? Grade I Grade II Grade III Unsure Section 3: Rehabilitation particulars First Hamstring Injury 3a) Did you have any scheduled rehabilitation for this injury? Yes No 3b) Who prescribed your rehabilitation? Rehabilitation/conditioning coach Physiotherapist Other If other, please specify: 3c) What was the length of your rehabilitation (the time taken to resume full competition following injury)? 3d) Describe in as much detail as possible the type of rehabilitation employed: For those who have suffered only one hamstring this is the end of the questionnaire. When more than one hamstring injury has occurred please continue to complete this questionnaire for the relevant number of hamstring strains you have had. Appendices 224

240 Second Hamstring Injury 2a) What date did this hamstring injury occur? / / 2b) What were you doing when the injury occurred? Sprinting Kicking Picking up the ball Other If other, please specify: 2c) Were you: Training on your own Training with the team Competing 2d) On which leg did the injury occur? Left Right 2e) Who diagnosed this injury? Doctor Physiotherapist Other If other, please specify: 2f) Were any imaging investigations performed on this injury? MRI Ultrasound Other If other, please specify: 2g) What was the severity of the strain? Grade I Grade II Grade III Unsure Rehabilitation Particulars 3a) Did you have any scheduled rehabilitation for this injury? Yes No 3b) Who prescribed your rehabilitation? Rehabilitation/conditioning coach Physiotherapist Other If other, please specify: 3c) What was the length of your rehabilitation (the time taken to resume full competition following injury)? 3d) Describe in as much detail as possible the type of rehabilitation employed: Appendices 225

241 Third Hamstring Injury 2a) What date did this hamstring injury occur? / / 2b) What were you doing when the injury occurred? Sprinting Kicking Picking up the ball Other If other, please specify: 2c) Were you: Training on your own Training with the team Competing 2d) On which leg did the injury occur? Left Right 2e) Who diagnosed this injury? Doctor Physiotherapist Other If other, please specify: 2f) Were any imaging investigations performed on this injury? MRI Ultrasound Other If other, please specify: 2g) What was the severity of the strain? Grade I Grade II Grade III Unsure Rehabilitation Particulars 3a) Did you have any scheduled rehabilitation for this injury? Yes No 3b) Who prescribed your rehabilitation? Rehabilitation/conditioning coach Physiotherapist Other If other, please specify: 3c) What was the length of your rehabilitation (the time taken to resume full competition following injury)? 3d) Describe in as much detail as possible the type of rehabilitation employed: Appendices 226

242 Fourth Hamstring Injury 2a) What date did this hamstring injury occur? / / 2b) What were you doing when the injury occurred? Sprinting Kicking Picking up the ball Other If other, please specify: 2c) Were you: Training on your own Training with the team Competing 2d) On which leg did the injury occur? Left Right 2e) Who diagnosed this injury? Doctor Physiotherapist Other If other, please specify: 2f) Were any imaging investigations performed on this injury? MRI Ultrasound Other If other, please specify: 2g) What was the severity of the strain? Grade I Grade II Grade III Unsure Rehabilitation Particulars 3a) Did you have any scheduled rehabilitation for this injury? Yes No 3b) Who prescribed your rehabilitation? Rehabilitation/conditioning coach Physiotherapist Other If other, please specify: 3c) What was the length of your rehabilitation (the time taken to resume full competition following injury)? Appendices 227

243 3d) Describe in as much detail as possible the type of rehabilitation employed: Appendices 228

244 Appendix C. Magnetic Resonance Imaging Screening Form Appendices 229