A philosophical approach to applied sport science with novel insights and applications for sports performance and rehabilitation

Sport science is a multidisciplinary field dedicated to improving sports performance. As interdisciplinary hybrids, applied sport scientists use scientific inquiry to address challenges faced across the various disciplines within sports organizations. This dissertation presents three standalone studies (i.e., articles), which span performance, training, and rehabilitation. In the first study, peak power (PP) and critical power (CP) were used to predict repeated-sprint ability (RSA) during cycling among seven highlytrained male athletes. Using multiple regression, PP and CP accounted for ≥ 92% of the variance in RSA. PP highly correlated (i.e., r ≥ .90) with nonfatigued sprint performance, whereas CP highly correlated with fatigued sprint performance. Overall, the findings suggest that complementary interplay between PP and CP maximize RSA. The purpose of the second study was to determine if single-leg cycling training could improve highintensity running performance among team-sport athletes. Twenty-seven lacrosse players were assigned to a control (CON), single-leg cycling training (SLC), or running training (RUN) group. All groups participated in the same preseason practice and resistance training schedule for 3 weeks. Only SLC and RUN performed high-intensity interval training twice per week. Linear mixed models indicated significant improvements in repeated-sprint ability but not running endurance within CON. Improvements in repeated-sprint ability and running endurance within SLC were 0.57 and 1.97 times greater than CON, respectively. SLC and RUN had similar improvements, indicating that iv single-leg cycling training is an effective cross-training modality that can improve highintensity running performance. In the third study, compensatory patterns after anterior cruciate ligament reconstruction (ACLR) were identified using submaximal cycling biomechanics. Before return to participation, 80% of the participants (n = 15) presented either interlimb or intralimb compensatory patterns, which were characterized by either whole-leg attenuation of power within the surgical limb and/or greater surgical side hip extension action which compensated for reduced surgical knee extension action. Exploratory analysis revealed potential side and sex-specific relationships. Overall, there is now proof-of-concept that submaximal cycling biomechanics can safely evaluate whole-leg dynamic function during early stages of ACLR rehabilitation. In conclusion, the variety of topics presented in this dissertation exemplify the interdisciplinary nature of applied sport science, while providing novel insights and applications for sports performance and rehabilitation.

A PHILOSOPHICAL APPROACH TO APPLIED SPORT SCIENCE WITH NOVEL INSIGHTS AND APPLICATIONS FOR SPORTS PERFORMANCE AND REHABILITATION by Ernest George Rimer A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Exercise and Sport Science Department of Health, Kinesiology, and Recreation The University of Utah August 2019 Copyright © Ernest George Rimer 2019 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Ernest George Rimer has been approved by the following supervisory committee members: James C. Martin , Chair 4/19/2019 Date Approved Nicole Detling , Member 4/19/2019 Date Approved Peter C. Fino , Member 4/19/2019 Date Approved David T. Martin , Member Date Approved Angus Ross , Member Date Approved and by Andrew Mark Williams the Department/College/School of , Chair/Dean of Health, Kinesiology, and Recreation and by David B. Kieda, Dean of The Graduate School. ABSTRACT Sport science is a multidisciplinary field dedicated to improving sports performance. As interdisciplinary hybrids, applied sport scientists use scientific inquiry to address challenges faced across the various disciplines within sports organizations. This dissertation presents three standalone studies (i.e., articles), which span performance, training, and rehabilitation. In the first study, peak power (PP) and critical power (CP) were used to predict repeated-sprint ability (RSA) during cycling among seven highlytrained male athletes. Using multiple regression, PP and CP accounted for ≥ 92% of the variance in RSA. PP highly correlated (i.e., r ≥ .90) with nonfatigued sprint performance, whereas CP highly correlated with fatigued sprint performance. Overall, the findings suggest that complementary interplay between PP and CP maximize RSA. The purpose of the second study was to determine if single-leg cycling training could improve highintensity running performance among team-sport athletes. Twenty-seven lacrosse players were assigned to a control (CON), single-leg cycling training (SLC), or running training (RUN) group. All groups participated in the same preseason practice and resistance training schedule for 3 weeks. Only SLC and RUN performed high-intensity interval training twice per week. Linear mixed models indicated significant improvements in repeated-sprint ability but not running endurance within CON. Improvements in repeated-sprint ability and running endurance within SLC were 0.57 and 1.97 times greater than CON, respectively. SLC and RUN had similar improvements, indicating that single-leg cycling training is an effective cross-training modality that can improve highintensity running performance. In the third study, compensatory patterns after anterior cruciate ligament reconstruction (ACLR) were identified using submaximal cycling biomechanics. Before return to participation, 80% of the participants (n = 15) presented either interlimb or intralimb compensatory patterns, which were characterized by either whole-leg attenuation of power within the surgical limb and/or greater surgical side hip extension action which compensated for reduced surgical knee extension action. Exploratory analysis revealed potential side and sex-specific relationships. Overall, there is now proof-of-concept that submaximal cycling biomechanics can safely evaluate whole-leg dynamic function during early stages of ACLR rehabilitation. In conclusion, the variety of topics presented in this dissertation exemplify the interdisciplinary nature of applied sport science, while providing novel insights and applications for sports performance and rehabilitation. iv TABLE OF CONTENTS ABSTRACT................................................................................................................... iii LIST OF TABLES ........................................................................................................ vii LIST OF FIGURES...................................................................................................... viii ACKNOWLEDGEMENTS ............................................................................................. x Chapters 1 A PHILOSOPHICAL APPROACH TO APPLIED SPORT SCIENCE AND REVIEW OF LITERATURE .......................................................................................... 1 Applied Sport Science ........................................................................................ 1 Review of Literature........................................................................................... 3 Summary............................................................................................................ 9 References.......................................................................................................... 9 2 PEAK POWER AND CRITICAL POWER PREDICT REPEATED-SPRINT ABILITY ...................................................................................................................... 17 Introduction...................................................................................................... 17 Methods ........................................................................................................... 21 Results ............................................................................................................. 29 Discussion ........................................................................................................ 34 Practical Applications....................................................................................... 41 References........................................................................................................ 43 3 SINGLE-LEG CYCLING TRAINING IMPROVES REPEATED-SPRINT AND ENDURANCE RUNNING PERFORMANCE .............................................................. 48 Introduction...................................................................................................... 48 Methods ........................................................................................................... 51 Results ............................................................................................................. 61 Discussion ........................................................................................................ 67 Practical Applications....................................................................................... 72 References........................................................................................................ 74 4 INTERLIMB AND INTRALIMB COMPENSATORY PATTERNS AFTER ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION ..................................... 80 Introduction ....................................................................................................... 80 Methods ............................................................................................................. 83 Results ............................................................................................................... 94 Discussion ....................................................................................................... 107 Practical Applications ...................................................................................... 113 References ....................................................................................................... 115 5 REFLECTIONS OF AN APPLIED SPORT SCIENTIST ....................................... 121 Opening Remarks ............................................................................................ 121 Historical Perspectives ..................................................................................... 122 Final Thoughts................................................................................................. 130 References ....................................................................................................... 131 vi LIST OF TABLES Table 2.1 Repeated-sprint activity in different sports ............................................................ 18 2.2 Reliability data ..................................................................................................... 30 2.3 Multiple regression model summaries ................................................................... 32 2.4 Multiple regression model coefficient summaries ................................................. 33 3.1 Participant characteristics ..................................................................................... 52 3.2 Weekly training schedule...................................................................................... 55 3.3 Single-leg cycling training data............................................................................. 57 3.4 High-intensity running interval training data ......................................................... 59 3.5 Group results from before to after the intervention period ..................................... 61 3.6 Results of linear mixed models ............................................................................. 63 3.7 Linear mixed model results for repeated-sprint ability after removal of influential subject ........................................................................................................................... 65 4.1 Participant background data .................................................................................. 86 4.2 ACL participant information................................................................................. 87 4.3 Relative contribution (%) of each leg to net double-leg power .............................. 95 4.4 Relative contribution of joint-specific actions to whole-leg work .......................... 97 LIST OF FIGURES Figure 2.1 Protocol for time-trials .......................................................................................... 24 2.2 Estimation of critical power .................................................................................. 26 2.3 Protocol for repeated-sprint trials .......................................................................... 27 2.4 Sprint-by-sprint performances .............................................................................. 30 2.5 Relationships of peak power and critical power with repeated-sprint ability and repeated-sprint fatigue ................................................................................................... 31 2.6 The relationship of peak power and critical power with individual sprint performance .................................................................................................................. 34 2.7 Theoretical continuum for repeated-sprint ability due to complementary interplay between peak power and critical power ......................................................................... 41 3.1 Group and individual changes in single-sprint performance .................................. 64 3.2 Group and individual changes in repeated-sprint ability ........................................ 64 3.3. Group and individual changes in repeated-sprint fatigue ....................................... 66 3.4 Group and individual changes in endurance running performance......................... 67 4.1 A schematic overview of the study protocol.......................................................... 84 4.2 Submaximal cycling biomechanics of the control group at 240 W ........................ 96 4.3 Between-subject data for the relative contribution (%) of the surgical leg to the net double-leg power........................................................................................................... 99 4.4 Between-subject data for the relative contribution (%) of the surgical side knee extension (KE) to total whole-leg mechanical work ..................................................... 101 4.5 Interlimb and intralimb compensatory patterns of an ACL participant during submaximal cycling..................................................................................................... 102 4.6 Between-subject data for the relative contribution (%) of the surgical side hip extension (HE) action to the total whole-leg mechanical work ..................................... 105 4.7 Exploratory analysis of interlimb compensatory patterns .................................... 106 ix ACKNOWLEDGEMENTS Countless people supported my doctoral education. Jim Martin, you are brilliant man. I fully appreciate your mentorship and our friendship. I’m grateful to the rest of my committee, including Nicole Detling, Peter Fino, David Martin, and Angus Ross, who were instrumental to the process. At times, Steve Elmer, Chee-Hoi Leong, and Jenna Link served as extended committee members. The critical power study was made possible by Emily Cook and Skullcandy®. Thank you, Josh Mines, Scott Willis, Cody Lockling, and all of his interns for helping with the laboratory protocols. A big high-five goes up for the athletes for their efforts. I would like to acknowledge Rick Kladis and the Utah Lacrosse team for participating in the single-leg cycling study. To the players: I appreciated your hard work and commitment. Gratitude goes to Jon Webster for offering training space, and Zac Harris, Parker Teagle, and Kyler Cipriano for helping with the training and testing. The generosity of Lee Romer from Brunel University made the ACL study possible. His force pedal, and the cooperation of all the athletic trainers, physical therapists, and surgeons facilitated the most fulfilling study of my dissertation. Utah Athletics, Dr. Hill, Kyle Brennan, Nona Richardson, Kyle Whittingham, Larry Krystkowiak, Beth Launiere, Malia Shoji, Phil Cullen, Charles Stephenson, Doug Elisaia, Trevor Jameson, and the rest of my circle of strength provided immense support. Alyssa, Veronika, Layla, and Max, you are the best and I love you. CHAPTER 1 A PHILOSOPHICAL APPROACH TO APPLIED SPORT SCIENCE AND REVIEW OF LITERATURE Applied Sport Science Sport science is a multidisciplinary field that uses scientific inquiry to improve outcomes in sport (e.g., improving performance, improving outcomes after rehabilitation, or reducing injury rate; Bishop, 2008; Haff 2010). By this definition, the role uses the scientific process to answer questions across various domains (e.g., physiology, biomechanics, nutrition, psychology, training, rehabilitation, coaching, endocrinology, motor control, etc.; Haff, 2010) that will translate into new practices aimed at improving sports performance (Bishop, 2008). To do so, the applied sport scientist develops a broad foundation of knowledge across a variety of disciplines in order to address the complex problem of improving sports performance. This dissertation presents three studies which span topics related to performance, training, and rehabilitation. To be clear, I chose breadth over depth for my doctoral research because I sought a broader foundation of knowledge across a variety of disciplines, with which I can apply toward scientific processes dedicated enhancing sports performance. Previous authors have proposed models to guide the applied sport science process (Atkinson, Batterham, & Drust, 2008; Bishop, 2008; Sanctuary, Meir, & Sadler, 2012), but they make little mention of the skills required to carry out the scientific method. The 2 basic steps outlined in elementary school textbooks (Eddleman, Carabatsos, Gutman, & McCallister, 2007) are the same as those in graduate school texts (Creswell, 2012): ask a question or state a problem, perform background research, state a purpose (and hypothesis) for the research, design an experiment, collect data, analyze and interpret data, and communicate results. Science is more complicated than a simple recipe, but developing competency in the above-listed steps will provide a stronger skill set for any scientist. As such, I propose for science to become the applied sport scientist’s greatest competency. The studies shared in the following chapters highlight my first steps toward becoming a better scientist, which was the primary reason I pursued a Ph.D. Indeed, my entire doctoral experience has sought greater expertise in using the scientific method so that I can provide greater value to sports organizations. Expertise is a highly-valued quality, but the tradeoff for expertise can be a lack of flexibility. Individuals with exceptionally high domain expertise can find it challenging to comprehend solutions through others’ perspectives (see Dane, 2010, for a review). Amidst extensive commentary that has questioned the impact sport science has had on sport (Bishop, Burnett, Farrow, Gabbett, & Newton, 2006; Drust & Green, 2013; Glazier, 2015; Martindale & Nash, 2013; Reade, Rodgers, & Spriggs, 2008), some sport scientists lack consideration for the traditions and culture within which they operate (Drust & Green, 2013). Furthermore, sport scientists have been criticized for the monodisciplinary nature of their research (Glazier, 2015) which often leads them to answering research questions that are not relevant to the sports, front offices, coaches, and other support staff they serve (Glazier, 2015; Martindale & Nash, 2013). By embracing science as the core competency, I argue that sport scientists become more flexible by answering questions 3 and providing solutions across a variety of areas of sport. By doing so, the sport scientist becomes an interdisciplinary hybrid (Glazier, 2015), and coincidently abides by the definition of sport science mentioned above, with one exception. As described above, sport science is a multidisciplinary field. Multidisciplinarity sounds attractive, but in actuality, it is a model of cooperation. In multidisciplinary research, scientists from various disciplines work together on a common problem, but they usually work independently, often in sequence (Balagué, Torrents, Hristovski, & Kelso, 2017; Rosenfield, 1992). Conversely, interdisciplinary practice joins professionals who combine knowledge and techniques from their disciplines while becoming full partners and interchangeable leaders collaborating to resolve common problems (Balagué et al., 2017; Rosenfield, 1992). Transformation into an interdisciplinary hybrid enables the sport scientist to cross disciplinary boundaries, which consequently facilitates greater collaboration while adding value to various stakeholders inside the sport organization. As I pursue my transformation into an interdisciplinary applied sport scientist, I felt it was necessary to consider research questions that spanned a variety of topics and disciplines. Review of Literature Study 1 The purpose of the first study presented in this dissertation was to determine if peak power and critical power could predict repeated-sprint ability among trained athletes. Repeated-sprint exercise is defined by two or more short sprints (≤ 10 s) with < 1 min of recovery between sprints (Girard, Mendez-Villanueva, & Bishop, 2011; Glaister, Howatson, Pattison, & McInnes, 2008), and characterized by a reduction in 4 performance across sprints (Girard et al., 2011; Wootton & Williams, 1983). Singlesprint performance is a key predictor of repeated-sprint ability (Bishop, Lawrence, & Spencer, 2003; Bishop & Spencer, 2004; Dawson, Fitzsimons, & Ward, 1993). However, performing consecutive sprints causes progressively greater depletion of muscle phosphocreatine (PCr; Dawson et al., 1997; Gaitanos, Williams, Boobis, & Brooks, 1993), and energy supply limitation is a critical factor related to repeated-sprint fatigue (see Girard et al., 2011, for a review). Given that PCr restoration is an oxidative process (Gabr, El-Sharkawy, Schär, Weiss, & Bottomley, 2011; Kemp, Ahmad, Nicolay, & Prompers, 2015; Mahler, 1985; Yoshida & Watari, 1997), the relationships between aerobic variables and repeated-sprint ability have been studied with mixed results. Previous reports have indicated significant (Bishop & Edge, 2006; Dawson et al., 1993) and nonsignificant (Bishop et al., 2003; Edge, Bishop, Hill-Haas, Dawson, & Goodman, 2005; Mendez-Villanueva, Hamer, & Bishop, 2008) relationships between maximal aerobic capacity and repeated-sprint fatigue. Maximal aerobic capacity depends on both central and peripheral factors, whereas peripheral respiratory factors primarily influence creatine rephosphorylation. The rate of PCr restoration is a direct indicator of the oxidative capacity of the muscle (Haseler, Hogan, & Richardson, 1999; Mahler, 1985; Prompers, Wessels, Kemp, & Nicolay, 2014). Indeed, there are stronger relationships between repeated-sprint fatigue and markers of muscle oxidative capacity measured from muscle biopsies (Thomas, Sirvent, Perrey, Raynaud, & Mercier, 2004). In search of less invasive ways to estimate oxidative capacity of lower extremity muscle, my dissertation committee introduced me to the concept of critical power. Critical power is the highest metabolic rate that can be sustained using wholly oxidative 5 metabolism (Jones & Vanhatalo, 2017; Jones, Vanhatalo, Burnley, Morton, & Poole, 2010; Poole, Burnley, Vanhatalo, Rossiter, & Jones, 2016) whilst preserving muscle PCr concentration (Jones, Wilkerson, DiMenna, Fulford, & Poole, 2008). By definition, then, critical power can be considered an indirect indicator of muscle oxidative capacity. The relationship between critical power and repeated-sprint ability has not been reported in the literature. Therefore, the purpose of my first study was to determine how much variance in repeated-sprint ability could be accounted for by peak power and critical power among trained athletes (See Chapter 2 below). Study 2 Repeated-sprint ability links the first two studies in this dissertation. However, the focus of my second study shifted from performance to training, and from the laboratory to the field. I investigated whether or not a novel training modality single-leg cycling could improve repeated-sprint ability and endurance running performance of team-sport athletes. Several team-sports involve episodes of repeated-sprint activity during competition, including but not limited American football (Iosia & Bishop, 2008), basketball (Matthew & Delextrat, 2009), rugby (Deutsch, Kearnery, and Rehrer, 2007), Australian football (Gray & Jenkins, 2010), soccer (Carling, Le Gall, & Dupont, 2012), field hockey (Spencer et al., 2004), and lacrosse (Polley, Cormack, Gabbett, & Polglase, 2015). In many of those sports, aerobic or intermittent running capacity is generally associated with high-intensity running volume and the frequency of athletic actions during competition (Datson et al., 2014; Helgerud, Engen, Wisløff, & Hoff, 2001; Johnston, Gabbett, Jenkins, & Hulin, 2015; Kempton & Coutts, 2016; Narazaki, Berg, 6 Stergiou, & Chen, 2009; Rampinini et al., 2007). Furthermore, elite team-sport competitors generally possess greater aerobic or high-intensity running capacity than subelite competitors within the same sport (Datson et al., 2014; Gabbett, Kelly, Ralph, & Driscoll, 2009; Jennings et al., 2012; Mohr et al., 2003). For the reasons listed above, others have recommended aerobic training using high-intensity interval training for teamsport athletes (Stone & Kilding, 2009). However, previous authors (Buchheit & Laursen, 2013) have cautioned that high-intensity interval training can increase overall musculoskeletal strain when implemented concurrently with higher volumes of sport participation. Indeed, greater high-intensity running volume can increase the risk of fatigue (Young et al., 2012) or overuse injury (Malone et al., 2018) among athletes. Alternative modes of exercise could support cardiorespiratory conditioning needs without adding greater running volume. Cycling is a possible alternative because it is a nonweight-bearing task that has proven to be an effective cross-training modality (Chan, Ho, & Yung, 2018; Etxebarria, Anson, Pyne, & Ferguson, 2014). Beyond double-leg cycling, two of my dissertation committee members contributed to a study (Abbiss et al., 2011) within which high-intensity single-leg cycling interval training induced greater peripheral adaptations than double-leg cycling interval training among trained cyclists. Likewise, several previous single-leg cycling training studies (Dela, Handberg, Mikines, Vinten, & Galbo, 1993; Henriksson, 1977; Klausen, Secher, Clausen, Hartling, & TrapJensen, 1982; MacInnis, Zacharewicz, et al., 2017; Rud, Foss, Krustrup, Secher, & Hallen, 2012; Saltin et al., 1976; Vincent et al., 2015) reported adaptations indicative of greater muscle oxidative capacity. As discussed above, energy supply is critical to repeated-sprint ability (Girard et 7 al., 2011), and PCr restoration is primarily associated with peripheral factors (Haseler, Hogan, & Richardson, 1999; Mahler, 1985; Prompers, Wessels, Kemp, & Nicolay, 2014). Single-leg cycling training has elicited significant peripheral adaptations that would facilitate faster PCr restoration. Therefore, it seemed plausible that single-leg cycling training could be successfully integrated into an existing sport participation regimen in order to improve measures of high-intensity running performance (see Chapter 3 below). Study 3 To further broaden my scope as an aspiring sport scientist, I shifted focus again for my final study, which focused on recovery from injury. Specifically, I explored whether or not submaximal cycling biomechanics could be used to detect interlimb and intralimb compensatory patterns after anterior cruciate ligament (ACL) reconstruction (ACLR). A large percentage of athletes who return to participation (RTP) after ACLR will suffer a second ACL injury (Wiggins et al., 2016). Functional impairments may be one factor (Paterno et al., 2010) contributing to the fact that athletes are ~5.7 times more likely to suffer another ACL injury after RTP compared to the risk associated with suffering a first ACL injury (Paterno, Rauh, Schmitt, Ford, & Hewett, 2014). Nevertheless, large functional asymmetries are frequently observed at the time of RTP (Myer et al., 2011; Palmieri-Smith & Lepley, 2015; Schmitt, Paterno, & Hewett, 2012;). The most common tests used to assess whole-leg function symmetry at the time of RTP involve maximal effort jumping tasks (see Abrams et al., 2014, for a review). A critical problem with current protocols is that maximal efforts during jumping exercises 8 do not begin until ~4–6 months after ACLR (Adams, Logerstedt, Hunter-Giordano, Axe, & Snyder-Mackler, 2012; van Grinsven, van Cingel, Holla, & van Loon, 2010). Consequently, early rehabilitation phases proceed without knowledge of asymmetry during dynamic whole-leg actions. During early stages of rehabilitation, it not possible to determine dynamic wholeleg functional asymmetries that may be due to muscular deficits (e.g., maximal force or power production), but it is possible to detect neuromuscular compensatory patterns using submaximal exercise. For example, Devita and colleagues (1998) reported altered walking biomechanics 3 weeks after ACLR, which were analogous to the protective compensatory patterns (i.e., reduced knee extension action) reported by Hunt and colleagues (2003, 2004) during submaximal cycling among ACL deficient patients. Specifically, during the stance phase of walking and during the extension phase of cycling, the involved limb among participants from both studies produced larger than usual ankle and hip extensor moments in conjunction with a smaller than usual knee extensor moment (Devita, Hortobagyi, & Barrier, 1998; Hunt, Sanderson, Moffet, & Inglis, 2003). The similarity in these patterns during walking and cycling may be due to the possibility that both locomotive tasks use similar spinal cord based central pattern generators (Zehr & Duysens, 2004). Using walking or cycling to identify neuromuscular compensatory patterns during early stages of rehabilitation may be important because it can be difficult to correct protective compensatory patterns during later stages of rehabilitation (Grooms, Appelbaum, & Onate, 2015; Kapreli & Athanasopoulos, 2006; Nyland, Wera, Klein, & Caborn, 2014). Cycling is often indicated before walking (van Grinsven et al., 2010) after 9 ACLR, and it may be an even more conservative way to identify compensatory patterns during whole-leg dynamic function after ACLR. No previous investigations have used submaximal cycling to identify interlimb and intralimb compensatory patterns after ACLR. In my final study, I sought to provide proof-of-concept that this could be done. More so, I tested the hypotheses that either interlimb or intralimb compensatory patterns would be present before RTP and that they would persist after RTP (see Chapter 4 below). Summary As described above, one goal for my doctoral education was to acquire a greater foundation of knowledge across a broad range of topics relevant to sport. Another goal was to develop greater competency in the skills required for scientific inquiry. The abbreviated review of literature provided above clearly illustrates the broad spectrum of topics covered herein, and my ability to identify questions using the published scientific record. The following chapters will merely provide a progress report on my quest toward becoming a better applied sport scientist. In the final chapter, I reflected on the experience while sharing the real-life stories which inspired each study presented— because it would be a fallacy to not share. Most importantly, I hope the outcomes of each study provide novel insights and applications for sports performance and rehabilitation. References Abbiss, C. R., Karagounis, L. G., Laursen, P. B., Peiffer, J. J., Martin, D. T., Hawley, J. A., … Martin, J. C. (2011). Single-leg cycle training is superior to double-leg cycling in improving the oxidative potential and metabolic profile of trained skeletal muscle. Journal of Applied Physiology, 110(5), 1248-1255. Abrams, G. D., Harris, J. D., Gupta, A. K., McCormick, F. M., Bush-Joseph, C. A., Verma, N. N., ... Bach Jr, B. R. (2014). Functional performance testing after 10 anterior cruciate ligament reconstruction: A systematic review. Orthopaedic Journal of Sports Medicine, 2(1), 1-10. Adams, D., Logerstedt, D., Hunter-Giordano, A., Axe, M. J., & Snyder-Mackler, L. (2012). Current concepts for anterior cruciate ligament reconstruction: A criterion-based rehabilitation progression. Journal of Orthopaedic & Sports Physical Therapy, 42(7), 601-614. Atkinson, G., Batterham, A., & Drust, B. (2008). Is it time for sports performance researchers to adopt a clinical-type research framework? International Journal of Sports Medicine, 29(09), 703-705. Balagué, N., Torrents, C., Hristovski, R., & Kelso, J. A. S. (2017). Sport science integration: An evolutionary synthesis. European Journal of Sport Science, 17(1), 51-62. Bishop, D. (2008). An applied research model for the sport sciences. Sports Medicine, 38(3), 253-263. Bishop, D., Burnett, A., Farrow, D., Gabbett, T., & Newton, R. (2006). Sports-science roundtable: Does sports-science research influence practice? International Journal of Sports Physiology and Performance, 1(2), 161-168. Bishop, D., & Edge, J. (2006). Determinants of repeated-sprint ability in females matched for single-sprint performance. European Journal of Applied Physiology, 97(4), 373-379. Bishop, D., Lawrence, S., & Spencer, M. (2003). Predictors of repeated-sprint ability in elite female hockey players. Journal of Science and Medicine in Sport, 6(2), 199209. Bishop, D., & Spencer, M. (2004). Determinants of repeated-sprint ability in well-trained team-sport athletes and endurance-trained athletes. Journal of Sports Medicine and Physical Fitness, 44(1), 1-7. Buchheit, M., & Laursen, P. B. (2013). High-intensity interval training: Solutions to the programming puzzle. Sports Medicine, 43(10), 927-954. Carling, C., Le Gall, F., & Dupont, G. (2012). Analysis of repeated high-intensity running performance in professional soccer. Journal of Sports Sciences, 30(4), 325-336. Chan, H. C.-K., Ho, W. K.-Y., & Yung, P. S.-H. (2018). Sprint cycling training improves intermittent run performance. Asia-Pacific Journal of Sports Medicine, Arthroscopy, Rehabilitation and Technology, 11, 6-11. 11 Creswell, J. W. (2012). Educational research: Planning, conducting, and evaluating quantitative and qualitative research. 4th ed. Boston, Massachusetts: Pearson Education. Dane, E. (2010). Reconsidering the trade-off between expertise and flexibility: A cognitive entrenchment perspective. Academy of Management Review, 35(4), 579-603. Datson, N., Hulton, A., Andersson, H., Lewis, T., Weston, M., Drust, B., & Gregson, W. (2014). Applied physiology of female soccer: An update. Sports Medicine, 44(9), 1225-1240. Dawson, B., Fitzsimons, M., & Ward, D. (1993). The relationship of repeated sprint ability to aerobic power and performance measures of anaerobic work capacity and power. Australian Journal of Science and Medicine in Sport, 25, 88-93. Dawson, B., Goodman, C., Lawrence, S., Preen, D., Poglaze, T., Fitzsimons, M., & Fournier, P. (1997). Muscle phosphocreatine repletion following single and repeated short sprint efforts. Scandinavian Journal of Medicine & Science in Sports, 7(4), 206-213. Dela, F., Handberg, A., Mikines, K. J., Vinten, J., & Galbo, H. (1993). GLUT 4 and insulin receptor binding and kinase activity in trained human muscle. The Journal of Physiology, 469(1), 615-624. Deutsch, M. U., Kearney, G. A., & Rehrer, N. J. (2007). Time-motion analysis of professional rugby union players during match-play. Journal of Sports Sciences, 25(4), 461-472. Devita, P., Hortobagyi, T., & Barrier, J. (1998). Gait biomechanics are not normal after anterior cruciate ligament reconstruction and accelerated rehabilitation. Medicine and Science in Sports and Exercise, 30(10), 1481-1488. Drust, B., & Green, M. (2013). Science and football: Evaluating the influence of science on performance. Journal of Sports Sciences, 31(13), 1377-1382. Eddleman, S., Carabatsos, M. J., Gutman, S., & McCallister, K. (2007). Focus on life science. Nashua, New Hamshire: CPO Science. Edge, J., Bishop, D., Hill-Haas, S., Dawson, B., & Goodman, C. (2006). Comparison of muscle buffer capacity and repeated-sprint ability of untrained, endurance-trained and team-sport athletes. European Journal of Applied Physiology, 96, 225-234. Etxebarria, N., Anson, J. M., Pyne, D. B., & Ferguson, R. A. (2014). High-intensity cycle interval training improves cycling and running performance in triathletes. European Journal of Sport Science, 14(6), 521-529. 12 Gabbett, T., Kelly, J., Ralph, S., & Driscoll, D. (2009). Physiological and anthropometric characteristics of junior elite and sub-elite rugby league players, with special reference to starters and non-starters. Journal of Science and Medicine in Sport, 12(1), 215-222. Gabr, R. E., El-Sharkawy, A. M. M., Schär, M., Weiss, R. G., & Bottomley, P. A. (2011). High-energy phosphate transfer in human muscle: Diffusion of phosphocreatine. American Journal of Physiology-Cell Physiology, 301(1), C234C241. Gaitanos, G. C., Williams, C., Boobis, L. H., & Brooks, S. (1993). Human muscle metabolism during intermittent maximal exercise. Journal of Applied Physiology, 75(2), 712-719. Girard, O., Mendez-Villanueva, A., & Bishop, D. (2011). Repeated-sprint ability—Part I. Sports Medicine, 41(8), 673-694. Glaister, M., Stone, M. H., Stewart, A. M., Hughes, M., & Moir, G. L. (2004). The reliability and validity of fatigue measures during short-duration maximalintensity intermittent cycling. Journal of Strength & Conditioning Research, 18(3), 459-462. Glazier, P. S. (2015). Towards a Grand Unified Theory of sports performance. Human Movement Science, 56, 139-156. Gray, M. A. J., & Jenkins, D. G. (2010). Match analysis and the physiological demands of Australian football. Sports Medicine, 40(4), 347-360. Grooms, D., Appelbaum, G., & Onate, J. (2015). Neuroplasticity following anterior cruciate ligament injury: A framework for visual-motor training approaches in rehabilitation. Journal of Orthopaedic & Sports Physical Therapy, 45(5), 381393. Haff, G. (2010). Roundtable discussion: Sport science. Strength & Conditioning Journal, 32(2), 33-45. Haseler, L. J., Hogan, M. C., & Richardson, R. S. (1999). Skeletal muscle phosphocreatine recovery in exercise-trained humans is dependent on O2 availability. Journal of Applied Physiology, 86(6), 2013-2018. Helgerud, J., Engen, L. C., Wisløff, U., & Hoff, J. (2001). Aerobic endurance training improves soccer performance. Medicine and Science in Sports and Exercise, 33(11), 1925-1931. Henriksson, J. (1977). Training induced adaptation of skeletal muscle and metabolism during submaximal exercise. The Journal of Physiology, 270(3), 661-675. 13 Hunt, M. A., Sanderson, D. J., Moffet, H., & Inglis, J. T. (2003). Biomechanical changes elicited by an anterior cruciate ligament deficiency during steady rate cycling. Clinical Biomechanics, 18(5), 393-400. Hunt, M. A., Sanderson, D. J., Moffet, H., & Inglis, J. T. (2004). Interlimb asymmetry in persons with and without an anterior cruciate ligament deficiency during stationary cycling. Archives of Physical Medicine and Rehabilitation, 85(9), 1475-1478. Iosia, M. F., & Bishop, P. A. (2008). Analysis of exercise-to-rest ratios during division IA televised football competition. Journal of Strength & Conditioning Research, 22(2), 332-340. Johnston, R. D., Gabbett, T. J., Jenkins, D. G., & Hulin, B. T. (2015). Influence of physical qualities on post-match fatigue in rugby league players. Journal of Science and Medicine in Sport, 18(2), 209-213. Jones, A. M., & Vanhatalo, A. (2017). The ‘Critical Power’ concept: Applications to sports performance with a focus on intermittent high-intensity exercise. Sports Medicine, 47(1), 65-78. Jones, A. M., Vanhatalo, A., Burnley, M., Morton, R. H., & Poole, D. C. (2010). Critical power: Implications for determination of VO2max and exercise tolerance. Medicine and Science in Sports and Exercise, 42(10), 1876-90. Jones, A. M., Wilkerson, D. P., DiMenna, F., Fulford, J., & Poole, D. C. (2008). Muscle metabolic responses to exercise above and below the “critical power” assessed using 31P-MRS. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 294(2), R585-R593. Kapreli, E., & Athanasopoulos, S. (2006). The anterior cruciate ligament deficiency as a model of brain plasticity. Medical Hypotheses, 67(3), 645-650. Kemp, G. J., Ahmad, R. E., Nicolay, K., & Prompers, J. J. (2015). Quantification of skeletal muscle mitochondrial function by 31P magnetic resonance spectroscopy techniques: A quantitative review. Acta Physiologica, 213(1), 107-144. Kempton, T., & Coutts, A. J. (2016). Factors affecting exercise intensity in professional rugby league match-play. Journal of Science and Medicine in Sport, 19(6), 504508. Klausen, K., Secher, N. H., Clausen, J. P., Hartling, O., & Trap-Jensen, J. (1982). Central and regional circulatory adaptations to one-leg training. Journal of Applied Physiology, 52(4), 976-983. MacInnis, M. J., Zacharewicz, E., Martin, B. J., Haikalis, M. E., Skelly, L. E., Tarnopolsky, M. A., … Gibala, M. J. (2017). Superior mitochondrial adaptations 14 in human skeletal muscle after interval compared to continuous single-leg cycling matched for total work. The Journal of Physiology, 595(9), 2955-2968. Mahler, M. (1985). First-order kinetics of muscle oxygen consumption, and an equivalent proportionality between QO2 and phosphorylcreatine level. Implications for the control of respiration. The Journal of General Physiology, 86(1), 135-165. Malone, S., Owen, A., Mendes, B., Hughes, B., Collins, K., & Gabbett, T. J. (2018). High-speed running and sprinting as an injury risk factor in soccer: Can welldeveloped physical qualities reduce the risk? Journal of Science and Medicine in Sport, 21(3), 257-262. Martindale, R., & Nash, C. (2013). Sport science relevance and application: Perceptions of UK coaches. Journal of Sports Sciences, 31(8), 807-819. Matthew, D., & Delextrat, A. (2009). Heart rate, blood lactate concentration, and timemotion analysis of female basketball players during competition. Journal of Sports Sciences, 27(8), 813-821. Mendez-Villanueva, A., Hamer, P., & Bishop, D. (2008). Fatigue in repeated-sprint exercise is related to muscle power factors and reduced neuromuscular activity. European Journal of Applied Physiology, 103(4), 411-419. Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match performance of high-standard soccer players with special reference to development of fatigue. Journal of Sports Sciences, 21(7), 519-528. Myer, G. D., Schmitt, L. C., Brent, J. L., Ford, K. R., Barber Foss, K. D., Scherer, B. J., ... Hewett, T. E. (2011). Utilization of modified NFL combine testing to identify functional deficits in athletes following ACL reconstruction. Journal of Orthopaedic & Sports Physical Therapy, 41(6), 377-387. Narazaki, K., Berg, K., Stergiou, N., & Chen, B. (2009). Physiological demands of competitive basketball. Scandinavian Journal of Medicine & Science in Sports, 19(3), 425-432. Nyland, J., Wera, J., Klein, S., & Caborn, D. N. (2014). Lower extremity neuromuscular compensations during instrumented single leg hop testing 2–10 years following ACL reconstruction. The Knee, 21(6), 1191-1197. Palmieri-Smith, R. M., & Lepley, L. K. (2015). Quadriceps strength asymmetry after anterior cruciate ligament reconstruction alters knee joint biomechanics and functional performance at time of return to activity. The American Journal of Sports Medicine, 43(7), 1662-1669. 15 Paterno, M. V., Rauh, M. J., Schmitt, L. C., Ford, K. R., & Hewett, T. E. (2014). Incidence of second ACL injuries 2 years after primary ACL reconstruction and return to sport. The American Journal of Sports Medicine, 42(7), 1567-1573. Paterno, M. V., Schmitt, L. C., Ford, K. R., Rauh, M. J., Myer, G. D., Huang, B., & Hewett, T. E. (2010). Biomechanical measures during landing and postural stability predict second anterior cruciate ligament injury after anterior cruciate ligament reconstruction and return to sport. The American Journal of Sports Medicine, 38(10), 1968-1978. Polley, C. S., Cormack, S. J., Gabbett, T. J., & Polglaze, T. (2015). Activity profile of high-level Australian lacrosse players. Journal of Strength & Conditioning Research, 29(1), 126-136. Poole, D. C., Burnley, M., Vanhatalo, A., Rossiter, H. B., & Jones, A. M. (2016). Critical power: An important fatigue threshold in exercise physiology. Medicine and Science in Sports and Exercise, 48(11), 2320-2334. Prompers, J. J., Wessels, B., Kemp, G. J., & Nicolay, K. (2014). Mitochondria: Investigation of in vivo muscle mitochondrial function by 31 P magnetic resonance spectroscopy. The International Journal of Biochemistry & Cell Biology, 50, 67-72. Rampinini, E., Bishop, D., Marcora, S. M., Bravo, D. F., Sassi, R., & Impellizzeri, F. M. (2007). Validity of simple field tests as indicators of match-related physical performance in top-level professional soccer players. International Journal of Sports Medicine, 28(3), 228-235. Reade, I., Rodgers, W., & Spriggs, K. (2008). New ideas for high performance coaches: A case study of knowledge transfer in sport science. International Journal of Sports Science & Coaching, 3(3), 335-354. Rosenfield, P. L. (1992). The potential of transdisciplinary research for sustaining and extending linkages between the health and social sciences. Social Science & Medicine, 35(11), 1343-1357. Rud, B., Foss, Ø., Krustrup, P., Secher, N. H., & Hallen, J. (2012). One-legged endurance training: Leg blood flow and oxygen extraction during cycling exercise. Acta Physiologica, 205(1), 177-185. Saltin, B., Nazar, K., Costill, D. L., Stein, E., Jansson, E., Essén, B., & Gollnick, P. D. (1976). The nature of the training response; peripheral and central adaptations to one-legged exercise. Acta Physiologica, 96(3), 289-305. Sanctuary, C. E., Meir, R., & Sadler, I. (2012). The seven step approach to the application of sports science in English professional rugby league: Practical 16 considerations in strength and conditioning. International Journal of Sports Science & Coaching, 7(1), 33-45. Schmitt, L. C., Paterno, M. V., & Hewett, T. E. (2012). The impact of quadriceps femoris strength asymmetry on functional performance at return to sport following anterior cruciate ligament reconstruction. Journal of Orthopaedic & Sports Physical Therapy, 42(9), 750-759. Spencer, M., Lawrence, S., Rechichi, C., Bishop, D., Dawson, B., & Goodman, C. (2004). Time-motion analysis of elite field hockey, with special reference to repeated-sprint activity. Journal of Sports Sciences, 22(9), 843-850. Stone, N. M., & Kilding, A. E. (2009). Aerobic conditioning for team sport athletes. Sports Medicine, 39(8), 615-642. Thomas, C., Sirvent, P., Perrey, S., Raynaud, E., & Mercier, J. (2004). Relationships between maximal muscle oxidative capacity and blood lactate removal after supramaximal exercise and fatigue indexes in humans. Journal of Applied Physiology, 97(6), 2132-2138. van Grinsven, S., Van Cingel, R. E. H., Holla, C. J. M., & Van Loon, C. J. M. (2010). Evidence-based rehabilitation following anterior cruciate ligament reconstruction. Knee Surgery, Sports Traumatology, Arthroscopy, 18(8), 11281144. Vincent, G., Lamon, S., Gant, N., Vincent, P., MacDonald, J., Markworth, J., … Hickey, A. (2015). Changes in mitochondrial function and mitochondria associated protein expression in response to 2-weeks of high intensity interval training. Frontiers in Physiology, 6(51), 1-8. Wootton, S., & Williams, C. (1983). The influence of recovery duration on repeated maximal sprints. In H. G. Knuttgen, J. A. Vogel, & J. Poortmans (Eds.), Biochemistry of exercise, international series in sports science (Vol. 8, pp. 269273). Champaign, Illinois: Human Kinetics. Yoshida, T., & Watari, H. (1997). Effect of circulatory occlusion on human muscle metabolism during exercise and recovery. European Journal of Applied Physiology and Occupational Physiology, 75(3), 200-205 Young, W. B., Hepner, J., & Robbins, D. W. (2012). Movement demands in Australian rules football as indicators of muscle damage. Journal of Strength & Conditioning Research, 26(2), 492-496. Zehr, E. P., & Duysens, J. (2004). Regulation of arm and leg movement during human locomotion. The Neuroscientist, 10(4), 347-361. CHAPTER 2 PEAK POWER AND CRITICAL POWER PREDICT REPEATED-SPRINT ABILITY Introduction Numerous sports involve repeated-sprint activity (Table 2.1), which is defined by a series of two or more sprints (≤ 10 s) interspersed by short recovery intervals (≤ 60 s), and characterized by sprint-by-sprint fatigue (Girard, Mendez-Villanueva, & Bishop., 2011; Wootton & Williams, 1983). Periods of repeated-sprint activity occur during critical moments of play, such as when goals are scored or toward the end of close matches (Girard et al., 2011). Because fatigue during repeated-sprint activity may affect outcomes of intense sports competition (Spencer, Bishop, Dawson, & Goodman, 2005), repeated-sprint ability has gained considerable interest in the sport science community. Cycling has commonly been used to assess repeated-sprint ability among athletes who play ground-based sports (Bishop & Edge, 2006; Bishop, Lawrence, & Spencer, 2003; Bishop & Spencer, 2004; Bishop, Spencer, Duffield, & Lawrence, 2001; Dawson, Fitzsimons, & Ward, 1993; Edge, Bishop, Hill-Haas, Dawson, & Goodman, 2006; Fitzsimons, Dawson, Ward, & Wilkinson, 1993; McGawley & Bishop, 2015; MendezVillanueva, Hamer, & Bishop, 2008). During cycling, repeated-sprint ability is the cumulative mechanical work of all sprints (Bishop et al., 2003; Bishop & Spencer, 2004; Dawson et al., 1993; Girard et al., 2011), whereas repeated-sprint fatigue is a measure of 18 Table 2.1. Repeated-sprint activity in different sports Sport Activity Pattern (citation) Tennis Each rally averages ~8 s (range: 3–15 s) per rally, with ~20 s between rallies (Kovacs, 2006) American Football Eacy play averages ~5 s (range: 2–15 s) per play, with ~36 s (SD: 7 s) between plays (Iosia & Bishop, 2008) Basketball ≤ 1.3 jumps and ≤ 2.8 sprints per minute of continuous play (Matthew & Delextrat, 2009) Rugby ~30% of all hi-intensity efforts lasting ~5 s had ≤ 35 s recovery between efforts (Deutsch, Kearnery, & Rehrer, 2007) Australian Football ≤ 185 max effort bursts per match with ≤ 60 s recovery between ~70% of all max efforts performed (Gray & Jenkins, 2010) Soccer 2–8 repeated-sprint episodes per player per match (Carling, Le Gall, & Dupont, 2012) Field Hockey ~30 sprints lasting 1.5–10 s per match with ≤ 40 s recovery between ~40% of all sprints performed (Spencer et al., 2004) Lacrosse ≤ 2–4 moderate to high-intensity accelerations per minute of continuous play (Polley, Cormack, Gabbett, & Polglase, 2015) Note. Repeated-sprint activity is indicated by evidence suggesting two or more brief sprints (≤ 10 s) or other explosive actions (e.g., jumps or accelerations) with ≤ 60 s recovery between efforts (Girard et al., 2011). As such, the criteria for repeated-sprint activity would be met in sports with more than one explosive action per minute of continuous play. the sprint-by-sprint reduction in mechanical work (Bishop et al., 2001; Dawson et al., 1993, Girard et al., 2011). Repeated-sprint ability is highly correlated with peak mechanical power (Bishop et al., 2003; Bishop & Spencer, 2004; Dawson et al., 1993). However, peak mechanical power is also highly correlated with repeated-sprint fatigue (Bishop et al., 2001; Bishop et al., 2003; Bishop & Spencer, 2004; Dawson et al., 1993; Mendez-Villanueva et al., 2008), that is, repeated-sprint fatigue increases as peak power increases. Therefore, considerable attention has been paid to the various underpinning factors related to repeated-sprint fatigue. 19 Energy supply limitations may be one of the most important factors associated with repeated-sprint fatigue (see Girard et al., 2011, for a review). During a single maximal sprint (≤ 10 s), over 50% of the total energy is supplied by cytosolic turnover of adenosine triphosphate via the phosphagen system (Gaitanos, Williams, Boobis, & Brooks, 1993). Consequently, muscle phosphocreatine (PCr) concentration ([PCr]) can plummet to as low as ~43–55% of resting levels after a single 6 s cycling sprint. After 5– 10 repeated sprints, [PCr] can further decrease to as low as ~16–25% (Dawson et al., 1997; Gaitanos et al., 1993), indicating incomplete PCr resynthesis between sprints. PCr resynthesis follows a monoexponential time-course (Baker, McCormick, & Robergs, 2010) with time-constants ranging from approximately 20 to 80 s (Bogdanis, Nevill, Boobis, Lakomy, & Nevill, 1995; Haseler, Hogan, & Richardson, 1999; McCully, Vandenborne, DeMeirleir, Posner, & Leigh Jr, 1992; Yoshida & Watari, 1997). Interestingly, Elmer and colleagues (2013) observed a monoexponential time-course of power recovery after a 30 s cycling sprint with a time-constant (~44 s) that was consistent with others (Bogdanis et al., 1995) who have reported that power recovery mirrors the time-course of PCr resynthesis. As such, I agree with the consensus (Girard et al., 2011) that sprint-by-sprint reductions in performance are closely related to availability of PCr. PCr resynthesis is primarily an oxidative process (Gabr, El-Sharkawy, Schär, Weiss, & Bottomley, 2011; Kemp, Ahmad, Nicolay, & Prompers, 2015; Mahler, 1985; Yoshida & Watari, 1997). In fact, rate of PCr resynthesis is considered a valid indicator of muscle oxidative capacity (Haseler et al., 1999; Mahler, 1985; Prompers, Wessels, Kemp, & Nicolay, 2014). Muscle oxidative capacity is negatively correlated to repeatedsprint fatigue, that is, as muscle oxidative capacity increases, sprint-by-sprint fatigue 20 decreases (Thomas, Sirvent, Perrey, Raynaud, & Mercier, 2004). Assessing PCr restoration or muscle oxidative capacity involves expensive and/or invasive techniques such as 31-Phosphorus magnetic resonance spectroscopy (Haseler et al., 1999; Kemp et al., 2015; Yoshida & Watari, 1997) or muscle biopsy (Bogdanis et al., 1995; Dawson et al., 1997; (Gaitanos et al., 1993; Thomas et al., 2004). In applied settings, it may be useful to establish a relationship between repeated-sprint fatigue and indicators of muscle oxidative capacity using less invasive, more readily available methods. One viable option is the estimation of critical power. Critical power is considered an indirect indicator of the oxidative capacity of the muscle because it is the highest metabolic rate that can be sustained using primarily oxidative metabolism (Jones & Vanhatalo, 2017; Jones, Vanhatalo, Burnley, Morton, & Poole, 2010; Poole, Burnley, Vanhatalo, Rossiter, & Jones, 2016) while preserving [PCr] (Jones, Wilkerson, DiMenna, Fulford, & Poole, 2008). It is possible that critical power may relate to repeated-sprint ability, but to the best of my knowledge, this has never been studied. Together, peak power and critical power may serve complementary roles in repeated-sprint ability. Whereas greater peak power facilitates greater cumulative work across sprints (Bishop et al., 2003; Bishop & Spencer, 2004; Dawson et al., 1993), critical power may facilitate more rapid PCr resynthesis between sprints or greater oxidative ATP supply during sprints (McGawley & Bishop, 2015) performed with reduced muscle [PCr] (Dawson et al., 1997; Gaitanos et al., 1993). Therefore, in this study, I investigated the relationships of peak power and critical power with repeatedsprint ability among trained athletes. Additionally, I explored whether or not peak power and critical power could predict repeated-sprint ability during successive series of 21 repeated sprints using multiple linear regression. To further elucidate this hypothesis, I determined if the relationships of peak power and critical power with individual sprint performance would change with each successive sprint. Specifically, I expected peak power to have a stronger relationship to performance during earlier sprints, and that critical power would become a stronger predictor during later sprints. Methods Experimental overview The methods used in this study were approved by the University of Utah’s Institutional Review Board. Seven highly trained competitive male athletes visited the Neuromuscular Function Laboratory on six separate occasions within 21 days. They performed cycling time-trials during the first three visits to determine critical power. During each of the final three visits, two consecutive series of repeated sprints were completed to determine peak power, repeated-sprint ability, and repeated-sprint fatigue. Multiple regression analyses were used to determine how much variance in repeatedsprint ability and fatigue could be accounted for by peak power and critical power. Pearson’s correlation coefficients were also calculated between each sprint and either peak power or critical power to determine if their relationships with sprint performance changed across successive sprints. Participants Seven highly trained male athletes (mean [SD]: 26.2 [3.4] years; 178 [5] cm; 84.2 [9.1] kg) from different sports provided written informed consent to participate in this study. They had highly competitive backgrounds, including two National Collegiate Athletics Association (NCAA) alpine skiers and an NCAA cross-country skier, all three 22 of whom formerly competed for their respective national teams, a 6-year Association of Volleyball Professionals beach volleyball veteran, a 9-year National Hockey League veteran, and two USA Bobsled racers who converted after completing their NCAA playing careers. They also celebrated significant accolades including a four-time NCAA All-American, an NCAA overall title winner, a world-championship team member, an eventual Olympian, a two-time Stanley Cup winner, and a world-record holder in a physical performance feat (not revealed to maintain confidentiality). At the time of the study, all participants followed a year-round strength and conditioning regimen, and they were free of injury. Familiarization During the first visit, participants were fitted for cleated cycling shoes (Specialized Bicycle Components Inc., Morgan Hill, GA, USA), seat height, and handlebar dimensions which were held constant throughout the study. They were familiarized with the isokinetic cycle ergometer used during all laboratory visits. This ergometer was described previously (Martin & Brown, 2009). Briefly, the flywheel of a Monark cycle ergometer (Vansbro, Sweden) was driven via pulleys and a belt with a 3750-W direct-current motor (model CDP3605; Baldor Electric Company, Fort Smith, AR) operated by a speed controller equipped with regenerative braking capability (model RG5500U; Minarik, Glendale, CA). When participants applied power to the ergometer, the motor acted as a generator and the resulting current was dissipated by a resistor and heat sink built into the speed controller. The controller could, therefore, maintain a specific pedaling rate while resisting up to 3750 W. The ergometer was equipped with an SRM power meter (Schoberer Rad Messtechnik, Ju ̈lich, Germany) capable of providing 23 valid and reliable measures (Gardner et al., 2004; Martin, Milliken, Cobb, McFadden, & Coggan, 1998). Power during the exercise trials was recorded every 0.5 s (i.e., 2 Hz), and total work (J) was calculated by summing the incremental work values sampled at each data point (incremental work = sampled power x 1/f). After becoming familiar with the isokinetic cycle ergometer, participants were introduced to custom audio files that guided them through the exercise protocols using similar audial tones during all sessions. Specifically, a soft triple-beep and soft doublebeep respectively indicated 1 min and 30 s warnings prior to the start of an exercise trial. Soft single-beeps were used to indicate 10 s before the start of exercise, to give a 4 s countdown before sprints, and to provide 1 min notifications during the warm-up and time-trials. A hard single-beep and hard double-beep indicated the beginning and end of the time-trials and sprints. The audio files guided each session from the start of the warm-up until the end of the time-trial or repeated-sprint trials. The warm-up was identical across all six sessions (Figure 2.1). Briefly, it began with 5 min of isokinetic cycling at 90 rpm and self-selected work rate (mean: 125 W; SD: 22 W). A 2 min passive rest preceded three baseline sprints (3 s), with each sprint separated by 2 min of passive recovery. Those sprints were performed seated and included a rolling start, initiated by a 4 s countdown using single soft-beeps each second. During that countdown, the participants were instructed to “catch up” to the speed of the isokinetic flywheel (120 rpm) while producing minimal mechanical power. With the sound of a hard single-beep, the participants were instructed to give maximal efforts until 3 s later when hard double-beep indicated when to stop. It should be noted that noncyclists require 3 days of practice before producing reliable 24 Power (W) VISIT 1 Sprint Rehearsal Repeated-Sprint 3x3s Rehearsal w/2 min Rest 2 x (2 x 8 s w/22 s Rest) 3 min Time-Trial Warm-Up 5 min Rest 3 min Rest 5 min VISIT 2 Power (W) 6 min Time-Trial Power (W) VISIT 3 15 min Time-Trial Time Figure 2.1. Protocol for time-trials. The exercise protocols, including the warm-up, sprint rehearsal, and time-trials were the same across the first three laboratory visits. The only difference was that a 3-min, 6-min, and 15-min time-trial was performed during the first, second, and third visits, respectively. measures of maximal cycling power (Martin, Diedrich, & Coyle, 2000). Therefore, performing the baseline sprints as part of the warm-up during the first three sessions insured greater reliability when they performed the repeated-sprint testing during the final three sessions. After the warm-up, participants were also familiarized with the format of the repeated-sprint protocol (Figure 2.1). To prevent fatigue during the upcoming time-trial, however, participants did not rehearse the entire repeated-sprint protocol. Instead, they rehearsed the first two sprints of each series of repeated sprints. Specifically, they rested 2 mins after the third baseline sprint, and then performed two maximal sprints (8 s) separated by 22 s of passive recovery. After 3 min of passive rest, two more sprints (8 s) 25 were performed. Similar to the baseline sprints, the 8 s sprints were seated and initiated using a rolling start and the same audio signals as previously described. Critical power Similar to previous authors (Quod, Martin, Martin, & Laursen, 2010) who have used fixed-duration time-trials to estimate critical power, participants performed a 3 min, 6 min, and 15 min time-trial during the first, second, and third sessions, respectively (Figure 2.1). Those time-trials began 5 min after completion of the repeated-sprint familiarization. During each time-trial, participants remained seated and were instructed to give a best effort in order to achieve the greatest average power possible. Continuous encouragement was provided throughout the time-trials, and a soft beep alerted participants of each elapsed minute. To estimate critical power, the hyperbolic relationship between power and duration (Jones et al., 2010) was converted into a linear model using first principles (Jones & Vanhatalo, 2017). Specifically, the total cumulative work (J) of each trial was graphed over the duration of the effort, and linear regression (Figure 2.2) was used to determine critical power (CP) and work prime (W’), the finite amount of work that can be performed above CP (Jones et al., 2010; Jones & Vanhatalo, 2017; Poole et al., 2016). To avoid placing additional time-demands on the participants, familiarization trials did not precede the time-trial sessions, but previous authors have indicated acceptable typical error of average power (~3.5%) during fixed duration time-trials (Quod et al., 2010). CP was expressed relative to body weight (W/kg) and used as a predictor for repeated-sprint ability and fatigue, and each of the 16 sprints. 26 4500 4000 W = CPt + W' 3500 Work (J) 3000 2500 2000 1500 1000 500 0 0 200 400 Time (s) 600 800 1000 Figure 2.2. Estimation of critical power. By regressing the total work (measured in J, and indicated by W) on time (measured in s, and indicated by t). Critical power (measured in W) was estimated by the slope (CP), and work prime (measured in J) was estimated by the intercept (W’). Repeated-sprint ability The 1st day of repeated-sprint testing began at least 48 h after the third time-trial. All three repeated-sprint sessions (Figure 2.3) were performed at the same time of day to account for the diurnal rhythm in maximal short-term power, which is highly correlated with the variation in core body temperature (Chtourou & Souissi, 2012). Although each session was scheduled between similar training activities, baseline sprints were required before the repeated-sprint protocol on each day because ongoing athletic activities (e.g., training and competition) can cause significant fluctuation in maximal cycling power (McLean, Petrucelli, & Coyle, 2012). The repeated-sprint protocol consisted of two series of eight cycling sprints (8 s) beginning every 30 s (i.e., 22 s recovery), with a 3 min recovery between each series. Similar to the baseline sprints, each sprint in the repeated-sprint trial started with a 4-s rolling start guided by soft single-beeps. During the rolling start, participants were asked to watch the SRM power meter and stay below their CP (estimated from the three time- 27 Power (W) VISITS 4-6 Warm Up 5 min Repeated-Sprint Trials 2 x (8 x 8 s w/ 22 s Rest) w/3 min Rest Criterion Sprints 3x3s w/2 min Rest Rest 3 min Rest 5 min Time Figure 2.3. Protocol for repeated-sprint trials. After a brief warm-up at a self-selected intensity that remained below critical power, there was a 2-min passive recovery before three criterion sprints which were used as a pacing index for the subsequent series of repeated sprints. trials) in order to avoid expending muscle PCr (Jones et al., 2008). As described, the sprints began with a hard single-beep, and the participants were instructed to give maximal efforts for 8 s at which time a hard double-beep alerted them to stop pedaling. By this time (i.e., the fourth session), participants were completely familiar with the audial signals, which facilitated excellent control over the repeated-sprint protocol. To prevent pacing during the repeated-sprint protocol, participants were encouraged to achieve a peak power in their best 8-s sprint that was at least 95% of their best 3-s baseline sprint. Because there was not enough time to examine the SRM power data between sprints, participants completed the entire repeated-sprint trial, and were given feedback immediately after the session about whether or not they met the 95% criteria. Overall, peak power in the best 8-s sprint was 98.7% (SD: 2.4%) of their best 3-s sprint. Data from the repeated-sprint testing days were processed to determine peak power, repeated-sprint ability, and repeated-sprint fatigue for each individual. In each session, peak power (PP) was determined by the greatest power (W/kg) achieved during 28 the 3-s baseline sprints or the 8-s repeated sprints. Repeated-sprint ability was determined by the total cumulative work (J/kg; Bishop et al., 2003; Bishop & Spencer, 2004; Dawson et al., 1993; Girard et al., 2011) across the first eight sprints (RSA1-8) and the final eight sprints (RSA9-16). Repeated-sprint fatigue for each series of sprints (RSF1-8 and RSF9-16) was expressed as a percentage using the sprint decrement score (Dawson et al., 1993; Fitzsimons et al., 1993): RSF (%) = (IRSA-RSA) / IRSA x 100 (1) where IRSA is the ideal repeated-sprint ability by which the best single-sprint performance (J/kg) would be maintained across all sprints. The usefulness of describing repeated-sprint fatigue using a relative measure has been questioned (see Oliver, 2009, for a review). Nevertheless, the sprint decrement score is considered the most valid index of repeated-sprint fatigue (Girard et al., 2011; Glaister, Stone, Stewart, Hughes, & Moir, 2004). To provide the best estimate of each participant’s performance and to reduce the effect of day to day variability, the average performance across all 3 days of testing were used for analysis. Statistical analysis Statistical analyses were performed using Microsoft Excel (Microsoft Corporation; Redmond, WA) and RStudio (RStudio, Inc.; Boston, MA). All three repeated-sprint trials were used to calculate the intraclass correlation coefficient (ICC), typical error of measurement (TE), and coefficient of variation (CV%) for PP, RSA1-8, RSA9-16, RSF1-8, and RSF9-16 according to Hopkins (2000). Pearson’s correlation coefficients (r) were calculated between the independent (PP and CP) and dependent variables (RSA1-8, RSA9-16, RSF1-8, and RSF9-16; α = .05). Separate multiple regression 29 analyses were used to test the hypothesis that PP and CP could account for a significant amount of variance in RSA1-8, RSA9-16, RSF1-8, and RSF9-16. With the limited sample size used in this study (N = 7), good predictability was indicated by an R2 ≥ .90 (Knofczynski & Mundfrom, 2008), which was supported by conservative a priori power analysis (α = .01; β = .10) using freeware software (G* Power version 3.1; Faul, Erdfelder, Lang, & Buchner, 2007). The same software was also used to perform post hoc analyses of the actual statistical power (1-β) for each regression model. The global validation of linear model assumptions (Peña & Slate, 2006) was used to confirm whether or not the assumptions of the multiple regression models were met. Standardized regression coefficients of each statistically significant model (p ≤ .01) were used to compare changes in the relative importance of each predictor from the first to second series of repeated sprints (Field, 2013). To further realize potential relationships between PP or CP with nonfatigued or fatigued sprint performance, Pearson’s correlation coefficients (r) were also calculated between PP or CP and performance (J/kg) during each of the 16 sprints (α = .05). Results The 3 days of repeated-sprint testing yielded high reliability (i.e., ICC ≥ .90; Atkinson & Nevill, 1998) for PP, RSA1-8, and RSA9-16, but questionable reliability (i.e., ICC < .80; Atkinson & Nevill, 1998) for RSF1-8 and RSF9-16 (Table 2.2). During the repeated-sprint protocol, a gradual sprint-by-sprint reduction in performance occurred across the first eight sprints (Figure 2.4). After the 3 min recovery interval, sprint performance (J/kg) returned to ~92.5% (SD: 4.7%) of the first sprint, and then gradually decreased through the 16th sprint (Figure 2.4). Overall RSA9-16 was 89.1% (SD: 30 Table 2.2. Reliability data PP RSA1-8 RSA9-16 RSF1-8 RSF9-16 ICC .908 .916 .937 .218 .748 TE 0.4 W/kg 23.0 J/kg 23.5 J/kg 3.2% 2.7% 3.0 3.4 4.1 26.5 19.2 CV (%) Note. ICC = intraclass correlation coefficient; TE = typical error of measurement expressed in the units of measurement; CV = coefficient of variation. 110.0 100.0 Total Work (J/kg) 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 Figure 2.4. Sprint-by-sprint performances. The average cumulative work (J/kg) of each sprint (S1, S2, etc.) is illustrated. Note that error bars indicate the between-subject SD. 5.1%) of RSA1-8. The relationship between PP and repeated-sprint ability decreased from RSA1-8 (r = .963; p = .0005) to RSA9-16 (r = .851; p = .015; Figure 2.5). Conversely, CP was more highly correlated to RSA9-16 (r = .924; p = .003) than to RSA1-8 (r = .862; p = .013; Figure 2.5) suggesting that muscle oxidative capacity, as represented by CP, became more important in the second set of repeated sprints. Further, PP did not correlate with RSF1-8 (r = -.258; p = .576) or RSF9-16 (r = -.317; p = .488), but CP had large negative correlations with RSF1-8 (r = -.701; p = .079) or RSF9-16 (r = -.721; p = .068). The multiple regression models (Table 2.3) indicate that PP and CP yielded good 31 Repeated-Sprint Ability (J/kg) 850 800 850 A r1-8 = 0.963 * 750 750 700 700 650 r9-16 = 0.851 * 600 550 500 450 450 400 12 12.5 13 13.5 14 14.5 15 15.5 16 16.5 Repeated-Sprint Fatigue (%) 2 30 C 25 20 r9-16 = 0.924 * 600 500 30 r1-8 = 0.862 * 650 550 400 0 B 800 2.5 3 3.5 4 4.5 D 25 r9-16 = -0.317 15 r9-16 = -0.721 20 15 10 10 r1-8 = -0.258 5 5 0 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 0 0 Peak Power (W/kg) r1-8 = -0.701 2.0 0 2.5 3.0 3.5 4.0 4.5 Critical Power (W/kg) Figure 2.5. Relationships of peak power and critical power with repeated-sprint ability and repeated-sprint fatigue. The relationships between repeated-sprint ability (J/kg) and peak power (W/kg; panel A), repeated-sprint ability and critical power (W/kg; panel B), repeated-sprint fatigue (% decrement) and peak power (panel C), and repeatedsprint fatigue and critical power (panel D) are illustrated. In each panel, the closed (●) and open circles ( ), and the solid and dotted lines of best fit respectively correspond to the first and second series of sprints. The correlations for the first (r1-8) and second (r916) series of sprints are displayed with indication (*) of statistically significant relationships (p ≤ .05). 32 Table 2.3. Multiple regression model summaries Model df SS MS F p R2 SEE 1-β 94.38 .0004 .979 14.0 .993 22.38 .007 .918 32.8 .917 3.51 .132 .637 2.7 .195 3.25 .145 .619 4.1 .232 RSA1-8 Regression 2 37060.1 18530.0 Residual 4 785.3 Total 6 37845.4 Regression 2 48122.5 24061.2 Residual 4 4300.0 Total 6 52422.4 Regression 2 50.3 25.1 Residual 4 28.6 7.2 Total 6 78.9 Regression 2 111.4 55.7 Residual 4 68.6 17.1 Total 6 180.0 196.3 RSA9-16 1075.0 RSF1-8 RSF9-16 Note. In each analysis, peak power and critical power were used to predict repeated-sprint ability during the first (RSA1-8) and second series of eight sprints (RSA9-16), as well as repeated-sprint fatigue across each series of sprints (RSF1-8 and RSF9-16). The ANOVA and model summary are provided for each analysis: df = degrees of freedom; SS = sum of squares; MS = mean squares; F = F ratio; p = p-value; R2 = multiple coefficient of determination; SEE = standard error of the estimate; 1-β = statistical power. predictability (i.e., R2 ≥ .90; Knofczynski & Mundfrom, 2008) of RSA1-8 and RSA9-16, and the global validation of linear model assumptions were met for each model (Peña & Slate, 2006). The fact that the standardized regression coefficients (Table 2.4) indicated that PP was a more influential predictor of RSA1-8 while CP was a more influential predictor of RSA9-16 further suggests that PP and CP have complementary roles in repeated-sprint ability. Finally, PP was highly predictive of the initial individual sprint 33 Table 2.4. Multiple regression model coefficient summaries DV IV b beta t p PP 41.52 0.72 6.75 .003 CP 37.48 0.33 3.14 .035 Int. -17.41 -0.27 .800 PP 25.36 0.37 1.76 .153 CP 85.90 0.65 3.07 .037 Int. -12.66 -0.08 .937 PP 1.49 0.56 1.26 .275 CP -5.72 -1.12 -2.51 .066 Int. 9.65 0.79 .476 PP 1.86 0.47 1.02 .365 CP -8.23 -1.06 -2.33 .080 Int 15.98 0.84 .448 RSA1-8 RSA9-16 RSF1-8 RSF9-16 Note. The standardized regression coefficients (beta) for peak power (PP) and critical power (CP) are bold in the sections for repeated sprint ability during the first (RSA1-8) and second series of sprints (RSA9-16) to emphasize how their relative predictive influence switched from from the first to second series of repeated sprints. DV = dependent variable; IV = Independent variables; b = the unstandardized regression coefficients; beta = the standardized regression coefficients; t = the t distribution; p = p-value; = repeated-sprint ability during the second series of eight sprints; RSF1-8 = repeated-sprint fatigue across the first series of sprints; RSF9-16 = repeated-sprint fatigue across the second series of sprints; PP = peak power; CP = critical power, Int. = the y-intercept of each model. 34 performances but became less predictive as more sprints were performed. In contrast, CP became more highly predictive as the trials progressed (Figure 2.6). Discussion The main finding of this study was that peak power (an estimate of muscle power) and critical power (an estimate of muscle oxidative capacity) accounted for over 90% of the variance in repeated-sprint ability (Table 2.3) among highly trained competitive athletes from various sports. More importantly, PP was more highly correlated with “nonfatigued” sprint performance, whereas CP was more highly correlated with “fatigued” sprint performance (Table 2.4; Figure 2.6). The results of this study support previous investigations reporting strong relationships between repeated-sprint ability and both single-sprint performance and aerobic fitness qualities—for brevity, only cycling studies among athletic populations will be discussed—but this appears to be the first investigation which has explored the relationship between critical power and repeatedsprint ability in isolation or in combination with other factors. Overall, my findings 1.000 0.991 * 0.981 * 0.978 * 0.937 * 0.908 * 0.900 0.880 * 0.911 * 0.945 * 0.906 * 0.935 * 0.880 * 0.889 * 0.913 * 0.937 * 0.925 * 0.915 * A Correlation (r) * * 0.865 * 0.869 0.880 0.858 * 0.800 0.838 0.854 *0.853 * 0.819 * 0.843 * 0.786 0.754 * 0.729 0.700 0.600 * * 0.778 * 0.808 * 0.823 * B C 0.651 0.000 0.500 S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 Figure 2.6. The relationship of peak power and critical power with individual sprint performance. Pearson’s correlation coefficients (r) are illustrated for the relationship between sprint-by-sprint (S1–S16) performance (J/kg) and peak power (W/kg; ●), and for the relationship between sprint-by-sprint performance and critical power (W/kg; ). Statistically significant relationships (p ≤ .05) are indicated (*). 35 suggest complementary interplay between peak power and critical power during repeated sprints. That is, the ability of the muscle to both produce peak power and facilitate creatine phosphorylation form a continuum for repeated-sprint ability. Peak power is the rested muscle’s ability to rapidly consume ATP, whereas critical power represents the exercising muscle’s ability to resynthesize PCr. Thus, maximizing these two factors can enhance repeated-sprint ability, and critical power may have even greater relevance to sports involving multiple series of repeated sprints during competition. These findings could have important implications for sport scientists interested in helping athletes achieve greater repeated-sprint ability. Both PP and CP were highly correlated with repeated-sprint ability (Figure 2.5). Likewise, when used in multiple regression, PP and CP accounted for ~98% and ~92% of the variance in RSA1-8 and RSA9-16, respectively (Table 2.3). Without measuring muscle [PCr], it was assumed that the demanding repeated-sprint protocol would cause large decreases in [PCr] (Dawson et al., 1997; Gaitanos et al., 1993), and that repeated-sprint ability would be affected by energy supply limitations (see Girard et al., 2011, for a review). Power recovery after cycling sprints heavily depends on the degree of PCr resynthesis (Bogdanis et al., 1995) during recovery. In other words, a greater rate of PCr resynthesis, which is primarily an oxidative process (Gabr et al., 2011; Kemp et al., 2015; Mahler, 1985; Yoshida & Watari, 1997), would support greater subsequent sprint performance. In this study, a linear model (Jones & Vanhatalo, 2017) was used to estimate the parameters that describe the hyperbolic relationship between power and duration (Jones et al., 2010; Jones & Vanhatalo, 2017; Poole et al., 2016): critical power and W’ (Figure 36 2.2). Again, critical power is the greatest work rate that can be achieved using wholly oxidative energy provision, and W’ is the finite amount of work that can be performed above the CP using primarily PCr and glycogen (Jones et al., 2010). Morton and Billat (2004) modeled that greater high-intensity intermittent exercise tolerance (i.e., time to exhaustion) is possible with a greater magnitude of either critical power or W’. However, a careful examination of their model also indicates that when recovery interval duration and intensity are held constant, an equal relative increase (e.g., 10%) in critical power compared to W’ would yield a much larger increase in intermittent exercise tolerance. For example, one participant in this study had a PP of 16.1 W/kg and a CP of 4.3 W/kg. According to their model (Morton & Billat, 2004), intermittent exercise tolerance for that participant would increase by ~5.5% with a 10% increase in critical power, but exercise tolerance would only increase by ~1% with a 10% increase in W’. To extend upon the findings of Morton and Billat (2004), Skiba and colleagues (2012) modeled that W’ recharge during high-intensity intermittent cycling occurs monoexponentially with a time-constant that depends on the difference between critical power and the recovery exercise intensity. In the current study, passive recovery was used with a relatively constant recovery duration (~22 s). Therefore, critical power was the primary factor contributing to the theoretical restoration of W’, and in turn, PCr resynthesis, because both W’ discharge and reconstitution corresponds with both PCr depletion (Jones et al., 2010) and restoration (Skiba, Fulford, Clarke, Vanhatalo, & Jones, 2015). As such, while this is the first study to predict repeated-sprint ability from peak power and critical power, my results yield a practical demonstration of what others have elegantly modeled, while supporting the hypothesis that sprint-by-sprint performance is 37 closely related to availability of PCr. Many have attributed the observed reductions in sprint-by-sprint performance during repeated-sprint exercise to the rate of PCr resynthesis between sprints. Repeatedsprint fatigue is a moderator of repeated-sprint ability, and the current study used the sprint decrement score to describe RSF1-8 and RSF9-16. This study revealed inadequate reliability for both RSF1-8 and RSF9-16 (Table 2.2), which is consistent with previous research that has subsequently questioned the usefulness of describing repeated-sprint fatigue with a relative measure (Oliver, 2009). Regardless, the sprint decrement score was chosen for this study because it is considered the most valid way to describe repeated-sprint fatigue (Girard et al., 2011; Glaister et al., 2004). Aside from questionable reliability, the surprisingly small relationship between PP and repeated-sprint fatigue (Figure 2.5) is the primary reason why PP and CP yielded inadequate predictability (i.e., R2 ≤ .90; Knofczynski & Mundfrom, 2008) of RSF1-8 or RSF9-16 (Table 2.3). My results oppose several others who have reported very high relationships (r = .70 .85) between peak cycling power and measures of repeated-sprint fatigue among athletic populations (Bishop et al., 2001; Bishop et al., 2003; Dawson et al., 1993; Mendez-Villanueva et al., 2008). The difference between my results and those from others may be due to the fact that other underpinning factors can affect repeated-sprint fatigue (Girard et al., 2011). For example, muscle buffer capacity is a unique trainable quality (see Bishop, Girard, & Mendez-Villanueva, 2011, for a review) that may relate to repeated-sprint fatigue independent of single-sprint performance (Bishop et al., 2003; Bishop & Edge, 2006; Bishop & Spencer, 2004). Muscle buffer capacity varies with different training statuses 38 and participation in different types of sports (Edge et al., 2006; Sahlin & Henriksson, 1984). Thus, it is possible that the diverse training and competitive backgrounds of the motley crew of athletes who participated in this study may have led to other adaptations, like muscle buffer capacity, that could have caused the small observed relationship between PP and repeated-sprint fatigue. PP did not correlate with repeated-sprint fatigue in this study, but there were strong negative relationships between CP and both RSF1-8 (r = -.701; p = .079) and RSF916 (r = -.721; p = .068; Figure 2.5), albeit those relationships were statistically nonsignificant due to small sample size. The idea that aerobic variables might relate to repeated-sprint fatigue is not a new one. The most commonly investigated factor has been the maximal aerobic capacity, with which some have reported moderate relationships with repeated-sprint fatigue (Bishop & Edge, 2006; Dawson et al., 1993), while others have not (Bishop et al., 2003; Edge et al., 2006; Mendez-Villanueva et al., 2008). Inconsistencies may be due to the fact that maximal aerobic capacity is a product of both oxygen transport and utilization. Historical debate has considered whether oxygen delivery to the muscle cell affects PCr resynthesis to a greater extent than the ability of the muscle cell to utilize oxygen (McMahon & Jenkins, 2002). While conjecture remains, several others have reported smaller relationships between maximal aerobic capacity and repeated-sprint fatigue (-.56 ≤ r ≤ -.17; Bishop et al., 2003; Bishop & Edge, 2006; Dawson et al., 1993; Edge et al., 2006; Mendez-Villanueva et al., 2008) than the range of correlations observed between repeated-sprint fatigue and direct markers of muscle oxidative capacity measured from muscle biopsies (-.74 ≤ r ≤ -.46; Thomas et al., 2004). Likewise, my study revealed a very high relationship between CP and repeated-sprint 39 fatigue. Critical power is an indirect indicator of muscle oxidative capacity because it can be sustained without progressive increases of blood lactate or breakdown of muscle PCr (Jones et al., 2008; Jones et al., 2010; Poole et al., 2016). Therefore, my findings are consistent with previous research suggesting that greater muscle oxidative capacity can mitigate repeated-sprint fatigue. The novelty in the current findings are amplified by the fact that the strength of the relationships between PP or CP and sprint performance switched from the first to the second series of sprints. This was subtly true for the Pearson’s correlation coefficients (Figure 2.5). More importantly, the standardized coefficients (Table 2.4) indicated that a one standard deviation increase in PP or CP would elicit an increase in RSA1-8 by 0.718 or 0.334 standard deviations, respectively. Conversely, during the second series of sprints, the standardized coefficients indicated that RSA9-16 would increase by 0.372 or 0.650 standard deviations with a one standard deviation increase in either PP or CP (Field, 2013). In other words, the relative importance of PP was ~2.1 times the relative importance of CP in the model predicting RSA1-8. In contrast, the relative importance of CP was ~1.7 times that for PP in the regression model for RSA9-16 (Table 2.4). For reasons described above, greater critical power would likely facilitate greater PCr resynthesis during the 3-min recovery interval between the first and second series of sprints. Typically, 3 min is not long enough to fully restore PCr after maximal cycling exercise (Baker et al., 2010; Haseler et al., 1999). Given that muscle [PCr] was not measured in this study, I assumed that the participants initiated the second series of sprints with some degree of reduced muscle [PCr]. In a PCr depleted state, performance during later sprints would also depend on greater oxidative ATP supply during sprints 40 (Girard et al., 2011; McGawley & Bishop, 2015). Indeed, others have proposed that oxidative ATP may supply up to 40% of the energy required during the final sprint performed in a series (Girard et al., 2011). As the greatest metabolic rate supported by wholly oxidative metabolism (Jones et al., 2010; Jones & Vanhatalo, 2017; Poole et al., 2016), a greater critical power would logically faciliate greater mechanical output during repeated sprints through greater aerobic energy provision. For example, two individuals in this study had similar PP (~1.5% different) but different CP (by ~20%). Interestingly, RSA1-8 was nearly identical (~2% different) for both participants, but RSA9-16 was ~12.5% greater for the individual with greater CP, who may have been able to achieve greater total mechanical work from greater oxidative energy provision. That is just a practical example between two participants, but across all participants, PP highly correlated (i.e., r ≥ .90) to each of the first five sprints during the first series, and the relationship between CP and sprint performance gradually increased until there were very large correlations with each of the final four sprints during the second series of sprints (Figure 2.6). Overall, it appears that greater critical power facilitates greater performance during sprints that would depend on greater PCr resynthesis between sprints and/or greater aerobic ATP provision during sprints. In conclusion, my findings suggest that peak power and critical power complement each other in a way that forms a continuum for repeated-sprint ability. Examination of the regression coefficients for each model (Table 2.4) suggest that those with lower peak power and critical power would have inferior repeated-sprint ability. Moderate repeated-sprint ability can be expressed by those with average peak power and critical power, or those with a greater magnitude of one characteristic and less of the 41 other. Superior repeated-sprint ability would be expressed by those with greater peak power and critical power. To illustrate (Figure 2.7), I simulated this continuum by estimating repeated-sprint ability from 100 randomly generated pairs of PP and CP and applied the derived regression coefficients (Table 2.4). Indeed, future research with a larger sample is needed to confirm the presence of this theoretical continuum. Practical Applications The current findings indicate a simple message, that peak power and critical power complement repeated-sprint ability. Importantly, episodes of repeated-sprint 850 A C B Repeated-Sprint Ability 800 750 700 650 600 550 0 5 10 15Combined 20 25 Rank 30 35of Simulated 40 45 50Peak 55 Power & Critical 60 65 70 75 Power 80 85 90 95 100 Figure 2.7. Theoretical continuum for repeated-sprint ability due to complementary interplay between peak power and critical power. Observed ranges from seven highly trained male athletes were used to randomly assign 100 pairs of peak power and critical power. Each pair of randomly generated dependent variables were applied to the regression equation derived to estimate repeated-sprint ability from the first eight sprints (Table 2.4), and then the resulting repeated-sprint ability was graphed over the combined rank of the simulated peak power and critical power. This simulation presents a theoretical continuum for repeated-sprint ability, by which those with exceptional peak power and critical power (A) would have superior repeated-sprint ability. Those with average critical power and peak power, or those with greater performance in one attribute and poorer performance in the other would have moderate repeated-sprint ability (B). Finally, those with poor peak power and critical power would have inferior repeatedsprint ability (C). 42 activity can happen several times during competition (Table 2.1), especially during critical moments of play (Girard et al., 2011). If recovery intervals between episodes of repeated sprints are inadequate to fully restore [PCr] (Baker et al., 2010; Haseler et al., 1999), performance during subsequent episodes of repeated sprints will likely be impaired. This was true in the current study as RSA9-16 was only ~89% of RSA1-8, and participants with lesser critical power had greater repeated-sprint fatigue. Repeated-sprint fatigue may affect performance during intense competition, especially toward the end of highly competitive contests (Spencer et al., 2005). In this study, there was a complementary interplay between peak power and critical power which formed a continuum for repeated-sprint ability. While more research with a larger sample size is needed to further validate this continuum, if it does exist, then three training program varieties may sufficiently cater to the individual needs of athletes seeking greater repeated-sprint ability. Specifically, a training program that primarily focuses on improving single-sprint performance (i.e., peak power) would favor athletes who have greater critical power and limited single-sprint performance, or those who have exceptional peak power and critical power and play a sport that favors explosive actions over pace of play. A concurrent training program that simultaneously improves both peak power and critical power would benefit those with average ranges of peak power or critical power. Finally, a training program designed to improve critical power would benefit athletes with exceptional peak power and critical power if their sport favors pace of play over explosiveness, as well as athletes with greater peak power and limited critical power. More research is warranted to further define the intricacies and efficacy of 43 those training variations. Nevertheless, it is clear from the findings in this study that athletes who desire greater repeated-sprint ability can benefit from greater levels of both types of power. References Atkinson, G., & Nevill, A. M. (1998). Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Medicine, 26(4), 217-238. Baker, J. S., McCormick, M. C., & Robergs, R. A. (2010). Interaction among skeletal muscle metabolic energy systems during intense exercise. Journal of Nutrition and Metabolism, 2010, 1-13. Bishop, D., & Edge, J. (2006). Determinants of repeated-sprint ability in females matched for single-sprint performance. European Journal of Applied Physiology, 97(4), 373-379. Bishop, D., Girard, O., & Mendez-Villanueva, A. (2011). Repeated-sprint ability—Part II. Sports Medicine, 41(9), 741-756. Bishop, D., Lawrence, S., & Spencer, M. (2003). Predictors of repeated-sprint ability in elite female hockey players. Journal of Science and Medicine in Sport, 6(2), 199209. Bishop, D., & Spencer, M. (2004). Determinants of repeated-sprint ability in well-trained team-sport athletes and endurance-trained athletes. Journal of Sports Medicine and Physical Fitness, 44(1), 1-7. Bishop, D., Spencer, M., Duffield, R., & Lawrence, S. (2001). The validity of a repeated sprint ability test. Journal of Science and Medicine in Sport, 4(1), 19-29. Bogdanis, G. C., Nevill, M. E., Boobis, L. H., Lakomy, H. K., & Nevill, A. M. (1995). Recovery of power output and muscle metabolites following 30 s of maximal sprint cycling in man. The Journal of Physiology, 482(2), 467-480. Carling, C., Le Gall, F., & Dupont, G. (2012). Analysis of repeated high-intensity running performance in professional soccer. Journal of Sports Sciences, 30(4), 325-336. Chtourou, H., & Souissi, N. (2012). The effect of training at a specific time of day: A review. Journal of Strength & Conditioning Research, 26(7), 1984-2005. Dawson, B., Fitzsimons, M., & Ward, D. (1993). The relationship of repeated sprint ability to aerobic power and performance measures of anaerobic work capacity 44 and power. Australian Journal of Science and Medicine in Sport, 25, 88-93. Dawson, B., Goodman, C., Lawrence, S., Preen, D., Poglaze, T., Fitzsimons, M., & Fournier, P. (1997). Muscle phosphocreatine repletion following single and repeated short sprint efforts. Scandinavian Journal of Medicine & Science in Sports, 7(4), 206-213. Deutsch, M. U., Kearney, G. A., & Rehrer, N. J. (2007). Time-motion analysis of professional rugby union players during match-play. Journal of Sports Sciences, 25(4), 461-472. Edge, J., Bishop, D., Hill-Haas, S., Dawson, B., & Goodman, C. (2006). Comparison of muscle buffer capacity and repeated-sprint ability of untrained, endurance-trained and team-sport athletes. European Journal of Applied Physiology, 96, 225-234. Elmer, S. J., Marshall, C. S., Amann, M., & Martin, J. C. (2013). Neuromuscular fatigue, reserve, and recovery following exhaustive high-intensity endurance exercise. Medicine and Science in Sports and Exercise, 45(5), 182-183. Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G* Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39(2), 175-191. Field, A. (2013). Discovering statistics using IBM SPSS statistics. London, England: Sage. Fitzsimons, M., Dawson, B., Ware, D., & Wilkinson, A. (1993). Cycling and running tests of repeated sprint ability. Australian Journal of Science and Medicine in Sport, 25(4), 82-87. Gabr, R. E., El-Sharkawy, A. M. M., Schär, M., Weiss, R. G., & Bottomley, P. A. (2011). High-energy phosphate transfer in human muscle: Diffusion of phosphocreatine. American Journal of Physiology-Cell Physiology, 301(1), C234C241. Gaitanos, G. C., Williams, C., Boobis, L. H., & Brooks, S. (1993). Human muscle metabolism during intermittent maximal exercise. Journal of Applied Physiology, 75(2), 712-719. Gardner, A. S., Stephens, S., Martin, D. T., Lawton, E., Lee, H., & Jenkins, D. (2004). Accuracy of SRM and power tap power monitoring systems for bicycling. Medicine and Science in Sports and Exercise, 36(7), 1252-1258. Girard, O., Mendez-Villanueva, A., & Bishop, D. (2011). Repeated-sprint ability—Part I. Sports Medicine, 41(8), 673-694. Glaister, M., Stone, M. H., Stewart, A. M., Hughes, M., & Moir, G. L. (2004). The reliability and validity of fatigue measures during short-duration maximal- 45 intensity intermittent cycling. Journal of Strength & Conditioning Research, 18(3), 459-462. Gray, M. A. J., & Jenkins, D. G. (2010). Match analysis and the physiological demands of Australian football. Sports Medicine, 40(4), 347-360. Haseler, L. J., Hogan, M. C., & Richardson, R. S. (1999). Skeletal muscle phosphocreatine recovery in exercise-trained humans is dependent on O2 availability. Journal of Applied Physiology, 86(6), 2013-2018. Hopkins, W. G. (2000). Measures of reliability in sports medicine and science. Sports Medicine, 30(1), 1-15. Iosia, M. F., & Bishop, P. A. (2008). Analysis of exercise-to-rest ratios during division IA televised football competition. Journal of Strength & Conditioning Research, 22(2), 332-340. Jones, A. M., & Vanhatalo, A. (2017). The ‘Critical Power’ concept: Applications to sports performance with a focus on intermittent high-intensity exercise. Sports Medicine, 47(1), 65-78. Jones, A. M., Vanhatalo, A., Burnley, M., Morton, R. H., & Poole, D. C. (2010). Critical power: Implications for determination of VO2max and exercise tolerance. Medicine and Science in Sports and Exercise, 42(10), 1876-90. Jones, A. M., Wilkerson, D. P., DiMenna, F., Fulford, J., & Poole, D. C. (2008). Muscle metabolic responses to exercise above and below the “critical power” assessed using 31P-MRS. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 294(2), R585-R593. Kemp, G. J., Ahmad, R. E., Nicolay, K., & Prompers, J. J. (2015). Quantification of skeletal muscle mitochondrial function by 31P magnetic resonance spectroscopy techniques: A quantitative review. Acta Physiologica, 213(1), 107-144. Knofczynski, G. T., & Mundfrom, D. (2008). Sample sizes when using multiple linear regression for prediction. Educational and Psychological Measurement, 68(3), 431-442. Kovacs, M. S. (2006). Applied physiology of tennis performance. British Journal of Sports Medicine, 40(5), 381-386. Mahler, M. (1985). First-order kinetics of muscle oxygen consumption, and an equivalent proportionality between QO2 and phosphorylcreatine level. Implications for the control of respiration. The Journal of General Physiology, 86(1), 135-165. Martin, J. C., & Brown, N. A. (2009). Joint-specific power production and fatigue during maximal cycling. Journal of Biomechanics, 42(4), 474-479. 46 Martin, J. C., Diedrich, D., & Coyle, E. F. (2000). Time course of learning to produce maximum cycling power. International Journal of Sports Medicine, 21(7), 485487. Martin, J. C., Milliken, D. L., Cobb, J. E., McFadden, K. L., & Coggan, A. R. (1998). Validation of a mathematical model for road cycling power. Journal of Applied Biomechanics, 14, 276-291. Matthew, D., & Delextrat, A. (2009). Heart rate, blood lactate concentration, and timemotion analysis of female basketball players during competition. Journal of Sports Sciences, 27(8), 813-821. McCully, K. K., Vandenborne, K., DeMeirleir, K., Posner, J. D., & Leigh Jr., J. S. (1992). Muscle metabolism in track athletes, using 31P magnetic resonance spectroscopy. Canadian Journal of Physiology and Pharmacology, 70(10), 13531359. McGawley, K., & Bishop, D. J. (2015). Oxygen uptake during repeated-sprint exercise. Journal of Science and Medicine in Sport, 18(2), 214-218. McMahon, S., & Jenkins, D. (2002). Factors affecting the rate of phosphocreatine resynthesis following intense exercise. Sports Medicine, 32(12), 761-784. Mendez-Villanueva, A., Hamer, P., & Bishop, D. (2008). Fatigue in repeated-sprint exercise is related to muscle power factors and reduced neuromuscular activity. European Journal of Applied Physiology, 103(4), 411-419. Morton, R. H., & Billat, L. V. (2004). The critical power model for intermittent exercise. European Journal of Applied Physiology, 91(2-3), 303-307. Oliver, J. L. (2009). Is a fatigue index a worthwhile measure of repeated sprint ability? Journal of Science and Medicine in Sport, 12(1), 20-23. Peña, E. A., & Slate, E. H. (2006). Global validation of linear model assumptions. Journal of the American Statistical Association, 101(473), 341-354. Polley, C. S., Cormack, S. J., Gabbett, T. J., & Polglaze, T. (2015). Activity profile of high-level Australian lacrosse players. Journal of Strength & Conditioning Research, 29(1), 126-136. Poole, D. C., Burnley, M., Vanhatalo, A., Rossiter, H. B., & Jones, A. M. (2016). Critical power: An important fatigue threshold in exercise physiology. Medicine and Science in Sports and Exercise, 48(11), 2320-2334. Prompers, J. J., Wessels, B., Kemp, G. J., & Nicolay, K. (2014). Mitochondria: Investigation of in vivo muscle mitochondrial function by 31 P magnetic resonance spectroscopy. The International Journal of Biochemistry & Cell Biology, 50, 67-72. 47 Quod, M. J., Martin, D. T., Martin, J. C., & Laursen, P. B. (2010). The power profile predicts road cycling MMP. International Journal of Sports Medicine, 31(6), 397401. Sahlin, K., & Henriksson, J. (1984). Buffer capacity and lactate accumulation in skeletal muscle of trained and untrained men. Acta Physiologica, 122(3), 331-339. Skiba, P. F., Chidnok, W., Vanhatalo, A., & Jones, A. M. (2012). Modeling the expenditure and reconstitution of work capacity above critical power. Medicine and Science in Sports and Exercise, 44(8), 1526-1532. Skiba, P. F., Fulford, J., Clarke, D. C., Vanhatalo, A., & Jones, A. M. (2015). Intramuscular determinants of the ability to recover work capacity above critical power. European Journal of Applied Physiology, 115(4), 703-713. Spencer, M., Bishop, D., Dawson, B., & Goodman, C. (2005). Physiological and metabolic responses of repeated-sprint activities. Sports Medicine, 35(12), 10251044. Spencer, M., Lawrence, S., Rechichi, C., Bishop, D., Dawson, B., & Goodman, C. (2004). Time-motion analysis of elite field hockey, with special reference to repeated-sprint activity. Journal of Sports Sciences, 22(9), 843-850. Thomas, C., Sirvent, P., Perrey, S., Raynaud, E., & Mercier, J. (2004). Relationships between maximal muscle oxidative capacity and blood lactate removal after supramaximal exercise and fatigue indexes in humans. Journal of Applied Physiology, 97(6), 2132-2138. Wootton, S., & Williams, C. (1983). The influence of recovery duration on repeated maximal sprints. In H. G. Knuttgen, J. A. Vogel, & J. Poortmans (Eds.), Biochemistry of exercise, international series in sports science (Vol. 8, pp. 269273). Champaign, Illinois: Human Kinetics. Yoshida, T., & Watari, H. (1997). Effect of circulatory occlusion on human muscle metabolism during exercise and recovery. European Journal of Applied Physiology and Occupational Physiology, 75(3), 200-205 CHAPTER 3 SINGLE-LEG CYCLING TRAINING IMPROVES REPEATED-SPRINT AND ENDURANCE RUNNING PERFORMANCE Introduction During team-sport competition in a variety of sports, total distance and high-speed running activity tends to decrease from earlier to later portions of a contest (Datson et al., 2014; Jennings, Cormack, Coutts, & Aughey, 2012; Mohr, Krustrup, & Bangsbo, 2003; Polley, Cormack, Gabbett, & Polglaze, 2015; Stojanović et al., 2017). Such fatigue could affect outcomes if players cannot meet the greater physical demands observed during highly competitive play or when facing stronger opponents (Kempton & Coutts, 2016; Trewin, Meylan, Varley, & Cronin, 2017; Vinson, Gerrett, & James, 2018). Highintensity running activity and the frequency of athletic actions during competition are generally associated with greater high-intensity running capacity (Datson et al., 2014; Helgerud, Engen, Wisløff, & Hoff, 2001; Johnston, Gabbett, Jenkins, & Hulin, 2015; Kempton & Coutts, 2016; Narazaki, Berg, Stergiou, & Chen, 2009; Rampinini et al., 2007). While exceptions apply (Bradley et al., 2013), there is a precedent for athletes to achieve greater cardiorespiratory fitness because elite team-sport competitors generally possess greater aerobic or high-intensity running capacity than sub-elite competitors from the same sports (Datson et al., 2014; Gabbett, Kelly, Ralph, & Driscoll, 2009; Jennings et al., 2012; Mohr et al., 2003). Greater aerobic running capacity is also associated with 49 reduced incidence of injury (Malone et al., 2018) during participation and decreased markers of fatigue before (i.e., from preceding practices; Hunkin, Fahrner, & Gastin, 2014) and after (Johnston et al., 2015) team-sport competition. Consequently, aerobic training has been recommended for team-sport athletes (Stone & Kilding, 2009). High-intensity interval training has become a standard for improving aerobic performance among team-sport athletes (Stone & Kilding, 2009). Various high-intensity interval training protocols (Buchheit & Laursen, 2013) are capable of improving physical qualities that may benefit performance during team-sport competition including running endurance (Cicioni-Kolsky, Lorenzen, Williams, & Kemp, 2013; Stone & Kilding, 2009) and repeated-sprint ability (Bishop, Girard, & Mendez-Villanueva, 2011; Cicioni-Kolsky et al., 2013). One potential limitation involved with high-intensity interval training is that running can increase overall musculoskeletal strain (Buchheit & Laursen, 2013), particularly when it coincides with practice, competition, and other activities common to the team-sport training process (Stone & Kilding, 2009). For example, common athletic actions like acceleration (Bezodis, North, & Razavet, 2017) and cutting (Spiteri, Cochrane, Hart, Haff, & Nimphius, 2013), and typical high-intensity interval running speeds (~10–21 km/h) cause peak ground reaction forces that are ~2–3 times bodyweight (Clark & Weyand, 2014; Keller et al., 1996). Team-sport players already experience greater muscle damage after competitions that demanded greater high-speed running (~10–25 km/h) volume (Young, Hepner, & Robbins, 2012). Therefore, alternative modes of high-intensity interval training exercise are warranted when greater fitness is desired, but increased running volume could exacerbate fatigue (Young et al., 2012) or risk of overuse injury (Malone et al., 2018). 50 Cycling is one possible alternative that could enhance high-intensity running performance because it is a nonweight-bearing task involving coordinated actions of the ankle, knee, and hip joints. Intense cycling (~150–200 W per leg) elicits peak pedal reaction forces (~550–700 N; Hoes, Binkhorst, Smeekes-Kuyl, & Vissers, 1968) that would equate to ~0.65–1.35 times bodyweight depending on an individual’s bodyweight. Despite historical debate over the potential cross-training effects of cycling training to improve running performance (Tanaka, 1994), recent studies have reported greater highintensity running performances after short-term high-intensity cycling interval training (Chan, Ho, & Yung, 2018; Etxebarria, Anson, Pyne, & Ferguson, 2014). Even greater improvements in running endurance could occur after single-leg cycling training because it elicits physiological adaptations beyond those from double-leg cycling (Abbiss et al., 2011). Single-leg cycling stimulates greater blood flow to the exercising muscle compared to multi-limb exercise involving larger amounts of muscles mass (e.g., running or double-leg cycling; Burns, Pollock, LaScola, & McDaniel, 2014; Klausen, Secher, Clausen, Hartling, & Trap-Jensen, 1982). Access to a greater proportion of the aerobic capacity during single-leg cycling (Abbiss et al., 2011; Bundle, Ernst, Bellizzi, Wright, & Weyand, 2006; Klausen et al., 1982; MacInnis, Morris, et al., 2017; MacInnis, Zacharewicz, et al., 2017; Rud, Foss, Krustrup, Secher, & Hallen, 2012; Saltin et al., 1976) facilitates greater exercise intensity per leg (Abbiss et al., 2011; Bundle et al., 2006; MacInnis, Morris, et al., 2017). Consequently, single-leg cycling training has repeatedly facilitated significant peripheral adaptations and greater cycling exercise performances (Abbiss et al., 2011; Dela, Handberg, Mikines, Vinten, & Galbo, 1993; Henriksson, 1977; Klausen et al., 1982; MacInnis, Zacharewicz, et al., 2017; Rud et al., 51 2012; Saltin et al., 1976; Vincent et al., 2015). A variety of single-leg cycling training protocols have been investigated, ranging from low- to moderate-intensity continuous training (Dela et al., 1993; Henriksson, 1977; Klausen et al., 1982; MacInnis, Zacharewicz, et al., 2017; Rud et al., 2012; Saltin et al., 1976) to high-intensity interval training (Abbiss et al., 2011; MacInnis, Zacharewicz, et al., 2017; Saltin et al., 1976; Vincent et al., 2015). Previous authors have reported that six high-intensity single-leg cycling interval training sessions performed over the course of 2–3 weeks were superior to moderate-intensity continuous training (MacInnis, Zacharewicz, et al., 2017) and double-leg cycling training (Abbiss et al., 2011). Taken together, previous research suggests that high-intensity single-leg cycling interval training can improve muscle respiratory capacity, even among highly trained individuals (Abbiss et al., 2011). As such, single-leg cycling may be an efficacious cross-training modality for ground-based team-sport athletes, but this has never been studied. The purpose of this study was to determine if short-term, high-intensity single-leg cycling training could improve field-measures of high-intensity running performance. Methods Overview The University of Utah Institutional Review Board approved the methods in this study. To determine the effects of high-intensity single-leg cycling training on repeatedsprint and endurance running performance, collegiate club lacrosse players rehearsed the repeated-sprint and endurance running protocols during an 8-week strength and conditioning and familiarization period coinciding with the final phase of their off-season training. Thereafter, the 5-week study commenced in conjunction with the team’s 52 preseason schedule. During the 1st week of the study, they performed the assessments and were assigned to either a control group (CON), a single-leg cycling training group (SLC), or a high-intensity interval running training group (RUN). All three groups participated as a team in their regularly scheduled resistance training and sport practice activities for the next 3 weeks. During that period, CON performed no additional activity, while SLC and RUN performed six high-intensity interval training sessions (i.e., two sessions per week). Repeated-sprint ability and endurance running performance was retested during the 5th week. Linear mixed models were used to compare effects between CON and the training groups, and independent samples t tests were used to compare effects between SLC and RUN. Participants After providing written, informed consent, 27 experienced collegiate club lacrosse players (Table 3.1) with recreational training background (Rhea, 2004) volunteered to participate in this study. At the start of the study, all participants were free of injury and were able to perform all study activities without limitation. Table 3.1. Participant characteristics Group Age (years) Height (cm) Body Mass (kg) Lacrosse (years) Training (years) CON (n = 9) 20.6 (2.8) 179 (6) 81.1 (9.8) 8.9 (3.2) 4.2 (2.0) SLC (n = 9) 19.8 (1.4) 183 (5) 82.2 (8.4) 8.0 (2.6) 4.0 (1.7) RUN (n = 9) 20.5 (1.9) 182 (6) 78.1 (7.5) 8.6 (2.9) 3.8 (2.2) All Participants (N = 27) 20.3 (1.8) 181 (6) 80.5 (8.5) 8.5 (2.8) 4.0 (1.9) Note: All values are mean (SD). Group designations refer to the control (CON), singleleg cycling training (SLC), and high-intensity running interval training (RUN) groups. Lacrosse (years) and Training (years) indicates total self-reported years of participation in organized lacrosse and strength and conditioning, respectively. 53 Study procedures Repeated-sprint ability. Before and after the intervention period, repeated-sprint ability was assessed on a hard rubberized sports floor at the same time of the day to minimize the influence of circadian variations on performance (Chtourou & Souissi, 2012). Testing involved ten 40 m shuttle-sprints (10x40) performed similarly to that proposed by others (Rampinini et al., 2007). Briefly, participants jogged for 5 min before performing a series of dynamic stretches, which progressed up to an untimed 40-m shuttle-sprint. After a 3-min passive recovery, participants performed a timed 40-m shuttle-sprint that served as a criterion sprint to avoid pacing during the 10x40. After an additional 3-min recovery, participants began the 10x40. For each sprint, a custom sound file played over a loudspeaker gave the participants a 5-s warning (i.e., the toll of a bell) to initiate a staggered stance, with the front toe placed at a line marked 0.5 m behind a second line where a pair of electronic timing gates (Brower Timing Systems, Salt Lake City, UT) were set 1 m above the floor. After 5 s elapsed, a second sound (i.e., the blast of a cannon) indicated when to start sprinting. They maximally accelerated through the timing gates toward a line on the floor 20 m away, where they changed directions, and then would bring it on home by maximally accelerating through the timing gates again. To eliminate any effect from pacing, all participants were required to perform the first sprint of the 10x40 within 2.5% (~0.18 s) of the time it took to complete the criterion sprint (Rampinini et al., 2007), and they met this requirement during testing before (mean: -0.09 s; SD: 0.14 s) and after (mean: -0.04 s; SD: 0.19 s) the intervention. The same series of sounds were played every 30 s until the 10x40 was completed. Depending on ability and fatigue, each sprint lasted ~7–9 s, resulting in ~21–23 s of passive rest 54 between sprints. Three 10x40 courses were set up approximately 10 m apart to encourage greater motivation by allowing participants to compete with one another. From the 10x40, single-sprint performance, repeated-sprint ability, and repeatedsprint fatigue were recorded for each participant and used for analysis. Specifically, single-sprint performance was the fastest sprint (s) during the 10x40. Repeated-sprint ability was the sum of all 10 sprint times (Cicioni-Kolsky et al., 2013; Girard, MendezVillanueva, & Bishop, 2011). Repeated-sprint fatigue was expressed as a percentage using the sprint decrement score (Dawson et al., 1993), which is considered the most valid index of repeated-sprint fatigue (Girard et al., 2011; Glaister, Howatson, Pattison, & McInnes, 2008). Endurance running performance. Endurance running performance was assessed using the Multi-Stage Fitness Test (MSFT; Berthoin et al., 1994), which is a widely used field test that has reliably estimated maximal aerobic capacity (Ramsbottom, Brewer, & Williams, 1988) in a variety of populations (Mayorga-Vega, Aguilar-Soto, & Viciana, 2015). Briefly, participants ran back and forth between two lines marked 20 m apart, touching each line with the sound of a beep played over a loudspeaker. The first stage occurred at a speed of 8.5 km.h, and each subsequent stage increased by 0.5 km/h every minute until volitional exhaustion. Endurance running performance in the MSFT was indicated by the total distance covered (m; Jennings et al., 2012; Mayorga-Vega et al., 2015) and used for analysis. Group assignments. After performing the 10x40 and MSFT, participants were grouped into nine triads (n = 3 per triad) according to their average rank across singlesprint performance, repeated-sprint ability, and endurance running performance. A 55 random number generator was used to assign one participant in each triad to either CON, SLC, or RUN (n = 9 per group). This grouping technique insured no between-group differences in single-sprint performance, repeated-sprint ability, or endurance running performance at the start of the study. Control group. During the 3-week preseason intervention, all participants (N = 27) partook in their team’s required weekly activities, including four lacrosse practice sessions and two supervised resistance training sessions (Table 3.2). CON performed no additional training activities, but it should be noted that they fully participated (along with SLC and RUN) in any occasional conditioning activity directed by their coaches during team practice sessions. Single-leg cycling training. In addition to participating in their team’s required weekly training activities, SLC performed two high-intensity single-leg cycling interval training sessions per week (i.e., six total sessions; Table 3.2). During each session, two participants trained together using a two-ergometer and two-treadmill system, which Table 3.2. Weekly training schedule Time of Day Monday Tuesday Wednesday Thursday A.M. SLC SLC SLC SLC P.M. RT & RUN Practice & RUN RT Practice Friday Saturday Sunday Practice Off Practice Note. Technical skills practices occurred in the afternoon (P.M.) on Tuesdays, Thursdays, and Fridays, and in the morning (A.M.) on Saturdays. Supervised resistance training (RT) occurred on Mondays and Wednesdays. The control group performed no additional physical activities, but the high-intensity interval training group (RUN) ran after RT on Mondays, and after practice on Tuesdays. The single-leg cycling training group (SLC) sessions were scheduled individually from Monday–Thursday, and there was 24–72 hr of recovery between sessions, which varied based on individual schedules. 56 allowed them to alternate legs between training and recovery intervals by rotating from one single-leg cycle ergometer to a treadmill, and then to the opposite single-leg cycle ergometer, etc. Each 35–40 min session started with approximately 3–5 min of treadmill walking (~5.6 km/h), followed by a transition to ride on one of the friction-braked singleleg cycle ergometers (Monark Exercise AB, Vansbro, Sweden) placed directly behind each treadmill. The left crank of one ergometer, and the right crank of the other ergometer was equipped with a 10.2 kg counterweight (Abbiss et al., 2011; Bundle et al., 2006; Burns et al., 2014; Elmer, McDaniel, & Martin, 2016) which facilitates smooth single-leg cycling and prevents premature fatigue of the hip-flexor muscles during the final half of the pedal cycle (Elmer et al., 2016). A platform was placed between the two ergometers to support the foot of the inactive leg and protect the leg from the rotating counterweight. Three high-intensity intervals lasting 3 min were performed with each leg, in alternate fashion. Similar to others (Abbiss et al., 2011), participants were instructed to give maximal self-paced efforts and were continuously encouraged to achieve the highest average power throughout each 3 min interval. Continuous feedback for mean cycling power, pedal rate (RPM), and heart rate was provided with a power meter (Schoberer Rad Messtechnik, SRM, Jülich, Germany), which also recorded heart rate transmitted from monitors (Polar USA, New York, USA) worn during all training sessions. After each interval, participants quickly transitioned to a treadmill and walked (~5.6 km/h) for 2 min minus the time required to transition to the other ergometer (~15–20 s). It is important to note that we assumed walking in-between intervals would facilitate greater venous blood return of the previously exercised, vasodilated limb (Gordon, Abbiss, Ihsan, Maiorana, & 57 Peiffer, 2018) via the muscle pump (Laughlin, 1987). The session concluded with 5–10 minutes of walking on the treadmill at self-selected pace. The leg used for the first interval was alternated each session, and participants were encouraged to match or exceed their average power from the previous session. Group training data from each session are reported (Table 3.3). High-intensity interval training. During the 3-week intervention period, RUN performed two high-intensity running interval training sessions per week. All sessions were performed as a group, and different training locations were used to accommodate the team’s existing schedule. The first session of the week was performed on treadmills (Woodway, Eugene, USA) after the team’s first resistance training session of the week (Table 3.2). Each 25–30 min treadmill session consisted of a 5-min jog, which gradually progressed to a 30-s run at each participant’s predetermined training speed, which for the first interval of the first session was the estimated maximal aerobic speed derived from Table 3.3. Single-leg cycling training data Duration (s) Power (W) Cadence (rpm) Session # Int 1 Int 2 Int 3 Int 1 Int 2 Int 3 1 179 (5) 179 (5) 180 (6) 159 (23) 143 (21) 134 (18) 83 (9) 85 (6) 2 181 (2) 182 (4) 181 (4) 178 (21) 150 (19) 146 (21) 90 (5) 3 182 (4) 183 (5) 182 (3) 180 (20) 163 (26) 153 (22) 4 182 (2) 181 (2) 181 (2) 185 (19) 176 (18) 5 181 (2) 180 (3) 180 (4) 193 (22) 6 181 (3) 180 (2) 180 (2) 197 (22) Int 1 Int 2 Int 3 HR15 (bpm) HR180 (bpm) Int 1 Int 2 Int 3 Int 1 Int 2 Int 3 87 (4) 174 (12) 176 (12) 179 (10) 163 (13) 171 (9) 172 (9) 88 (4) 88 (5) 173 (13) 177 (11) 179 (11) 163 (13) 168 (12) 170 (11) 91 (7) 89 (6) 89 (7) 173 (13) 178 (14) 179 (9) 163 (15) 170 (12) 170 (10) 168 (18) 90 (9) 89 (8) 88 (5) 172 (11) 181 (11) 181 (11) 158 (12) 171 (11) 171 (10) 175 (18) 166 (19) 88 (8) 86 (8) 84 (7) 173 (12) 178 (12) 178 (10) 158 (13) 169 (12) 169 (10) 180 (22) 183 (23) 89 (7) 87 (5) 87 (5) 170 (11) 177 (11) 179 (10) 158 (11) 168 (10) 169 (9) Note. All values are mean (SD). The duration (s), power (W), pedal cadence (rpm), average heart rate (bpm) during the final 15 s (HR15) and entire 180 s (HR180) of each interval (Int) are listed for each session (#1–6). 58 the MSFT (Berthoin et al., 1994). After a 2-min passive rest, participants performed three high-intensity running intervals lasting 3 min and then walked (~5.6 km/h) for 3 min inbetween each interval. Vigorous encouragement was provided to maintain the prescribed speed for the entire interval. Thereafter, speeds were voluntarily increased during subsequent intervals and across sessions in order to achieve the greatest running speed by the end of the intervention. Participants walked at self-selected speed for 5–10 min to cool down at the end of each session. Heart rate monitors (Polar USA, New York, USA) were worn during each treadmill running session, and a single heart rate measurement for each was recorded during the final 15 s of each interval due to equipment limitations. The second weekly training session was performed on ~100 m of artificial grass surface after the team’s first lacrosse practice of the week. This 20–25 min session began with 2–3 min of jogging followed by 1–2 min of passive recovery. Three high-intensity running intervals lasting ~3 min were interspersed by ~3 min of walking at self-selected pace. To guide the prescribed pace during the artificial turf sessions, individual running distances were calculated in order to yield the intended average speed that could be achieved if the interval was completed in exactly 3 min. The calculated distance for each individual was divided by eight and a cone was placed at the corresponding distance. Each participant ran to his individual cone and back four times and was instructed to complete each roundtrip in 45 s. Similar to the treadmill sessions, vigorous encouragement was provided to complete the prescribed distance within 3 min. The actual time it took to run the prescribed distance was recorded in order to calculate the average speed for each interval. Interval distances were adjusted within each session and across sessions in order to achieve the greatest average speed by the end of the training 59 intervention. It should be noted that even though running on turf at a given speed is more physiologically demanding than running the same speed on a treadmill (Di Michelle, Di Renzo, Ammazzalorso, & Merni, 2009), participants were still encouraged to match or exceed the average speed achieved the treadmill session they performed earlier in the week because they were expected to improve each session. Group training data for all six sessions are reported (Table 3.4). Data analysis The final two rehearsal trials performed during the 8-week training and familiarization period were used to calculate the intraclass correlation coefficient (ICC) and typical error of measurement (TE) for single-sprint performance, repeated-sprint ability, repeated-sprint fatigue, and endurance running performance (Hopkins, 2000). Table 3.4. High-intensity running interval training data Duration (s) Speed (km/h) HR15 (bpm) Session # Surface Int 1 Int 2 Int 3 Int 1 Int 2 Int 3 Int 1 Int 2 Int 3 1 Treadmill 180 (–) 180 (–) 180 (–) 14.1 (0.7) 14.0 (0.9) 13.9 (1.1) 178 (9) 184 (9) 183 (12) 2 Artificial 166 (6) 180 (11) 177 (10) 15.4 (1.4) 14.2 (1.2) 14.4 (1.1) – – – 3 Treadmill 180 (–) 180 (–) 180 (–) 14.7 (1.1) 14.7 (1.1) 14.7 (1.1) 182 (6) 187 (5) 189 (5) 4 Artificial 173 (4) 178 (2) 175 (6) 14.9 (0.8) 14.5 (1.0) 14.7 (1.1) – – – 5 Treadmill 180 (–) 180 (–) 180 (–) 15.2 (1.1) 15.2 (1.1) 15.2 (1.1) 183 (10) 186 (6) 190 (6) 6 Artificial 178 (4) 185 (7) 187 (8) 15.4 (1.0) 14.8 (1.2) 14.7 (1.0) – – – Note. All values are mean (SD). The duration (s), speed (km/h), and heart rate (bpm) during the final 15 s (HR15) of each interval (Int) are listed for each session (#1–6). Note that interval duration was assumed to be precisely 180 s during all treadmill sessions, and heart rate monitors were not worn during training sessions on artificial turf due to equipment limitations. 60 Changes in single-sprint performance, repeated-sprint ability, repeated-sprint fatigue, and endurance running performance were analyzed using linear mixed models (α = .05), with fixed effects for group (CON, SLC, or RUN) and time (PRE & POST), their interaction, and a random intercept for each subject. Additionally, individual changes in performance within SLC and RUN from before to after the intervention were compared using independent samples t tests (α = .05). To complement the statistical hypothesis testing, all differences were considered in light of the smallest detectible change (SDC; Bruynesteyn, Boers, Kostense, van der Linden, & van der Heijde, 2005). Calculated from the TE, the SDC represents the minimal difference (in the units measured) between two measurements that is required to be confident (± 90%) that the difference is beyond the error within both measurements (Bruynesteyn et al., 2005). A difference of 1 SDC does not necessarily indicate a smallest worthwhile difference (or change), that is, the minimal difference that would truly affect athletic performance. However, identifying the smallest worthwhile difference in teamsport is difficult because the published literature has not readily reported clear links between fitness test performance changes and player performance during competition. Many have arbitrarily defined a fixed effect size of 0.2 standard deviations (Cohen, 1992) as the smallest worthwhile difference. Such methodology would be considered aggressive, especially if the TE, let alone the SDC, equated to a larger effect size. Therefore, due to lack of a better indication, the SDC was used as the smallest worthwhile difference for each dependent variable, and all within-group and betweengroup effects were expressed relative to 1 SDC. 61 Results All data are expressed as mean (SD) unless otherwise noted. According to standards suggested by Atkinson and Nevill (1998), the familiarization trials yielded good reliability (ICC = .80–.89) for single-sprint performance (ICC = .87; TE = 0.12 s), high reliability (ICC ≥ .90) for repeated-sprint ability (ICC = .97; TE = 0.73 s), and endurance running performance (ICC = .91; TE = 104 m), and questionable reliability (ICC = .70–.79) for repeated-sprint fatigue (ICC = .75; TE = 1.26%). As stated above, the smallest worthwhile difference was indicated by a change that exceeded 1 SDC for each variable, which equated to 0.21 s for single-sprint performance, 1.26 s for repeated-sprint ability, 2.17% for repeated-sprint fatigue, and 178 m for endurance running performance. Only seven participants from CON finished the study because two of them suffered sport-related injuries. All nine participants from SLC and RUN finished the study. The results for each group from before to after the intervention are provided (Table 3.5). Before the intervention, the linear mixed models indicated that single-sprint Table 3.5. Group results from before to after the intervention period Single-Sprint Performance (s) Repeated-Sprint Ability (s) Repeated-Sprint Fatigue (%) Endurance Running Performance (m) Before After Before After Before After Before After CON 7.07 (0.26) 6.98 (0.14) 79.05 (3.28) 78.03 (3.45) 11.8 (4.6) 12.0 (5.3) 1642 (303) 1691 (429) SLC 7.04 (0.27) 7.09 (0.30) 79.62 (3.61) 76.53 (2.97) 13.1 (3.4) 7.9 (1.1) 1649 (273) 1878 (234) RUN 7.15 (0.40) 7.14 (0.41) 80.12 (5.92) 77.17 (4.48) 12.0 (4.1) 8.2 (2.5) 1711 (301) 1924 (412) Note. Results for the control group (CON), single-leg cycling training group (SLC), and running interval training group (RUN) are listed (mean [SD]) from before to after the intervention. The post-intervention values for CON are italicized because they were calculated from the seven out of nine participants in that group who finished the study. 62 performance, repeated-sprint ability, repeated-sprint fatigue, and endurance running performance were not statistically difference between, SLC, RUN, and CON (Table 3.6). In the 10x40, single-sprint performance within CON did not significantly change from before to after the intervention and there were no group by time interaction effects for SLC or RUN compared to CON (Table 3.6; Figure 3.1). Within-group changes in single-sprint performance were < 1 SDC for all groups. Within CON, three participants improved and one participant slowed down by more than 1 SDC. One participant from SLC and RUN improved by at least 1 SDC, and two participants in SLC slowed down by more than 1 SDC (Figure 3.1). For repeated-sprint ability, one individual within CON was the only participant in the study whose total time in the 10x40 did not decrease after the intervention period (Figure 3.2). Before removing that participant, the linear mixed model indicated a nonsignificant change in repeated-sprint ability within CON, and there was a significant group by time interaction indicating greater improvements within RUN and SLC compared to the change within CON (Table 3.6). However, another linear mixed model without that subject revealed that there was a statistically significant improvement (mean: -1.97 s; SD: 1.61 s; p = .003) in repeated-sprint ability within CON that was equivalent to 1.57 SDC, and there was no longer a significant group by time interaction effect for SLC (p = .168) and RUN (p = .22) compared to CON (Figure 3.2; Table 3.7). Four out of the six remaining participants in CON improved their total time in the 10x40 by at least 1 SDC (Figure 3.2). SLC had a large within-group change (mean: -3.09 s; SD: 1.80 s) equivalent to 2.46 SDC, and all nine participants improved by at least 1 SDC (Figure 3.2). Similarly, the average within-group change (mean: -2.96 s; SD: 1.69 s) was 63 Table 3.6. Results of linear mixed models beta (SE) 90% CI p Single-Sprint Performance (s) Intercept 7.07 (0.10) [6.88, 7.26] < .001 SLC -0.03 (0.14) [-0.30, 0.24] .818 RUN 0.08 (0.14) [-0.19, 0.35] .563 TIME -0.09 (0.06) [-0.22, 0.03] .169 SLC*TIME 0.14 (0.09) [-0.02, 0.31] .102 RUN*TIME 0.07 (0.09) [-0.09, 0.24] .404 Intercept 79.05 (1.27) [76.56, 81.55] < .001 SLC 0.56 (1.80) [-2.97, 4.09] .757 RUN 1.07 (1.80) [-2.46, 4.60] .557 Repeated-Sprint Ability (s) TIME -1.22 (0.67) [-2.52, 0.09] .080 SLC*TIME -1.87 (0.89) * [-3.62, -0.13] .046 RUN*TIME -1.74 (0.89) [-3.49, 0.01] .062 Intercept 11.8 (1.1) [9.6, 14.1] < .001 SLC 1.3 (1.6) [-1.9, 4.5] .439 RUN 0.1 (1.6) [-3.1, 3.3] .935 Repeated Sprint Fatigue (%) TIME -0.1 (1.3) [-2.6, 2.3] .921 SLC*TIME -5.1 (1.7) * [-8.4, -1.8] .006 RUN*TIME -3.7 (1.7) * [-7.0, -0.4] .039 1642 (102) [1443, 1841] < .001 SLC 7 (144) [-275, 288] .963 RUN 69 (144) [-213, 350] .635 TIME 77 (49) [-19, 173] .127 SLC*TIME 152 (65) * [24, 280] .029 RUN*TIME 136 (65) * [8, 264] .047 Endurance Running Performance (m) Intercept Note. Beta coefficients in bold indicate when the within-group change was greater than the smallest worthwhile difference from before to after the training intervention. Beta coefficients in bold italics indicate when the single-leg cycling (SLC) or running (RUN) training groups had differences compared to the control group (Intercept) that were beyond the smallest worthwhile difference. Negative beta coefficients indicate improvements for single-sprint performance, repeated-sprint performance, and repeatedsprint fatigue, and a positive beta coefficient indicates an improvement in endurance running performance. The linear mixed model estimated a beta coefficient for the control group from before to after (TIME) the intervention based on the fact that two participants were excluded due to sport related injuries. SE = standard error; CI = confidence interval. * significantly different than the control group (p ≤ .05) 64 CON SLC RUN Single-Sprint Performance (s) 8.00 7.75 7.50 7.25 7.00 6.75 6.50 6.25 6.00 5.75 0.00 5.50 Before After Before After Before After Figure 3.1. Group and individual changes in single-sprint performance. Individual ( ) and group (columns) changes in single-sprint performance are illustrated for the control (CON), single-leg cycling interval training (SLC), and high-intently running interval training (RUN) groups from before to after the intervention period. Changes that were greater than the smallest worthwhile change are indicated by solid black lines, whereas grey lines indicate trivial differences. Note that errors bars are ± SD. CON SLC RUN Repeated-Sprint Ability (s) 95.00 90.00 85.00 * * * 80.00 75.00 70.00 65.00 60.00 55.00 0.00 Before After Before After Before After Figure 3.2. Group and individual changes in repeated-sprint ability. Solid black lines indicate individual ( ) and group (columns) changes that were greater than the smallest worthwhile change, whereas grey lines indicate trivial differences from before to after the intervention. One participant in the control group (CON) was the only individual in the study whose total time did not decrease. Before removal of that influential participant the linear mixed model indicated a nonsignificant change (column with grey border). After excluding that participant, the linear mixed model yielded a significant change that was beyond the smallest worthwhile difference (*; column with black border). Changes within the single-leg cycling (SLC) and high-intensity running (RUN) training groups were also greater than the smallest worthwhile difference, but there was not a significant group by time interaction compared to CON. Note that error bars are ± SD. 65 Table 3.7. Linear mixed model results for repeated-sprint ability after removal of influential subject Repeated-Sprint Ability (s) beta (SE) 90% CI p Intercept 79.05 (1.27) [76.57, 81.53] < .001 SLC 0.56 (1.79) [-2.95, 4.07] .755 RUN 1.07 (1.79) [-2.44, 4.58] .554 TIME -1.97 (0.61)* [-3.16, -0.78] .003 SLC*TIME -1.12 (0.79)* [-2.66, 0.42] .168 RUN*TIME -0.98 (0.79)* [-2.52, 0.56] .223 Note. The beta coefficient for the control group from before to after the intervention (TIME) was estimated from all nine participants before the intervention (Intercept) and six participants after the intervention because two could not complete the study, and another was excluded due to an uncharacteristic result. Beta coefficients in bold indicate when the within-group change was greater than the smallest worthwhile difference from before to after the training intervention. SE = standard error; CI = confidence interval. * significantly different within-group change from before the intervention (p ≤ .05) equivalent to 2.35 SDC for RUN, within which eight out of nine participants improved by ≥ 1 SDC (Figure 3.2). Even though there was not a significant group by time interaction after removal of the influential subject, the relative improvements within SLC and RUN were at least 50% greater than the improvement within CON (Table 3.7). The changes within SLC and RUN were nearly identical (p = .87), indicating that SLC was equally as effective as RUN. Repeated-sprint fatigue significantly decreased (i.e., improved) within SLC (mean: -5.22%; SD: 3.76%; SDC: -2.40; p = .006) and RUN (mean: -3.80%; SD: 2.80%; SDC: -1.75; p = .039) compared to CON (mean: 0.12%; SD: 4.38%; SDC: 0.06; p = .921; Table 3.6, Figure 3.3). Overall, the respective improvements within SLC and RUN were 2.34 SDC and 1.69 SDC greater than the change within CON. There were mixed individual responses within CON, that is, three individuals had less repeated-sprint fatigue (i.e., improvement by at least 1 SDC), and two individuals had greater repeatedsprint fatigue. For SLC and RUN, individual responses were more uniform, as seven out 66 CON SLC RUN Repeated-Sprint Fatigue (%) 22.5 20.0 17.5 15.0 12.5 * 10.0 * 7.5 5.0 2.5 0.0 Before After Before After Before After Figure 3.3. Group and individual changes in repeated-sprint fatigue. Individual ( ) and group (columns) changes that were greater than the smallest worthwhile difference are illustrated with black lines, and gray lines indicate trivial changes. Repeated-sprint fatigue did not change within the control group (CON) from before to after the intervention. Compared to CON, there was a significant group by time interaction (*) for both the single-leg cycling (SLC) and running (RUN) groups. Note that error bars are ± SD. of nine participants in each group improved by at least 1 SDC (Figure 3.3). The mean difference (0.65 SDC) between SLC and RUN was not significantly different (p = .38), indicating that SLC and RUN elicited similar responses. Endurance running performance (i.e., distance covered during the MSFT) did not significantly change (mean: 77 m; SD: 132 m; SDC: 0.43; p = .127) within CON from before to after the intervention (Table 3.6). Compared to CON, there was a significant group by time interaction for both SLC (p = .029) and RUN (p = .047), who improved by 1.29 SDC (mean: 229 m; SD: 136 m) and 1.20 SDC (mean: 213 m; SD: 147 m), respectively (Table 3.6, Figure 3.4). Despite the significant group by time interaction for both SLC and RUN, the mean differences between their improvements and the change within CON were within 1 SDC. Nevertheless, there were only three participants from CON who improved by at least 1 SDC, compared to seven out of nine participants in SLC and RUN who improved (Figure 3.4). Similar to the other variables, SLC and RUN 67 Endurance Running Performance (m) CON SLC RUN 2,500 * 2,250 * 2,000 1,750 1,500 1,250 1,000 750 500 0 Before After Before After Before After Figure 3.4. Group and individual changes in endurance running performance. Individual ( ) and group (columns) changes that were greater than the smallest worthwhile difference are illustrated with black lines, and gray lines indicate trivial changes. Endurance running performance did not change within the control group (CON), within which only three individuals ( ) improved by the smallest worthwhile change. Compared to CON, there was a significant group by time interaction (*) for both the single-leg cycling (SLC) and running (RUN) groups (columns with black lines). Likewise, seven out of nine individuals ( ) in both groups improved by more than the smallest worthwhile change. Note that error bars are ± SD. elicited very similar responses (p = .82). Discussion To my knowledge, this is the first investigation of the effect of single-leg cycling training on high-intensity running performance. The main finding of this study was that short-term, high-intensity single-leg cycling interval training can improve repeated-sprint ability and endurance running performance among ground-based team-sport athletes. The combined effect of preseason team-sport practice and resistance training led to significant improvements in repeated-sprint ability (Table 3.7, Figure 3.2), but not endurance running performance (Table 3.6, Figure 3.4). Adding single-leg cycling training facilitated a practical trend toward even greater repeated-sprint ability, which was due to unequivocally larger reductions in repeated-sprint fatigue (Table 3.6, Figure 3.3). Single- 68 leg cycling training also contributed to an improvement in running endurance that was significantly greater than sport participation alone (Table 3.6, Figure 3.4). Finally, singleleg cycling training elicited comparable improvements in field-measures of high-intensity running performance to those resulting from high-intensity running interval training (Figures 3.2–3.4). In this study, adding single-leg cycling training to a 3-week preseason training period elicited large improvements in both repeated-sprint ability (2.46 SDC) and endurance running performance (1.29 SDC). My findings extend upon previous research reporting positive adaptations from single-leg cycling training. Abbiss and colleagues (2011) reported that single-leg cycling induced dramatic increases in muscle respiratory proteins COX II and IV among highly trained cyclists. While I did not measure respiratory proteins, the protocol in my study was quite similar and thus likely elicited similar or even greater increases given the recreational training state (Rhea, 2004) of the current participants. In conjunction with other studies reporting that single-leg cycling training elicits adaptations indicative of greater muscle oxidative capacity (Dela et al., 1993; Henriksson, 1977; Klausen et al., 1982; MacInnis, Zacharewicz, et al., 2017; Rud et al., 2012; Saltin et al., 1976; Vincent et al., 2015), it is therefore plausible that the observed improvements in both repeated-sprint ability and endurance running performance after single-leg cycling training were due to peripheral adaptations rather than central cardiac changes. More research is needed to confirm this hypothesis, but peripheral adaptations would facilitate greater recovery between sprints (see Girard et al., 2011 and Stone & Kilding, 2009, for reviews). Indeed, the improvement in repeated-sprint ability after 69 single-leg cycling training was due to a systematic reduction (2.40 SDC) in repeatedsprint fatigue (Figure 3.3). Repeated-sprint fatigue is negatively correlated with muscle oxidative capacity (Thomas, Sirvent, Perrey, Raynaud, & Mercier, 2004), which can be measured by the rate of phosphocreatine restoration (Haseler, Hogan, & Richardson, 1999; Mahler, 1985; Prompers, Wessels, Kemp, & Nicolay, 2014). Previous authors (Forbes, Slade, & Meyer, 2008) reported a 14% decrease in the time-constant of phosphocreatine recovery after performing six high-intensity double-leg cycling interval sessions across 2 weeks, and they attributed the changes to peripheral adaptations. Compared to double-leg cycling, Abbiss and colleagues (2011) reported significantly greater peripheral adaptations. Thus, peripheral adaptations may have contributed to the observed reductions in repeated-sprint fatigue after single-leg cycling training. Likewise, similar adaptations may have facilitated greater endurance running performance (1.29 SDC) within SLC. Briefly, greater distance covered during the MSFT positively correlates with increased maximal aerobic capacity (Mayorga-Vega et al., 2015). Maximal aerobic capacity reflects whole body aerobic consumption, which is the product of central and peripheral aspects. Compared to large muscle mass exercise (e.g., running or double-leg cycling), single-leg cycling exercise stimulates greater blood flow to the exercising muscle (Burns et al., 2014; Klausen et al., 1982) at a lower central cardiac demand (Abbiss et al., 2011; Gleser, 1973; Klausen et al., 1982). Indeed, heart rates during the final 15 s of the single-leg cycling intervals (Table 3.3) were consistently lower than those during the final 15 s of the high-intensity running intervals (Table 3.4). More blood flow enables greater intensity to the exercising limb during single-leg cycling (Abbiss et al., 2011; Bundle et al., 2006; MacInnis, Morris, et al., 2017), which would 70 contribute, in part, to why single-leg cycling training has repeatedly elicited significant peripheral adaptations. By the sixth single-leg cycling session, the cycling naive participants in this study achieved average single-leg cycling interval work rates that matched those achieved by the highly trained cyclists who participated in the study by Abbiss and colleagues (2011)—albeit their intervals were 4 min. In other words, the training likely elicited a profound peripheral stress to the legs, and any potential improvements in the ability of the lower extremity musculature to extract and utilize oxygen would facilitate an improvement in whole-body aerobic capacity. While previous research supports the notion that single-leg cycling elicits significant peripheral adaptations, more research is needed to truly verify the extent by which central and/or peripheral adaptations from single-leg cycling training would contribute to greater endurance running performance. SLC elicited an improvement in repeated-sprint ability that was ~57% greater than that within CON. However, it was novel to observe an improvement (1.57 SDC) in repeated-sprint ability (Table 3.7) within CON because there is not readily available research that has reported the effect of general sport practice participation and resistance training on repeated-sprint ability. Certainly, previous research has indicated improvements in repeated-sprint ability from sport-specific conditioning activities, like small-sided games, that were performed in place of traditional conditioning activities (see Bujalance-Moreno, Latorre-Román, & García-Pinillos, 2019, for a review). However, those studies would not apply to this one because the coaches responsible for the practice activities in this study did not implement a structured conditioning regimen during the intervention period. 71 It was not immediately clear from the data whether or not the improvement in repeated-sprint ability within CON was due to changes in single-sprint performance or repeated-sprint fatigue because both variables did not significantly change. Coincidently, repeated-sprint fatigue did not improve among the three CON participants who had better single-sprint performance, and single-sprint performance did not improve among the three other CON participants who had less repeated-sprint fatigue. To expound on that observation, the three participants who improved their single-sprint performance also had the poorest single-sprint performance before the intervention, and the same was generally true for the three participants who reduced their repeated-sprint fatigue. In other words, it appears that team practice participation and resistance training provided the CON participants with “what they needed” to improve their repeated-sprint ability. Conversely, single-leg cycling training elicited a systematic reduction in repeatedsprint fatigue (Figure 3.3) which facilitated an average improvement in repeated-sprint ability that was 0.89 SDC greater than the average improvement within CON. In this study, 1 SDC was designated as the smallest worthwhile difference, but it is not known how much of a difference is truly relevant from a practical standpoint. Practically speaking, it appears sport participation and resistance training can improve repeatedsprint ability through unpredictable improvements in either single-sprint performance and/or repeated-sprint fatigue, whereas the addition of single-leg cycling training can elicit potentially greater improvements in repeated-sprint ability by systematically reducing repeated-sprint fatigue. Furthermore, adding single-leg cycling to an existing sport participation schedule can offer improvements in endurance running performance that may not be possible from general sport participation that does not include other 72 cardiorespiratory fitness training strategies. My data have exciting translational importance because they also demonstrated that single-leg cycling training yielded similar improvements in repeated-sprint ability and running endurance as high-intensity running interval training. Using high-intensity running intervals was not a primary focus of this study, as it is already a widely-used, time-efficient training strategy that elicits significant improvements in cardiorespiratory and metabolic function (Bishop et al., 2011; Buchheit & Laursen, 2013; Stone & Kilding, 2009). Assigning a group to RUN, however, enabled the ability to quantify the efficacy of single-leg cycling training in terms of the range of possible improvements between the status-quo (i.e., RUN) and sport participation (i.e., CON). Cross-training doesn’t appear to elicit more beneficial effects than mode-specific training (Tanaka, 1994), which was true in the current study. However SLC yielded the maximum possible benefit indicated by RUN. In light of recent studies reporting greater running performance after short-term double-leg cycling interval training (Chan et al., 2018; Etxebarria et al., 2014), my findings provide further support the concept of cross-training. Compared to double-leg cycling, however, single-leg cycling may be a more potent (Abbiss et al., 2011) way to improve high-intensity running performance. Practical Applications Single-leg cycling is a low-impact cross-training modality that can enhance highintensity running performance, particularly during times when additional running activities could increase fatigue (Young et al., 2012) or overuse injury risk (Malone et al., 2018). More research is needed to determine the smallest worthwhile differences in fitness that that would truly benefit sports performance, but previous authors have 73 reported that greater aerobic running capacity (Malone et al., 2018) and repeated-sprint ability (Malone, Hughes, Doran, Collins, & Gabbett, 2019) are associated with a lower incidence of injury during team-sport practice and competition. Therefore, the benefits of single-leg cycling training could mitigate risk of injury during play, while simultaneously accommodating crucial moments of play associated with greater repeated-sprint activity (Girard et al., 2011), as well as demanding competitive situations requiring greater highintensity physical activity (Kempton & Coutts, 2016; Trewin et al., 2017; Vinson et al., 2018). My study tested the efficacy of a short-term training intervention, and there is a need to investigate if single-leg cycling training can benefit athletes over the long run. Furthermore, my findings apply to seasoned team-sport athletes with recreational strength and conditioning training experience. Future research could determine if elite athletes with greater training status could also benefit from single-leg cycling training. One problem with single-leg cycling training is that the protocols can take twice as long as typical high-intensity interval training routines involving both limbs (i.e., double-leg cycling or running). As done in the current study, alternating legs between intervals enabled a more time-efficient single-leg cycling training protocol than any other previously reported protocol. In this study, single-leg cycling training was successfully integrated into an existing collegiate-athletics participation schedule, making it a viable training option for athletes with significant time-demands who seek greater high-intensity running performance. In conclusion, single-leg cycling is a potent cross-training modality that can improve field-measures of high-intensity running performance among team-sport athletes in a time-efficient manner. 74 References Abbiss, C. R., Karagounis, L. G., Laursen, P. B., Peiffer, J. J., Martin, D. T., Hawley, J. A., … Martin, J. C. (2011). Single-leg cycle training is superior to double-leg cycling in improving the oxidative potential and metabolic profile of trained skeletal muscle. Journal of Applied Physiology, 110(5), 1248-1255. Atkinson, G., & Nevill, A. M. (1998). Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Medicine, 26(4), 217-238. Berthoin, S., Gerbeaux, M., Turpin, E., Guerrin, F., Lensel-Corbeil, G., & Vandendorpe, F. (1994). Comparison of two field tests to estimate maximum aerobic speed. Journal of Sports Sciences, 12(4), 355-362. Bezodis, N. E., North, J. S., & Razavet, J. L. (2017). Alterations to the orientation of the ground reaction force vector affect sprint acceleration performance in team sports athletes. Journal of Sports Sciences, 35(18), 1817-1824. Bishop, D., Girard, O., & Mendez-Villanueva, A. (2011). Repeated-sprint ability—Part II. Sports Medicine, 41(9), 741-756. Bradley, P. S., Carling, C., Diaz, A. G., Hood, P., Barnes, C., Ade, J., … Mohr, M. (2013). Match performance and physical capacity of players in the top three competitive standards of English professional soccer. Human Movement Science, 32(4), 808-821. Bruynesteyn, K., Boers, M., Kostense, P., van der Linden, S., & van der Heijde, D. (2005). Deciding on progression of joint damage in paired films of individual patients: Smallest detectable difference or change. Annals of the Rheumatic Diseases, 64(2), 179-182. Buchheit, M., & Laursen, P. B. (2013). High-intensity interval training: Solutions to the programming puzzle. Sports Medicine, 43(10), 927-954. Bujalance-Moreno, P., Latorre-Román, P. Á., & García-Pinillos, F. (2019). A systematic review on small-sided games in football players: Acute and chronic adaptations. Journal of Sports Sciences, 37(8) 921-949. Bundle, M. W., Ernst, C. L., Bellizzi, M. J., Wright, S., & Weyand, P. G. (2006). A metabolic basis for impaired muscle force production and neuromuscular compensation during sprint cycling. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 291(5), R1457-R1464. Burns, K. J., Pollock, B. S., LaScola, P., & McDaniel, J. (2014). Cardiovascular responses to counterweighted single-leg cycling: Implications for rehabilitation. European Journal of Applied Physiology, 114(5), 961-968. 75 Chan, H. C.-K., Ho, W. K.-Y., & Yung, P. S.-H. (2018). Sprint cycling training improves intermittent run performance. Asia-Pacific Journal of Sports Medicine, Arthroscopy, Rehabilitation and Technology, 11, 6-11. Chtourou, H., & Souissi, N. (2012). The effect of training at a specific time of day: A review. Journal of Strength & Conditioning Research, 26(7), 1984-2005. Cicioni-Kolsky, D., Lorenzen, C., Williams, M. D., & Kemp, J. G. (2013). Endurance and sprint benefits of high-intensity and supramaximal interval training. European Journal of Sport Science, 13(3), 304-311. Clark, K. P., & Weyand, P. G. (2014). Are running speeds maximized with simple-spring stance mechanics? Journal of Applied Physiology, 117(6), 604-615. Cohen, J. (1992). Statistical Power Analysis. Current Directions in Psychological Sciences, 1(3), 98-101. Datson, N., Hulton, A., Andersson, H., Lewis, T., Weston, M., Drust, B., & Gregson, W. (2014). Applied physiology of female soccer: An update. Sports Medicine, 44(9), 1225-1240. Dela, F., Handberg, A., Mikines, K. J., Vinten, J., & Galbo, H. (1993). GLUT 4 and insulin receptor binding and kinase activity in trained human muscle. The Journal of Physiology, 469(1), 615-624. Di Michele, R., Di Renzo, A. M., Ammazzalorso, S., & Merni, F. (2009). Comparison of physiological responses to an incremental running test on treadmill, natural grass, and synthetic turf in young soccer players. Journal of Strength & Conditioning Research, 23(3), 939-945. Elmer, S. J., McDaniel, J., and Martin, J. C. (2016). Biomechanics of counterweighted one-legged cycling. Journal of Applied Biomechanics, 32(1), 78-85. Etxebarria, N., Anson, J. M., Pyne, D. B., & Ferguson, R. A. (2014). High-intensity cycle interval training improves cycling and running performance in triathletes. European Journal of Sport Science, 14(6), 521-529. Forbes, S. C., Slade, J. M., & Meyer, R. A. (2008). Short-term high-intensity interval training improves phosphocreatine recovery kinetics following moderate-intensity exercise in humans. Applied Physiology, Nutrition, and Metabolism, 33(6), 11241131. Gabbett, T., Kelly, J., Ralph, S., & Driscoll, D. (2009). Physiological and anthropometric characteristics of junior elite and sub-elite rugby league players, with special reference to starters and non-starters. Journal of Science and Medicine in Sport, 12(1), 215-222. Gleser, M. A. (1973). Effects of hypoxia and physical training on hemodynamic 76 adjustments to one-legged exercise. Journal of Applied Physiology, 34(5), 655659. Girard, O., Mendez-Villanueva, A., & Bishop, D. (2011). Repeated-sprint ability—Part I. Sports Medicine, 41(8), 673-694. Glaister, M., Howatson, G., Pattison, J. R., & McInnes, G. (2008). The reliability and validity of fatigue measures during multiple-sprint work: An issue revisited. Journal of Strength & Conditioning Research, 22(5), 1597-1601. Gordon, N., Abbiss, C. R., Ihsan, M., Maiorana, A. J., & Peiffer, J. J. (2018). Active and inactive leg hemodynamics during sequential single-leg interval cycling. Medicine and Science in Sports and Exercise, 50(60), 1297-1304. Haseler, L. J., Hogan, M. C., & Richardson, R. S. (1999). Skeletal muscle phosphocreatine recovery in exercise-trained humans is dependent on O2 availability. Journal of Applied Physiology, 86(6), 2013-2018. Helgerud, J., Engen, L. C., Wisløff, U., & Hoff, J. (2001). Aerobic endurance training improves soccer performance. Medicine and Science in Sports and Exercise, 33(11), 1925-1931. Henriksson, J. (1977). Training induced adaptation of skeletal muscle and metabolism during submaximal exercise. The Journal of Physiology, 270(3), 661-675. Hoes, M. J., Binkhorst, R. A., Smeekes-Kuyl, A. E., & Vissers, A. C. (1968). Measurement of forces exerted on pedal and crank during work on a bicycle ergometer at different loads. European Journal of Applied Physiology and Occupational Physiology, 26(1), 33-42. Hopkins, W. G. (2000). Measures of reliability in sports medicine and science. Sports Medicine, 30(1), 1-15. Hunkin, S. L., Fahrner, B., & Gastin, P. B. (2014). Creatine kinase and its relationship with match performance in elite Australian rules football. Journal of Science and Medicine in Sport, 17(3), 332-336. Jennings, D. H., Cormack, S. J., Coutts, A. J., & Aughey, R. J. (2012). International field hockey players perform more high-speed running than national-level counterparts. Journal of Strength & Conditioning Research, 26(4), 947-952. Johnston, R. D., Gabbett, T. J., Jenkins, D. G., & Hulin, B. T. (2015). Influence of physical qualities on post-match fatigue in rugby league players. Journal of Science and Medicine in Sport, 18(2), 209-213. Keller, T. S., Weisberger, A. M., Ray, J. L., Hasan, S. S., Shiavi, R. G., & Spengler, D. M. (1996). Relationship between vertical ground reaction force and speed during walking, slow jogging, and running. Clinical Biomechanics, 11(5), 253-259. 77 Kempton, T., & Coutts, A. J. (2016). Factors affecting exercise intensity in professional rugby league match-play. Journal of Science and Medicine in Sport, 19(6), 504508. Klausen, K., Secher, N. H., Clausen, J. P., Hartling, O., & Trap-Jensen, J. (1982). Central and regional circulatory adaptations to one-leg training. Journal of Applied Physiology, 52(4), 976-983. Laughlin, M. H. (1987). Skeletal muscle blood flow capacity: Role of muscle pump in exercise hyperemia. American Journal of Physiology-Heart and Circulatory Physiology, 253(5), H993-H1004. MacInnis, M. J., Morris, N., Sonne, M. W., Zuniga, A. F., Keir, P. J., Potvin, J. R., & Gibala, M. J. (2017). Physiological responses to incremental, interval, and continuous counterweighted single-leg and double-leg cycling at the same relative intensities. European Journal of Applied Physiology, 117(7), 1423-1435. MacInnis, M. J., Zacharewicz, E., Martin, B. J., Haikalis, M. E., Skelly, L. E., Tarnopolsky, M. A., … Gibala, M. J. (2017). Superior mitochondrial adaptations in human skeletal muscle after interval compared to continuous single-leg cycling matched for total work. The Journal of Physiology, 595(9), 2955-2968. Malone, S., Hughes, B., Doran, D. A., Collins, K., & Gabbett, T. J. (2019). Can the workload-injury relationship be moderated by improved strength, speed and repeated-sprint qualities? Journal of Science and Medicine in Sport, 22(1), 29-34. Malone, S., Owen, A., Mendes, B., Hughes, B., Collins, K., & Gabbett, T. J. (2018). High-speed running and sprinting as an injury risk factor in soccer: Can welldeveloped physical qualities reduce the risk? Journal of Science and Medicine in Sport, 21(3), 257-262. Mahler, M. (1985). First-order kinetics of muscle oxygen consumption, and an equivalent proportionality between QO2 and phosphorylcreatine level. Implications for the control of respiration. The Journal of General Physiology, 86(1), 135-165. Mayorga-Vega, D., Aguilar-Soto, P., & Viciana, J. (2015). Criterion-related validity of the 20-m shuttle run test for estimating cardiorespiratory fitness: A meta-analysis. Journal of Sports Science & Medicine, 14(3), 536. Mohr, M., Krustrup, P., & Bangsbo, J. (2003). Match performance of high-standard soccer players with special reference to development of fatigue. Journal of Sports Sciences, 21(7), 519-528. Narazaki, K., Berg, K., Stergiou, N., & Chen, B. (2009). Physiological demands of competitive basketball. Scandinavian Journal of Medicine & Science in Sports, 19(3), 425-432. Polley, C. S., Cormack, S. J., Gabbett, T. J., & Polglaze, T. (2015). Activity profile of 78 high-level Australian lacrosse players. Journal of Strength & Conditioning Research, 29(1), 126-136. Prompers, J. J., Wessels, B., Kemp, G. J., & Nicolay, K. (2014). Mitochondria: Investigation of in vivo muscle mitochondrial function by 31 P magnetic resonance spectroscopy. The International Journal of Biochemistry & Cell Biology, 50, 67-72. Rampinini, E., Bishop, D., Marcora, S. M., Bravo, D. F., Sassi, R., & Impellizzeri, F. M. (2007). Validity of simple field tests as indicators of match-related physical performance in top-level professional soccer players. International Journal of Sports Medicine, 28(3), 228-235. Ramsbottom, R., Brewer, J., & Williams, C. (1988). A progressive shuttle run test to estimate maximal oxygen uptake. British Journal of Sports Medicine, 22(4), 141144. Rhea, M. R. (2004). Determining the magnitude of treatment effects in strength training research through the use of the effect size. Journal of Strength & Conditioning Research, 18, 918-920. Rud, B., Foss, Ø., Krustrup, P., Secher, N. H., & Hallen, J. (2012). One-legged endurance training: Leg blood flow and oxygen extraction during cycling exercise. Acta Physiologica, 205(1), 177-185. Saltin, B., Nazar, K., Costill, D. L., Stein, E., Jansson, E., Essén, B., & Gollnick, P. D. (1976). The nature of the training response; peripheral and central adaptations to one-legged exercise. Acta Physiologica, 96(3), 289-305. Spiteri, T., Cochrane, J. L., Hart, N. H., Haff, G. G., & Nimphius, S. (2013). Effect of strength on plant foot kinetics and kinematics during a change of direction task. European Journal of Sport Science, 13(6), 646-652. Stojanović, E., Stojiljković, N., Scanlan, A. T., Dalbo, V. J., Berkelmans, D. M., & Milanović, Z. (2017). The activity demands and physiological responses encountered during basketball match-play: A systematic review. Sports Medicine, 48(1), 111-135. Stone, N. M., & Kilding, A. E. (2009). Aerobic conditioning for team sport athletes. Sports Medicine, 39(8), 615-642. Tanaka, H. (1994). Effects of cross-training. Sports Medicine, 18(5), 330-339. Thomas, C., Sirvent, P., Perrey, S., Raynaud, E., & Mercier, J. (2004). Relationships between maximal muscle oxidative capacity and blood lactate removal after supramaximal exercise and fatigue indexes in humans. Journal of Applied Physiology, 97(6), 2132-2138. 79 Trewin, J., Meylan, C., Varley, M. C., & Cronin, J. (2017). The influence of situational and environmental factors on match-running in soccer: A systematic review. Science and Medicine in Football, 1(2), 183-194. Vincent, G., Lamon, S., Gant, N., Vincent, P., MacDonald, J., Markworth, J., … Hickey, A. (2015). Changes in mitochondrial function and mitochondria associated protein expression in response to 2-weeks of high intensity interval training. Frontiers in Physiology, 6(51), 1-8. Vinson, D., Gerrett, N., & James, D. V. (2018). Influences of playing position and quality of opposition on standardized relative distance covered in domestic women’s field hockey: Implications for coaches. Journal of Strength & Conditioning Research, 32(6), 1770-1777. Young, W. B., Hepner, J., & Robbins, D. W. (2012). Movement demands in Australian rules football as indicators of muscle damage. Journal of Strength & Conditioning Research, 26(2), 492-496. CHAPTER 4 INTERLIMB AND INTRALIMB COMPENSATORY PATTERNS AFTER ANTERIOR CRUCIATE LIGAMENT RECONSTRUCTION Introduction Nearly 25% of the athletes who return to participation (RTP) after anterior cruciate ligament (ACL) reconstruction (ACLR) will suffer a secondary ACL injury to either the ipsilateral (i.e., surgical) or contralateral (i.e., nonsurgical) side (Wiggins et al., 2016). Paterno and colleagues (2014) indicated that athletes returning to cutting and pivoting sports (i.e., high-risk sports) are ~5.7 times more likely to suffer another ACL injury compared to athletes without a history of ACL rupture. Secondary ACL injury may be related to functional impairments at the time of RTP (Paterno et al., 2010). Functional impairments are typically identified using whole-leg tasks involving coordinated maximal actions of the hip, knee, and ankle joints. Numerous functional assessments have been proposed (see Abrams et al., 2014, for a review), and the singleleg hop series (single-leg hop, single-leg triple hop, single-leg triple crossover hop, single-leg 6 m hop for time) is the most popular field-testing battery used during later stages of rehabilitation (Abrams et al., 2014). In these tests, the surgical and nonsurgical limbs are compared by calculating the limb symmetry index (Barber, Noyes, Mangine, & Hartman, 1990; Palmieri-Smith & Lepley, 2015) or asymmetry (Impellizzeri, Rampinini, Maffiuletti, & Marcora; 2007). Despite using symmetry measures to aid decisions to 81 RTP, large deficits in whole-leg dynamic function within the surgical leg (~79-92% of nonsurgical limb) persist among many athletes at the time of RTP (Myer et al., 2011; Palmieri-Smith & Lepley, 2015; Schmitt, Paterno, & Hewett, 2012). Maximal efforts are required for accurate comparisons of whole-leg functional performance using the standard evaluation techniques described above. However, jumping and change-of-direction exercises are not typically introduced until approximately 8–12 weeks after surgery (Adams, Logerstedt, Hunter-Giordano, Axe, & Snyder-Mackler, 2012; van Grinsven, van Cingel, Holla, & van Loon, 2010), and maximal efforts begin after an additional 2–3 months. Consequently, early rehabilitation phases proceed without knowledge of asymmetry during whole-leg dynamic function, which may be due to protective compensatory patterns. Compensatory strategies that develop during early stages of rehabilitation may be more difficult to correct during later stages (Grooms, Appelbaum, & Onate, 2015; Kapreli & Athanasopoulos, 2006; Nyland, Wera, Klein, & Caborn, 2014). Thus, asymmetries at the time of RTP may be due to the development of early post-surgical compensatory behaviors which persist throughout the rehabilitation process. Alternatively, assessing whole-leg dynamic function soon after surgery could facilitate identification and correction of compensatory behaviors in order to produce more successful outcomes. High forces associated with maximal whole-leg actions are inappropriate immediately after surgery, whereas submaximal activities could provide helpful metrics. Walking and submaximal cycling are two exercise modes indicated by surgeons as early as 2–3 weeks post-ACLR (van Grinsven et al., 2010). Walking elicits peak ground reaction forces equivalent to ~1.15–1.25 times bodyweight (Keller et al., 1996). 82 Likewise, submaximal cycling (~120–400 W) elicits peak pedal reaction forces (Hoes, Binkhorst, Smeekes-Kuyl, & Vissers, 1968) that can range from ~0.25–0.50 times bodyweight at 120 W to ~0.75–1.35 times bodyweight at 400 W depending on an individual’s bodyweight. Furthermore, peak knee extension moments during walking (Devita, Hortobagyi, & Barrier, 1998) are ~60% greater than the peak knee extension moments during cycling at 175 W (Hunt, Sanderson, Moffet, & Inglis, 2003). Cycling also elicits significantly less strain on the ACL compared to walking and other weightbearing and nonweight-bearing tasks (Escamilla, MacLeod, Wilk, Paulos, & Andrews, 2012). Therefore, submaximal cycling may be a more conservative strategy than walking to assess whole-leg function during early stages of rehabilitation after ACLR. Finally, Devita and colleagues (1998) reported altered walking gait kinematics after surgery whereas kinematics are generally constrained during cycling in ACL deficient participants (Hunt et al., 2003). As such, submaximal cycling could uniquely detect differences in coordination because the central nervous system can freely alter the remaining degrees of freedom in the two-leg system (i.e., the ankle, knee, and hip actions) to perform the locomotive task (Raasch & Zajac, 1999). Biomechanical analyses of joint moments and power during submaximal cycling could reveal interlimb and intralimb compensatory behavior. Indeed, previous authors have reported interlimb (Hunt, Sanderson, Moffet, & Inglis, 2004) and intralimb (Hunt et al., 2003) compensatory patterns among ACL deficient participants, but this has never been studied among those who have undergone ACLR. Early detection of possible biomechanical compensatory strategies during whole-leg dynamic action could help therapists advance their current rehabilitation protocols while improving outcomes, 83 especially among athletes who are at greater risk of secondary injury (Paterno, Rauh, Schmitt, Ford, & Hewett, 2014; Wiggins et al., 2016). Therefore, the purpose of this exploratory pilot study was to identify whole-leg and joint-specific compensatory patterns using inverse dynamics during isokinetic cycling after ACLR. In this study, I hypothesized that during submaximal cycling, (1) athletes who have not RTP after ACLR would exhibit compensatory behavior to protect the surgical knee (i.e., reduced knee extension power). Specifically, I expected there to be either an interlimb compensatory behavior characterized by whole-limb attenuation of power on the surgical side (Hunt et al., 2004; Rimer, Elmer, & Martin, 2014), or an intralimb compensatory pattern characterized by a lower relative contribution of the surgical knee extension (KE) action to the total work produced by the surgical leg, which would be compensated by greater relative contributions from other joint-specific actions within the same leg (Rimer et al., 2014). Additionally, based on previous reports of persistent quadriceps activation inhibition ~1–5 years after ACLR (Otzel, Chow, & Tillman, 2015; Urbach, Nebelung, Becker, & Awiszus, 2001), I expected that (2) the above described interlimb or intralimb compensatory patterns would be present in athletes who had RTP from ACLR. Methods Overview All methods used in this exploratory pilot study were approved by the University of Utah Institutional Review Board. The forthcoming description of the methods are illustrated by a schematic overview (Figure 4.1). Briefly, I used cycling biomechanics to identify the presence of interlimb or intralimb compensatory among athletes within ~2.5 84 Report to Lab ND CON Limb Control Group (n=12) Consent, Info., Meas. D CON Limb ND CON Limb Contribution Joint-Specific Contributions t test t test D CON Limb Contribution Joint-Specific Contributions WarmUp 240 W x2 InterLimb? Yes IntraLimb? Yes No Nonlinear Regression 120 W x2 180 W x2 Limb-Specific Standardized Differences (z) Process Kinematic & Kinetic Data For each Intensity ACLR Limb Before RTP (n=15) NS Limb 400 W x2 ACLR Limb After RTP (n=5) NS Limb ND ACLR Limb Contribution Joint-Specific Contributions D ACLR Limb Contribution Joint-Specific Contributions t test t test ND NS Limb Contribution Joint-Specific Contributions D NS Limb Contribution Joint-Specific Contributions ND ACLR Limb Contribution Joint-Specific Contributions D ACLR Limb Contribution Joint-Specific Contributions t test t test ND NS Limb Contribution Joint-Specific Contributions D NS Limb Contribution Joint-Specific Contributions Figure 4.1. A schematic overview of the study protocol. Briefly, participants reported to the laboratory and provided informed consent before collecting background information and taking anthropometric measurements. They warmed up and then performed trials at up to four cycling intensities. Kinematic and kinetic data were processed, and then partitioned for control group participants, anterior cruciate ligament reconstruction (ACLR) participants before return to participation (RTP), and after RTP. For the control group, data was processed and summarized for the nondominant (ND CON) and dominant (D CON) limbs, whereas ACLR participant data was processed and summarized for the ACLR and nonsurgical (NS) limbs. The relative whole-leg contributions to the net double-leg power and the relative joint-specific contributions to the total whole-leg mechanical work were compared between the ND CON and D CON limbs of the control group and between the ACLR and NS Limbs of the ACL groups using independent t tests (dashed arrows). The relative whole-leg and joint-specific contributions of nondominant and dominant ACLR limbs (ND ACLR; D ACLR) and nonsurgical limbs (ND NS; D NS) were transformed to limb-specific standardized differences using the mean and standard deviation of the same limbs within the control group (dashed arrows), and summarized for each variable. The proportion of participants who demonstrated either an interlimb or intralimb compensatory pattern was determined. Finally, nonlinear regression techniques were used to assess if the whole-leg and jointspecific limb-specific standardized differences between individuals improved from before to after RTP. 85 years of ACLR (i.e., the ACL group) compared to competitive athletes without a history of ACL injury (i.e., the control group). The relative contribution that each limb made to the net double-leg power, and the relative contribution that each joint-specific action made to the whole-leg work were compared between the surgical and nonsurgical limbs within the ACL group and between the nondominant and dominant limbs within the control group. Differences between groups were compared by interpreting the magnitude of the limb-specific standardized difference (defined below) between the surgical limb of the ACL group participants and the same limb (i.e., nondominant or dominant) of the control group participants. The limb-specific standardized differences were used to calculate the proportion of trials that had either an interlimb or intralimb compensatory pattern, and they were also used to explore compensatory patterns within subsets of the data (i.e., sex, side, and graft type). Finally, regression techniques were used to interpret whether or not deficits persisted between individuals who participated before or after RTP. Participants Each participant provided written informed consent before volunteering for this study. All 32 participants (Table 4.1) were involved in high risk activities, which were preferred because they pose greater risk of secondary ACL injury (Paterno et al., 2014). Overall, they had extensive experience within their sport and they had significant strength and conditioning training backgrounds. Also, to control for limb dominance, identified as the preferred kicking leg (Carpes, Mota, & Faria; 2010), only right-legged kickers were used because very few adults (~5–8% of the population) prefer to use their left leg (Gabbard & Iteya, 1996). 86 Table 4.1. Participant background data Group Age (years) Height (cm) Weight (kg) Male (n) Female (n) Sport (years) Training (years) CON 22.4 (1.9) 178 (7) 76.9 (8.3) 10 2 13 (3) 9 (4) ACL 23.4 (6.2) 174 (10) 71.5 (12.9) 9 12 14 (5) 8 (5) Note. All values are expressed as the group mean (SD). Sport refers to the years of experience that individuals within the control (CON) and ACL groups had in their chosen sport or recreational activity, and training refers to their experience (years) with strength & conditioning training activities. Participants in the control group (n = 12) were competitive athletes who had never torn an ACL. They were free of injury and had not suffered another lower extremity injury that caused a time-loss (> 1 week) from their sport within 2 years from the day of their visit. Participants in the ACL group (n = 20) had a variety of sport and recreational activity backgrounds (Table 4.2). Individuals who participated before RTP (n = 15) received clearance from their sports medicine team to perform submaximal cycling— threrapists usually attended the laboratory with the participants—and their single laboratory visit occurred at different times ranging from 17–215 days post-ACLR (i.e., ≤ 7 months). Those who participated after unrestricted RTP (n = 5) visited the laboratory from 468–890 days (i.e., ~1.3–2.5 years) post-ACLR. As this was an exploratory pilot study, I did not delimit the ACL group based on sex, side, graft type, or concomitant injuries. In total, there were 11 female and nine male ACL participants. Eight had ACLR on their nondominant limb and 12 had ACLR on their dominant limb, including 11 who had a had a bone-patellar tendon-bone autograft, eight who had a hamstring tendon autograft, and one had a cadaver allograft. Several ACL participants suffered concomitant injuries and two visited the laboratory after their second ACLR on the same leg. It should be noted that ACLR participants were only 87 Table 4.2. ACL participant information SUB Sex Sport Side Type post-ACLR (d) Concomitant # 1 Level of Participation RTPP (d) RTP (d) M Basketball Recreational R BPTB 17 Meniscus (NR) 2* F Ski Jumping World Cup L HT 39 Meniscus (R) L HT 72 Meniscus (R) 3 F Gymnastics NCAA DI L HT 44 None 4 F Aerials World Cup R HT 50 Meniscus (NR) bone bruise – – – – – – – – – – 5 M Football NCAA DI R BPTB 53 Meniscus (R) 6 M Lacrosse DI Club L BPTB 70 Meniscus (R) 7 M Aerials World Cup R BPTB 77 Meniscus (NR) Bone bruise Gastrocnemius strain – – – – – – 8 F Basketball NCAA DI R BPTB 89 Meniscus (NR) 9* F Alpine Continental R BPTB 97 Meniscus (R) Cup R BPTB 141 Meniscus (R) 10 F Soccer NCAA DI L BPTB 104 None 11 M Alpine World Cup R HT 113 Meniscus (NR) 12 F Basketball NCAA DI R HT 139 Meniscus (R) 13 M Climbing Recreational R CA 188 None 14 F Soccer NCAA DII R BPTB 207 2nd ACLR Meniscus (R) Bone bruise – – – – – – – – – – – – – – – – 15 M Football NCAA DI L BPTB 215 None 174 – 16 F Soccer NCAA DI L HT 468 None 269 294 17 M Alpine World Cup R HT 633 Meniscus (NR) MCL (NR) 216 348 18 F Soccer NCAA DI L BPTB 638 None 288 309 19 F Soccer NCAA DI L BPTB 784 None 300 321 20 M Lacrosse DI Club R HT 890 Meniscus (NR) MCL (NR) 229 406 Note. Participants are listed in order of days post-ACLR (anterior cruciate ligament reconstruction). ACLR of the left (L) and right (R) sides correspond with the nondominant and dominant limbs, respectively. NCAA = National Collegiate Athletics Association; DI = Division I; DII = Division II; BPTB = bone-patellar tendon-bone autograph; HT = hamstring tendon autograph; CA = cadaver allograft; NR = not repaired; R = repaired; MCL= medial collateral ligament; RTPP (d) = days post-ACLR when returned to participation progression; RTP (d) = days post-ACLR when returned to full participation. *participants who had two laboratory visits 88 accepted if they planned to or had already returned to the same level of participation in their given activity or sport. Experimental protocol Background information and anthropometric measurements. The protocol required participants to come to the Neuromuscular Function Laboratory for a single 1hour visit. Upon arrival, the procedures were explained verbally and participants provided written informed consent. Individual background information (date of birth, height, bodyweight, sex, sport or activity, level of participation, playing experience, and strength and conditioning training experience) and injury history (date of surgery, type of graft, concomitant injuries, and RTP status) was recorded. After measuring their foot length (heel to toe), participants were fitted for cycling shoes, seat height, and handlebar position on an isokinetic cycle ergometer according to their personal preference. They sat on the cycle ergometer (described below) in a stationary position in order to measure kinematic foot length (pedal spindle to lateral malleolus), leg length (lateral femoral condyle to lateral malleolus), and thigh length (greater trochanter to lateral femoral condyle). Cycling trials. To warm up, participants practiced using feedback provided by a power meter to stabilize at specific submaximal work rates. They started by practicing stabilizing at 120 W 3–4 times. As long as they felt comfortable to proceed, they performed 2–3 rehearsals at each of three progressively greater work rates: 180 W, 240 W, and 400 W. A faster cadence (90 rpm) was used at all cycling intensities because in order to minimize strain on the ACL compared to slower cadences (e.g., 60 rpm; Fleming et al., 1998). After a 1–2 min rest, the experimental trials began. Participants performed 89 two brief (6 s) submaximal isokinetic cycling trials at 120 W, and were allowed to perform two more trials at 180 W, 240 W, and 400 W as long as they did not experience discomfort during the warm-up. No participants experienced complications as they were free to progress at their own discretion or stop any time. Kinematic and kinetic data (described below) were recorded during each 6 s trial and used for biomechanical analysis. Cycle ergometer. All trials were performed on an isokinetic cycle ergometer which was described previously (Martin & Brown, 2009; Elmer, Barratt, Korff, & Martin, 2011). Briefly, the flywheel of a cycle ergometer (CycleOps, Madison, WI) was driven via pulleys and a belt with a 3750-W direct-current motor (model CDP3605; Baldor Electric Company, Fort Smith, AR) operated by a speed controller equipped with regenerative braking capability (model RG5500U; Minarik, Glendale, CA). When participants applied power to the ergometer, the motor acted as a generator and the resulting current was dissipated by a resistor and heat sink built into the speed controller. The controller could, therefore, maintain a specific pedaling rate while resisting up to 3750 W. The ergometer was equipped with a power meter (Schoberer Rad Messtechnik, Ju ̈lich, Germany) capable of providing valid and reliable measures (Gardner et al., 2004; Martin, Milliken, Cobb, McFadden, & Coggan, 1998). Each pedal was equipped with two three-component piezoelectric force transducers (Kistler 9251; Kistler USA, Amherst, NY) and digital position encoders (US Digital model S5S-1024; Vancouver, WA). Participants attached their feet to the pedals using cleated pedal interfaces (Speedplay Inc., San Diego, CA) and cycling shoes (Specialized Bicycle Components Inc., Morgan Hill, GA, USA). 90 Kinematic and kinetic data. Two-dimensional kinematic and kinetic data were obtained using methods first introduced by Martin and colleagues (2007). Briefly, pedal forces, pedal and crank positions, and the position of an instrumented spatial linkage system (ISL) was recorded at 120 Hz (Bioware 3.0; Kistler USA). Normal and tangential pedal forces, pedal position, crank position, and ISL position data were filtered using a fourth-order zero-lag low-pass Butterworth filter with a cutoff frequency of 8 Hz. On each side, pedal power was calculated as the dot product of pedal force and linear pedal velocity. Positions of the left and right greater trochanter and iliac crest were determined by collecting a static trial of each participant attached to the ISL, and the relative positions were assumed to remain constant (Neptune & Hull, 1995). During the cycling trials, iliac crest, pedal, and crank position coordinates of each side were recorded to allow geometrical determination of sagittal plane limb segment positions using the law of cosines. Position-time data were fit with a quadratic spline which allowed linear and angular velocities to be determined by differentiation of the fitted spline functions. Segmental masses, moments of inertia, and location of centers of mass were estimated using previously reported regression equations (de Leva, 1996). Sagittal plane joint reaction forces and net joint moments at the ankle, knee, and hip were determined using inverse dynamics (Elftman, 1939). Ankle, knee, and hip joint-specific powers were calculated as the product of net joint moments and joint angular velocities. Power transferred across the hip joint was calculated as the dot product of the hip joint reaction force and linear velocity. Joint-specific powers were averaged over all of the complete pedal cycles within each measurement interval (i.e., ~9 cycles in 6 s). The relative contribution of each joint- 91 specific action to the total mechanical work performed by each leg was calculated (McDaniel, Behjani, Elmer, Brown, & Martin, 2014; Rimer et al., 2014; Rimer, Marshall, Wehmanen, Farley, & Martin, 2012). Previous investigators (Broker & Gregor, 1994; Ericson, 1998; Elmer et al., 2011; Martin & Brown, 2009; van Ingen Schenau, van Woensel, Boots, Snackers, & De Groot, 1990) generally agree that AE, KF, KE, and HE are the four main actions contributing to total mechanical work during submaximal cycling. After ACLR, however, compensatory patterns may also manifest within the ankle flexion (AF; dorsiflexion) and hip flexion (HF) actions. Thus, this study evaluated potential asymmetries and compensatory patterns associated with all six joint-specific actions (AE, KE, HE, AF, KF, HF) within each leg. Data analysis All background and experimental data were summarized (mean [SD]). Kinematic and kinetic data from each trial were processed as described above, and then averaged across both trials at each prescribed work rate. The following variables were calculated and used for analysis: the relative contributions (%) of each limb to the net double-leg power, and the relative contribution (%) of the AE, KE, HE, AF, KF, HF actions to the total mechanical work performed by each limb—for brevity, whole-leg and joint-specific actions will be referred to as the relative contribution from now on. These data were summarized for the nondominant and dominant limbs within the control group and for the surgical and nonsurgical limbs within the ACL group. To account for the magnitude of any potential differences between ACL group participants and the control group, the relative whole-leg and joint-specific contributions of the surgical and nonsurgical limbs of each individual in the ACL group were 92 transformed to standardized values (z) of the same limb within control group. For example, if the surgical side was the dominant limb, then the whole-leg and joint-specific contributions were expressed (in units of SD) relative to the mean and standard deviation of the parameters within the dominant limb of the control group, and vice-versa for the nondominant limb. The benefit in using limb-specific standardized differences (as they will be referred to from here forward) is that they facilitate interpretation of the magnitude of difference between measures within the ACL group and the control group before and after RTP, while accounting for potential asymmetries that may exist between limbs of the control group. The practical benefit of using limb-specific standardized differences is that they facilitate the opportunity to give objective feedback to individual participants. To acknowledge potential differences between limbs within participants in the control group, the relative whole-leg and joint-specific contributions of their dominant and nondominant limbs were compared using independent t tests, rather than dependent t tests because previous authors have reported laterality during cycling (Carpes et al., 2010, Hunt et al., 2004). Accordingly, some minor degree of asymmetry was expected for the relative contributions of the whole-leg action. However, no assumptions were made about potential asymmetries between the relative contributions of the nondominant and dominant limb joint actions because no normative data has been published. To examine my first hypothesis, that athletes undergoing rehabilitation from ACLR would have either an interlimb or intralimb compensatory pattern, only participants before RTP were used. First, within-group comparisons were made between the relative contributions of whole-leg and joint-specific actions of the surgical and 93 nonsurgical limbs using independent t tests. To make comparisons to the control group, the magnitudes of the limb-specific standardized differences of the relative whole-leg and joint-specific contributions were summarized and interpreted; a limb-specific standard difference of 1 SD was assumed to be practically meaningful. Finally, the proportion of individuals with a protective compensatory pattern was determined from those who demonstrated either whole-leg attenuation of power (≤ -1 SD) or a reduced relative surgical KE contribution (≤ -1 SD). My second hypothesis was that the above-described interlimb and intralimb compensatory patterns would also be demonstrated by participants who were assessed after RTP. Similar to what was described above, within-group (independent t test) and between-group (limb-specific standardized differences) comparisons were made among those who had RTP. Additionally, to facilitate visual interpretation of potential trends between ACL participants from before to after RTP, the limb-specific standardized differences of the relative whole-leg and joint-specific contributions (KE and HE) from all ACLR limb trials were graphed over days since ACLR, with the central tendency of the control group (mean ± 1 SD) superimposed. Nonlinear regression techniques were used to determine the time-constant associated with the exponential function that would minimize the sum of squared error terms between the actual and model terms. This method accounted for the fact that participants tested at different time points after ACLR, and the corresponding coefficient of determination (R2) between the limb-specific standardized differences of each variable and days post-ACLR were used to interpret each regression model. Beyond my hypotheses, this study was not delimited by potentially confounding 94 factors. As this was the first study to investigate interlimb and intralimb compensatory patterns during submaximal cycling after ACLR, a diverse pilot dataset was desired so that it could be explored. Subsets of data were partitioned based on the presence of specific factors (sex, limb dominance, and graft type). For each subset, the mean limbspecific standardized difference was calculated and the proportion of participants who had a limb-specific standardized difference was determined. Results Overall, 20 participants completed trials at 120 W, 180 W, and 240 W, and only 18 participants completed trials at 400 W. It should also be noted that two ACL participants (SUBs #2 and #9 in Table 4.2) tested on two separate occasions and were treated independently in the regression analyses. Control group Within the control group, the relative contributions of the nondominant and dominant limbs to the net double-leg power (Table 4.3) were not statistically different at 120 W (p = .44), 180 W (p = .14), or 400 W (p = .08), but there was a significant difference at 240 W (p = .001; Table 4.3; Figure 4.2). The control group exhibited large variation in the relative contribution of each joint-specific action (AE, KE, HE, AF, KF, HF) to the whole-leg work (Table 4.4). For the most part, differences between the relative contributions of the nondominant and dominant limb joint-specific actions were nonsignificant across all work rates. However, there were a few notable trends. Specifically, the relative contributions of the KE and HF actions were consistently greater (p > .05) within the nondominant limb compared to the dominant limb (Table 4.4; Figure 4.2). In contrast, the relative contributions of the HE and KF actions were consistently 95 Table 4.3. Relative contribution (%) of each leg to net double-leg power Control Group ACL Group - Before RTP ACL Group - After RTP Power ND D ACLR NS ACLR NS 120 W 50.5 (3.0) 49.5 (3.0) 47.5 (6.1)* -0.5 (2.1) 52.5 (6.1) 0.5 (2.1) 50.4 (2.5) 0.2 (1.0) 49.2 (3.3) -0.2 (1.0) 180 W 49.2 (2.5) 50.8 (2.5) 46.8 (6.2) * -1.3 (2.5) 53.2 (6.2) 1.3 (2.5) 50.8 (2.1) 0.4 (1.2) 49.2 (2.1) -0.4 (1.2) 240 W 48.7 (1.7)ϟ 51.3 (1.7) 46.6 (5.9) * -2.1 (3.5) 53.4 (5.9) 2.1 (3.5) 50.7 (1.0) 0.6 (1.3) 49.3 (1.0) -0.6 (1.3) 400 W 49.5 (1.4) 50.5 (1.4) 46.4 (4.6) * -2.8 (3.4) 53.6 (4.6) 2.8 (3.4) 49.8 (1.1) -0.1 (0.9) 50.2 (1.1) 0.1 (0.9) Note. Data is summarized by the mean (SD) for the nondominant (ND) and dominant (D) limbs of the control group, and for the surgical (ACLR) and nonsurgical (NS) limbs of participants in the ACL group who had not returned to play (Before RTP) and those who did (After RTP). Italicized values indicate the limb-specific standardized difference (in units of SD) between the control group and the ACL group (expressed as mean [SD]), whereas values in bold represent instances when the mean standardized differences were greater than 1 SD. * p ≤ .05 within each group. *p ≤ .01 within each group. 96 A 600 ND CON D CON ND CON ±1SD D CON ±1SD Pedal Power (W) 500 400 300 200 100 0 -100 -200 Relative Ankle Power (%) B 0 90 180 Pedal Angle (º) 270 360 0 90 180 Pedal Angle (º) 270 360 0 90 180 Pedal Angle (º) 270 360 0 90 180 Pedal Angle (º) 270 360 300 250 200 150 100 50 0 -50 -100 Relative Knee Power (%) C 300 250 200 150 100 50 0 -50 -100 Relative Hip Power (%) D 300 250 200 150 100 50 0 -50 -100 Figure 4.2. Submaximal cycling biomechanics of the control group at 240 W. The mean pedal power (A), ankle power (B), knee power (C), and hip power (D) are displayed across the pedal cycle (0–360°) for the nondominant (CON-ND) and dominant limbs (CON-D) of the control group. Joint-specific power data (B–D) are expressed relative (%) to the average net double-leg cycling power across the pedal cycle. The typical ranges (mean ± SD) for each leg of the control group participants are also displayed (ND CON ±1SD; ND CON ±1SD). Crank angles of 0° and 360° indicate when the pedal was at top dead center, and 180° indicates when it was at bottom dead center. As such, leg extension and flexion generally occur from 0–180° and from 180–360°, respectively. Within the control group, the dominant limb typically produced slightly greater pedal power, relative knee flexion power, and relative hip extension power. In contrast, it was typical for the nondominant limb to produce slightly greater relative knee extension power and relative hip flexion power. 97 Table 4.4. Relative contribution of joint-specific actions to whole-leg work Control Group Nondominant Limb Joint-Specific Actions Power AE Dominant Limb Joint-Specific Actions KE HE AF KF HF AE KE HE AF KF HF 120 W 19 (10) 57 (21) 14 (15) -8 (5) 7 (9)* 8 (11) 19 (11) 47 (25) 24 (19) -10 (7) 16 (10) 1 (16) 180 W 14 (8) 53 (12) 16 (11) -4 (2) 13 (8) 5 (6) 14 (8) 46 (12) 24 (9) -5 (4) 17 (8) 1 (11) 240 W 12 (6) 53 (12) 14 (12)* -2 (2) 11 (8) 8 (9) 12 (5) 45 (11) 25 (7) -3 (2) 15 (7) 2 (8) 400 W 12 (6) 49 (9) 19 (7)* -2 (1) 13 (6) 2 (4) 12 (5) 43 (8) 26 (6) -2 (1) 15 (7) 1 (5) ACL Group - Before RTP Surgical Limb Joint-Specific Actions Power AE KE HE AF KF Nonsurgical Limb Joint-Specific Actions HF AE KE HE AF KF HF 120 W 13 (6) 46 (20) 29 (19) -6 (4) 8 (15) 6 (13) -0.6 (0.6) -0.2 (0.8) 0.5 (1.1) 0.5 (0.7) 0.5 (1.5) 0.1 (0.9) 13 (6) 55 (12) 20 (12) -6 (4) 3 (15) 11 (10) 0.6 (0.6) 0.1 (0.5) 0.2 (0.7) 0.5 (0.9) 0.8 (1.6) 0.4 (0.9) 180 W 12 (6) 38 (19) 33 (16)* -4 (3) 11 (12) 5 (9) -0.2 (0.8) -0.8 (1.5) 1.2 (1.7) 0.2 (1.1) 0.5 (1.4) 0.3 (0.9) 11 (5) 48 (11) 23 (11) -3 (3) 5 (11) 12 (9) 0.3 (0.7) 0.2 (0.8) 0.4 (1.0) 0.4 (1.2) 1.2 (1.4) 1.1 (1.3) 240 W 13 (5) 35 (18)* 33 (15)* -3 (2) 12 (11) 6 (6) 0.1 (0.9) -1.1 (1.5) 1.4 (2.3) 0.2 (1.2) 0.2 (1.4) 0.2 (0.8) 11 (5) 47 (11) 24 (9) -2 (2) 7 (9) 10 (8) 0.2 (0.9) 0.3 (0.9) 0.6 (0.9) 0.1 (1.2) 0.6 (1.2) 0.4 (0.8) 400 W 14 (5) 34 (13) 32 (10) -2 (2) 12 (8) 5 (5) 0.3 (0.9) -1.3 (1.7) 1.3 (1.7) 0.4 (1.4) 0.5 (1.3) 0.7 (1.0) 13 (5) 42 (10) 28 (8) -2 (1) 10 (6) 4 (4) 0.0 (1.0) 0.6 (1.1) 1.0 (1.2) 0.5 (1.2) 0.7 (1.1) 0.6 (0.8) ACL Group - After RTP Surgical Limb Joint-Specific Actions Power AE KE HE AF KF Nonsurgical Limb Joint-Specific Actions HF AE KE HE AF KF HF 120 W 17 (8) 62 (14) 19 (6) -8 (6) 12 (8) -6 (6) 0.2 (0.8) 0.4 (0.6) 0.1 (0.5) 0.3 (1.1) 0.1 (0.8) 0.9 (0.2) 16 (4) 64 (11) 11 (6) -7 (3) 9 (15) 2 (13) 0.2 (0.4) 0.5 (0.5) 0.5 (0.3) 0.2 (0.6) 0.3 (1.5) 0.1 (0.9) 180 W 15 (5) 56 (17) 22 (5)* -4 (3) 11 (11) -4 (3) 0.1 (0.7) 0.5 (1.4) 0.3 (0.3) 0.3 (1.0) 0.4 (1.6) 1.0 (0.4) 13 (4) 60 (11) 12 (5) -4 (2) 9 (14) 4 (10) 0.1 (0.5) 0.9 (0.9) 0.8 (0.7) 0.2 (0.5) 0.7 (1.7) 0.1 (1.1) 240 W 14 (4) 53 (13) 23 (2) -2 (2) 9 (9) -1 (2) 0.4 (0.8) 0.3 (1.3) 0.3 (0.7) 0.0 (0.9) 0.5 (1.4) 0.8 (0.3) 12 (2) 57 (6) 17 (5) -3 (1) 9 (10) 2 (4) 0.1 (0.4) 0.8 (0.5) 0.6 (0.9) 0.2 (0.7) 0.5 (1.3) 0.2 (0.7) 400 W 15 (2) 45 (9) 29 (5) -2 (1) 7 (8) -1 (2) 0.6 (0.4) 0.1 (1.4) 1.1 (1.0) 0.3 (0.7) 1.1 (1.3) 0.6 (0.3) 13 (3) 46 (8) 26 (8) -2 (1) 10 (8) 2 (4) 0.0 (0.6) 0.0 (0.6) 0.5 (0.9) 0.4 (0.8) -0.8 (1.2) 0.2 (0.9) Note. The relative contribution (%) of each joint-specific action to the total mechanical work performed by each leg are displayed (mean [SD]) for the control group and ACL participants who had not returned to play (ACL Group - Before RTP), and those who did (ACL Group - After RTP). In the ACL group sections, italicized values indicate the limbspecific standardized difference (in units of SD) between limbs of the control group and ACL groups (expressed as mean [SD]), whereas values in bold represent instances when the mean standardized differences of the ACL group were 1 SD beyond the mean of the control group. ACL = anterior cruciate ligament; AE = ankle extension; KE = knee extension; HE = hip extension; AF = ankle flexion; KF = knee flexions; AF = ankle flexion. * p ≤ .05 within each group. 98 greater within the dominant limb compared to the nondominant limb (Table 4.4, Figure 4.2), whereas the relative contribution of the HE contributions was significantly greater at 240 W (p = .02) and 400 W (p = .02), and the relative contribution of the KF action was significantly greater at 120 W ( p = .03; Table 4.4). It should be noted that the standardized differences between the relative whole-leg and joint-specific contributions of the nondominant and dominant limbs were within ±1 SD in all cases except those that were significantly different (mentioned above), and this observation supported using a limb-specific standardized difference of 1 SD as an indicator for whole-leg and jointspecific differences between the ACL and control groups. Before RTP In consideration of my first hypothesis, the surgical limb among individuals before RTP made significantly lower relative contributions than the nonsurgical limb at each cycling intensity (120 W: p = .02; ,180 W: p = .005; 240 W p = .03; 400 W: p = .0006; Table 4.3). Furthermore, the mean limb-specific standardized differences gradually increased in magnitude with increasing intensity from 120 W to 400 W (Table 4.3). These data indicate that there was whole-leg attenuation of power within the surgical limb before RTP, and that the deficits increased with greater work rates. Before RTP, the limb-specific standardized difference for the relative contribution of surgical leg was ≤ -1 SD among 47% of the participants (7 out of 15) at 120 W, 53% (8 out of 15) at 180 W and 240 W, and 69% of the participants who were able to test (9 out of 13) at 400 W (Figure 4.3). Notably, both participants who opted-out of 400 W demonstrated a protective compensatory pattern at each of the lower work rates. Therefore, there would have likely been a greater relative frequency (73%) at 400 W had 99 % Net Double-Leg Power (z) A 4 3 2 1 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 R2=0.32 ND ACLR Control (mean±SD) 0 % Net Double-Leg Power (z) B % Net Double-Leg Power (z) C % Net Double-Leg Power (z) 180 270 360 450 540 Days After ACLR 630 720 810 900 R2=0.37 0 90 180 270 4 3 2 1 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 360 450 540 Days After ACLR 630 720 810 900 R2=0.30 0 D 90 4 3 2 1 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 D ACLR Model Fit 90 180 270 4 3 2 1 0 -1 -2 -3 -4 -5 -6 -7 -8 -9 360 450 540 Days After ACLR 630 720 810 900 R2=0.23 0 90 180 270 360 450 540 Days After ACLR 630 720 810 900 Figure 4.3. Between-subject data for the relative contribution (%) of the surgical leg to the net double-leg power. Individual data are graphed over days since anterior cruciate ligament reconstruction (ACLR) at 120 W (A), 180 W (B), 240 W (C), and 400 W (D). Data for individuals with a nondominant side ACLR (ND ACLR) are illustrated with open circles (⚬), and those with a dominant side ACLR (D ACLR) are illustrated using closed circles (●). All data are expressed as limb-specific standardized differences, that is, data for participants with an ND ACLR are expressed as a standardized value (in units of SD) of the dominant limb within the control group, and vice-versa. The nonlinear function that best fit the data (model fit) is illustrated with a dotted line (⋯) and the coefficient of variation (R2) is displayed. 100 they tested. Intralimb compensatory patterns were also present. The relative contribution of the surgical KE action was less than that within the nonsurgical leg at each cycling intensity (120 W: p = .12; 180 W: p = .06; 240 W: p = .04; 400 W: p = .10), and the magnitudes of the mean limb-specific standardized differences systematically increased with work rate and were ≤ -1 SD at 240 and 400 W (Table 4.4). Compared to the number of participants who had whole-leg attenuation of power, fewer demonstrated a relative surgical leg KE contribution that equated to a limb-specific standardized difference of ≤ -1 SD: 20% at 120 W, 40% at 180 W and 240 W, and 46% at 400 W (Figure 4.4). Only one of those participants did not exhibit interlimb compensatory behavior at 180 W, 240 W, and 400 W. Therefore, either an interlimb or intralimb protective compensatory pattern was present among 47% at 120 W, 60% at 180 W and 240 W, and 77% at 400 W (likely 80% had all 15 participants tested). The reduced relative KE contribution on the surgical side was compensated by greater than usual relative contributions of other joint-specific actions within the same leg. Those compensatory patterns were highly individual, and the most common intralimb compensatory pattern was an overcompensation of the HE action due to reduced KE action (Figure 4.5). In fact, every participant with a reduced relative KE contribution (by ≤ -1 SD) had a greater than usual relative HE contribution (by ≥ 1 SD). Furthermore, the surgical side HE action was typically greater than the nonsurgical side (120 W: p = .12; 180 W: p = .05; 240 W: p = .05; 400 W: p = .29), and the mean limbspecific standardized differences were ≥ 1 SD at 180 W, 240 W, and 400 W (Table 4.4). Differences in relative contributions of all other joint-specific actions between the 101 % Whole-Leg Work (z) A 3 2 1 0 R2=0.10 -1 -2 -3 ND ACLR KE Control (mean±SD) -4 -5 0 % Whole-Leg Work (z) B 90 180 270 360 450 540 630 D ACLR KE Model Fit 720 810 900 Days After ACLR 3 2 1 0 R2=0.08 -1 -2 -3 -4 -5 0 % Whole-Leg Work (z) C 90 180 270 360 450 540 630 720 810 900 Days After ACLR 3 2 1 R2=0.11 0 -1 -2 -3 -4 -5 0 % Whole-Leg Work (z) D 90 180 270 360 450 540 630 720 810 900 Days After ACLR 3 2 1 R2=0.15 0 -1 -2 -3 -4 -5 0 90 180 270 360 450 540 Days After ACLR 630 720 810 900 Figure 4.4. Between-subject data for the relative contribution (%) of the surgical side knee extension (KE) to total whole-leg mechanical work. Individual data are graphed over days since ACL reconstruction (ACLR) at 120 W (A), 180 W (B), 240 W (C), and 400 W (D). Data for individuals with a nondominant side ACLR (ND ACLR KE) are illustrated with open circles (⚬), and those with a dominant side ACLR (D ACLR KE) are illustrated using closed circles (●). All data are expressed as limb-specific standardized differences, that is, ND ACLR KE data are expressed as a standardized value (in units of SD) of the relative dominant limb KE action within the control group, and vice-versa for D ACLR KE data. The nonlinear function that best fit the data (model fit) is illustrated with a dotted line (⋯) and the coefficient of variation (R2) is displayed. 102 Pedal Power (W) A 600 D ACLR ND NS D CON±1SD ND CON ±1SD 500 400 300 200 100 0 -100 -200 0 Relative Ankle Power (%) B 90 180 Pedal Angle (º) 270 360 180 Pedal Angle (º) 270 360 300 250 200 150 100 50 0 -50 -100 0 Relative Knee Power (%) C 90 300 250 200 150 100 50 0 -50 -100 Relative Hip Power (%) D 0 90 180 Pedal Angle (º) 270 360 0 90 180 Pedal Angle (º) 270 360 300 250 200 150 100 50 0 -50 -100 Figure 4.5. Interlimb and intralimb compensatory patterns of an ACL participant during submaximal cycling. The pedal power (A), relative ankle power (B), relative knee power (C), relative and hip power (D) are displayed across the pedal cycle (0–360°) for a participant who was assessed 144 days after a dominant leg ACL reconstruction (D ACLR). Data is also displayed for the nondominant nonsurgical limb (ND NS). Jointspecific power data (B–D) are expressed relative (%) to the average whole-leg cycling power across the pedal cycle. Typical ranges (mean ± SD) for the nondominant and dominant limbs of the control group participants are also displayed (ND CON ±1SD; ND CON ±1SD). During cycling at 240 W, the ND NS limb produced greater extension (0– 180°) power than both the D ACLR limb and the nondominant limb of the control group. This participant also demonstrated greater relative ankle extension, hip extension, and knee flexion power on the surgical side to accommodate a lower relative knee extension power. It is also worth noting that the nonsurgical limb had greater hip flexion power (180-360°). 103 surgical and nonsurgical limbs were not significant (Table 4.4), but some individuals compensated for reduced KE action using various combinations of greater ipsilateral AE, KF, and HE action, and even greater contralateral HF action (Figure 4.5). After RTP For my second hypothesis, I proposed that the interlimb and intralimb compensatory patterns would remain present among participants who had RTP. As stated above, the majority of participants undergoing rehabilitation (i.e., before RTP) demonstrated whole-leg attenuation of power on the surgical side, and that compensatory behavior increased in magnitude with greater cycling intensity. Conversely, among the five participants in this study who returned to full participation, there were no significant differences between the relative contributions that their surgical and nonsurgical limbs made toward the net double-leg power (Table 4.3). Additionally, the magnitudes of the mean limb-specific standardized differences at each cycling intensity were within typical ranges (i.e., within ± 1 SD) of the control group. These data suggest that interlimb compensatory patterns were not present after RTP. This observation was confirmed by the time-constants of the nonlinear regression functions for each work rate (R2 = .23– .37),which ranged from ~55 days at 120 W to ~90 days at 400 W (Figure 4.3). Regarding the presence of intralimb compensatory patterns after RTP, differences between the surgical and nonsurgical limb joint-specific actions were nonsignificant for the most part (Table 4.4). After RTP, there was a trend for greater relative HE contributions within the surgical limb compared to that of the nonsurgical leg (Table 4.4). Those differences were significant (p = .02) only at 180 W (Table 4.4), and the mean limb-specific standardized difference was ≥ 1 SD only at 400 W. As already alluded 104 above, there was large between-subject variation in the relative contribution made by each joint-specific action. Consequently, there was not a predictable pattern for the relative KE action (Figure 4.4), HE action (Figure 4.6), or other joint-specific contributions between participants over time following ACLR at any work rate (R2 ≤ .15 for all regression functions). At the least, the variation in the relative joint-specific contributions tended to decrease from before to after RTP (Table 4.4) and it appeared that ACL participants were fairly normal after RTP (Figures 4.4 and 4.6). Exploratory analysis Exploratory data analyses were only performed with data from participants before RTP. Data was subdivided by sex, limb-dominance, and graft type (Figure 4.7). Female participants (n = 8) demonstrated a protective compensatory pattern in ~84% of the trials performed across all intensities, compared to ~22% of the trials performed by male participants (n = 7; Figure 4.7)—male and female control participants had similar relative whole-leg contributions. Regarding the surgical side, the relative contributions of individuals with nondominant surgical limbs (n = 5) were ≤ -1 SD in 68% of the trials performed across all intensities compared to 49% of those with a dominant limb ACLR (n = 10; Figure 4.7). Finally, there were nine individuals with a bone-patellar tendonbone autograph and five with a hamstring tendon autograft. An interlimb compensatory pattern was demonstrated across 51% of the trials performed by participants (n = 9) with a bone-patellar tendon-bone autograft and 74% of the trials performed by those (n = 5) with a hamstring tendon autograph. It should be noted, however, that four out of five the hamstring tendon autograph participants were female participants. 105 % Whole-Leg Work (z) A 5 4 3 2 1 0 R2=0.02 -1 ND ACLR HE Control (mean±SD) -2 -3 0 % Whole-Leg Work (z) B 90 180 270 360 450 540 630 D ACLR HE Model Fit 720 810 900 Days After ACLR 5 4 3 2 1 0 R2=0.04 -1 -2 -3 0 % Whole-Leg Work (z) C 90 180 270 360 450 540 630 720 810 900 Days After ACLR 5 4 3 2 1 0 R2=0.04 -1 -2 -3 0 % Whole-Leg Work (z) D 90 180 270 360 450 540 630 720 810 900 Days After ACLR 5 4 3 2 R2=0.00 1 0 -1 -2 -3 0 90 180 270 360 450 540 Days After ACLR 630 720 810 900 Figure 4.6. Between-subject data for the relative contribution (%) of the surgical side hip extension (HE) action to the total whole-leg mechanical work. Individual data are graphed over days since ACL reconstruction (ACLR) at 120 W (A), 180 W (B), 240 W (C), and 400 W (D). Individuals with a nondominant side ACLR (ND ACLR HE) are illustrated with open circles (⚬), and those with a dominant side ACLR (D ACLR HE) are illustrated using closed circles (●). A more detailed explanation is provided above (Figure 4.4). 106 Before RTP (n = 15) -0.5, 47% -1.3, 53% -2.1, 53% -2.8, 69%* M (n = 7) 0.4, 29% 0.3, 14% 0.2, 14% -0.1, 33%* ND (n = 2) -0.5, 50% 1.2, 0% 1.4, 0% -0.3, 50% F (n = 8): -1.3, 63% -2.6, 88% -4.1, 88% -5.0, 100%* D (n = 5) 0.7, 20% -0.1, 20% -0.3, 20% -0.1, 25%* BPTB (n = 2) -0.5, 50% 1.2, 0% 1.4, 0% -0.3, 50% BPTB (n = 3) 0.0, 33% -1.1, 33% -1.7, 33% -1.6, 50%* ND (n = 5) 2.0, 80% -1.7, 60% -2.0, 60% -2.8, 75%* D (n = 10) 0.2, 30% -1.1, 50% -2.1, 50% -2.8, 67%* HT (n = 1) 1.9, 0% 0.6, 0% 0.9, 0% 0.2, 0% ND (n = 3) -3.0, 100% -3.6, 100% -4.4, 100% -5.2, 100%* CA (n = 1) 1.8, 0% 2.2, 0% 2.7, 0% 2.7, 0% BPTB (n = 1) -3.7, 100% -3.5, 100% -4.8, 100% -6.8, 100% HT (n = 2) -2.7, 100% -3.6, 100% -4.2, 100% -3.7, 100%* D (n = 5) -0.3, 40% -2.0, 80% -4.0, 80% -4.9, 100% BPTB (n = 3) 0.6, 33% -1.2, 67% -2.6, 67% -3.8, 100% HT (n = 2) -1.7, 50% -3.4, 100% -6.1, 100% -6.7, 100% BPTB (n = 9) -0.3, 44% -0.9, 44% -1.7, 44% -2.7, 75%* HT (n = 5) -1.4, 60% -2.7, 80% -3.9, 80% -4.2, 75%* Figure 4.7. Exploratory analysis of interlimb compensatory patterns. Surgical limb data for all participants before return to participation (RTP) are sub-divided by male (M) and female (F), and then by nondominant (ND) and dominant (D) surgical limbs, and again by either bone-patellar tendon-bone autograph, hamstring tendon autograph, or cadaver allograft. The sample size (n) is listed in each partition, whereas values in the first column indicate the mean limb-specific standardized difference for the whole-leg relative contribution to the net double-leg power. Values in the second column of each partition are the proportion of participants who had a limb-specific standardized difference that was ≤ -1 SD. Data for cycling intensities of 120 W, 180 W, 240 W, and 400 W are listed in each row, respectively, from top to bottom. Note that two participants did not perform trials at 400 W (*). 107 Discussion To my knowledge, this is the first investigation of interlimb and intralimb compensatory patterns after ACLR using cycling biomechanics. In this exploratory pilot study, the majority of participants undergoing rehabilitation from ACLR demonstrated an interlimb and/or intralimb compensatory pattern. The most common protective strategy was a whole-leg attenuation of power within the surgical limb. Several participants also demonstrated intralimb protective behavior, which usually presented as overcompensation of the HE action on behalf of reduced KE action (Table 4.4; Figure 4.5). After RTP, participants demonstrated fairly normal cycling biomechanics. The conclusion that ACLR participants return to normal after RTP must be considered with caution, however, as it was drawn from a between-subject analysis. Overall, this study provided proof-of-concept that submaximal cycling biomechanics can be used to safely evaluate whole-leg dynamic function during early stages of rehabilitation from ACLR. Data from this study offer several future research directions and practical applications. Before RTP Before RTP, 80% of the participants in this study demonstrated either an interlimb and/or intralimb compensatory pattern during submaximal cycling. Cycling involves whole-leg dynamic action by which the central nervous system can freely alter multiple degrees of freedom within the two-leg system to meet the demand of the locomotive task (Raasch & Zajac, 1999). After ACLR, there are at least two neuromuscular compensatory strategies that can protect the knee during cycling. The first strategy involves redistribution of the submaximal workload from the surgical to nonsurgical limb, which would generally coincide with less overall ankle, knee, and hip 108 work within the surgical limb. The second strategy involves reduction in the surgical KE action with an increase in effort from other joint actions within the same limb. Before RTP, participants’ surgical limb made a significantly lower contribution to the net double-leg power than the nonsurgical limb at all work rates (Table 4.3). Additionally, the relative KE contribution within the surgical leg was consistently less than that within the nonsurgical limb (Table 4.4), and 100% of the participants who had a reduced relative KE contribution action (≤ -1 SD) also had greater (≥ 1 SD) relative HE contribution. Overall, my findings are consistent with previous reports on ACL deficient individuals who presented interlimb (Hunt et al., 2004) and intralimb compensatory patterns (Hunt et al., 2003). Regarding the observed interlimb compensatory patterns, exploratory analysis revealed that participants with nondominant surgical limbs (n = 5) were ~40% more likely to present abnormal asymmetry compared to those with dominant surgical limbs (n = 10). In this study, limb-specific standardized differences between the ACL group and control group were used to account for whole-leg asymmetries that may be present after ACLR. Consistent with previous research (Carpes et al., 2010; Gabbard & Iteya, 1996; Hunt et al., 2004), there was a trend in this study for the dominant leg of the control group to make a greater relative contribution to the net double-leg power (Table 4.3; Figure 4.2). If noninjured nondominant limbs present lower relative contributions before an injury, then perhaps there could be even greater magnitudes of compensatory behavior within the nondominant limb after an injury. My data suggest that this may be the case. Further investigation is warranted because no previous studies have reported greater neuromuscular deficits in the nondominant limb after ACLR compared to the dominant 109 one. Sex-specific differences may explain, in part, the greater interlimb deficits that were observed within the nondominant surgical limbs. Before RTP, females were ~3.8 times more likely to have a large interlimb compensatory pattern (i.e., ≤ -1 SD) than the male participants (Figure 4.7). Furthermore, their mean limb-specific standardized differences were substantially greater in magnitude than the males at all cycling intensities (Figure 4.7). My data are consistent with previous authors who have reported greater neuromuscular deficits among females after ACLR compared to males. For example, Hewett and colleagues (2002) reported that females had greater single-leg stance instability than males for up to 12 months after ACLR. Also, compared to men, women have significantly greater knee and hip joint excursion asymmetry, and greater knee joint moment asymmetry during walking 6 months after ACLR (Di Stasi, Hartigan, & Snyder-Mackler, 2015). Taken together, my data sets a precedent for greater research into whether or not there are side and sex-specific relationships with the magnitude of interlimb compensatory patterns during whole-leg dynamic function after ACLR, especially since athletic females are at greater risk of ACL injury compared to their male counterparts (Wiggins et al., 2016), and because they may suffer ACL ruptures in their nondominant knee more often than their dominant one (Brophy, Silvers, Gonzales, & Mandelbaum, 2010; Ruedl et al., 2012). In conjunction with interlimb compensatory patterns, intralimb compensatory patterns were also present. As described above, reduced KE action was usually compensated by greater HE action (Figure 4.5). Participants in this study were generally naïve to cycling exercise, and they appeared to recruit a smaller relative HE contribution 110 compared to what has been reported among trained cyclists (Elmer et al., 2011; Martin & Brown, 2009). Nevertheless, Elmer and colleagues (2011) reported that HE was the dominant action during submaximal cycling. Therefore, it is not surprising that overcompensation with greater HE action was the most common strategy used to overcome reduced KE action after ACLR. Other than KE and HE, AE and KF are the two other main actions that contribute to total mechanical work during submaximal cycling (Broker & Gregor, 1994; Elmer et al., 2011; Ericson, 1998; Martin & Brown, 2009; van Ingen Schenau et al., 1990). While there were no systematic differences in the relative contributions of the AE and KF actions between the surgical and nonsurgical limbs (Table 4.4), some individuals compensated for reduced KE action with various combinations of greater ipsilateral AE, KF, and HE action, as well as greater contralateral HF action (Figure 4.7). In other words, intralimb compensatory patterns were highly individual, which is consistent with the fact that the central nervous system can freely alter the contributions made by each joint-specific action during cycling (Raasch & Zajac, 1999). Beyond the variety of interlimb and intralimb compensatory patterns that were present in this study, it was interesting to observe that the compensatory patterns progressed with work rate. Specifically, there was a negative correlation between cycling intensity and the mean limb-specific standardized differences of the relative whole-leg contributions (Table 4.3), suggesting that the protective mechanisms became more extreme with cycling intensity. Likewise, cycling intensity appeared to positively correlate with the mean limb-specific standardized differences of the relative contributions of the surgical limb AE and HE actions, and cycling intensity negatively 111 correlated with that of the KE action (Table 4.4). Muscular strength deficits could contribute, in part, to the greater whole-leg and KE action deficits that occurred at greater cycling intensities. Strength deficits could not have been a factor at 120 W, 180 W, and 240 W, however. At 400 W, the surgical limb produced a mean net cycling power across the pedal cycle (~180 W) that exceeded what was necessary at lower intensities. Therefore, it can be reasoned that the interlimb and knee-joint deficits were due to neuromuscular compensatory strategies, and not strength deficits. There was another compensatory pattern that was completely unexpected. When cycling intensities changed, the limb-specific standardized differences of the relative joint-specific contributions within the nonsurgical limb progressed in the same direction as the joint-specific actions within the surgical limb (Table 4.4). For example, the relative KE contribution of one participant (SUB # 4; Table 4.2) was -1.2 SD for the surgical limb and -0.2 SD for the nonsurgical limb at 120 W, -3.1 SD and -0.9 at 180 W, -3.3 SD and 1.3 SD at 240 W, and -4.0 SD and -1.9 SD at 400 W. The relative HE contribution within that same participant progressed from 1.4 SD for the surgical limb and -0.1 SD for the nonsurgical limb at 120 W, to 2.6 SD and 0.3 SD at 180 W, 3.2 SD and 0.6 SD at 240 W, and 2.9 SD and 1.6 SD at 400 W. Interestingly, previous authors have reported bilateral quadriceps activation failure in men and women after ACLR (see Hart, Pietrosimone, Hertel, & Ingersoll, 2010, and Hobson, 2018, for reviews). While it was not my aim, my data support existing research that unilateral ACLR can affect bilateral KE function, in addition to the other joint-specific actions involved with whole-leg dynamic function of both legs. 112 After RTP Contrary to my hypothesis, the five participants who visited the lab after RTP from ACLR demonstrated fairly normal interlimb (Table 4.3; Figure 4.3) and intralimb biomechanical patterns (Table 4.4; Figures 4.4 and 4.6). Indeed, ~23-37% of the variance in the limb-specific standardized difference of the relative contribution made by the surgical limb was accounted for by time since ACLR—those coefficients of variation would be considered large for the between-subject design used in this study. Jointspecific contributions varied to an even greater extent, and there was no distinguishable pattern (indicated by low R2 values) other than the fact that the limb-specific standardized differences of the relative joint-specific contributions fell within ± 1 SD of the mean of control group after RTP (Figures 4.4 and 4.6). At face value, my data suggest that interlimb and intralimb compensatory patterns are present before RTP and that they may disappear after RTP. My study cannot fully draw this conclusion, however. As indicated above, several athletes demonstrated interlimb and/or intralimb compensatory patterns before RTP, but not all of them. It is not known whether or not the athletes who participated after RTP had compensatory patterns before RTP because most participants only tested once. Conversely, a within-subject design could reveal that athletes may not fully improve. As mentioned above (see Results), two athletes performed follow-up visits. One of those athletes (SUB #9, Table 4.2) visited the lab for the first time at 97 days post-ACLR. After 6 weeks of rehabilitation under ideal conditions—due to her status as a national team athlete—she returned to the lab 141 days post-ACLR and demonstrated even more extreme interlimb and intralimb compensatory patterns compared to her initial visit. The 113 other participant (SUB #2; Table 4.2) was also a national team athlete. She demonstrated progressively greater whole-leg attenuation of power from her first (39 days post-ACLR) to second visit (72 days post-ACLR). Both examples provide evidence that it can be difficult to correct compensatory strategies that develop during early stages of rehabilitation (Grooms et al., 2015; Kapreli & Athanasopoulos, 2006; Nyland et al., 2014). In conclusion, it is clearly possible for athletes to demonstrate normal cycling biomechanics after RTP, but within-subject data from this exploratory pilot study reveal that some athletes may not improve over the course of 4–8 weeks of rehabilitation performed in ideal settings. More research is needed to determine what facilitates restoration of normal cycling biomechanics after ACLR. Practical Applications This exploratory pilot study demonstrated that submaximal cycling is a safe and efficacious way to identify interlimb and intralimb compensatory patterns during early phases of rehabilitation after ACLR when assessment of whole-leg dynamic function has not been traditionally performed. Cycling is one of the first multi-joint lower extremity activities indicated by surgeons after ACLR (van Grinsven et al., 2010). During early stages of rehabilitation, some individuals will present with interlimb and/or intralimb compensatory strategies at very low cycling intensities, while others will not. Therefore, I propose that it could become standard of practice to assess cycling biomechanics using the techniques outlined in this study before indicating cycling because it can be difficult to reverse compensatory strategies that develop during early stages of rehabilitation (Grooms et al., 2015; Kapreli & Athanasopoulos, 2006; Nyland et al., 2014). Indeed, cycling may be contraindicated until it can be performed with interlimb and intralimb 114 symmetry. This study also revealed that compensatory strategies progress with greater submaximal cycling work rates. As such, assessing cycling biomechanics over the course of rehabilitation could facilitate exercise prescription, particularly if cycling is to be used as a cross-training modality for cardiorespiratory conditioning instead of running. This point cannot be overstated because cycling and running likely utilize similar spinal cord based central pattern generators (Zehr & Duysens, 2004) and reinforcing a compensatory pattern during cycling could manifest during ground-based cyclical locomotion (i.e., running) after RTP. More research is needed to confirm this assumption, but overcompensation by the noninjured leg to protect the injured leg could increase risk of overuse injury, whereby reduced knee extension action, which can happen bilaterally, inevitably results in quadriceps muscle strength loss and greater risk of secondary ACL injury of both the ipsilateral and contralateral knees (Grindem, Snyder-Mackler, Moksnes, Engebretsen, & Risberg, 2016). In fact, it is possible that a greater proportion of the ~25% of athletes who suffer a second ACL injury after RTP (Wiggins et al., 2016) could be identified among the ~75–80% of athletes who present interlimb or intralimb compensatory patterns before RTP. Conversely, on the basis that cycling and walking use common central pattern generators (Zehr & Duysens, 2004), correcting compensatory patterns during submaximal cycling could be an efficacious way to improve outcomes after RTP from ACLR. Indeed, cycling with feedback has been successfully used in motor function rehabilitation of stroke patients (see Barbosa, Santos, and Martins, 2015, for a review). Coincidently, compensatory patterns observed during cycling in this study are similar to 115 those observed during walking and running after ACLR (Devita et al., 1998, Di Stasi et al., 2015; Patras et al., 2010), and more research could examine if there is a relationship. If so, perhaps cycling can similarly improve motor function for patients undergoing ACLR rehabilitation as it has for patients with other conditions. Previous research (Korff, Romer, Mayhew, & Martin, 2007), including work from our lab (Rimer et al., 2012), has reported that simple cycling technique instructions can acutely change relative jointspecific contributions toward the whole-leg effort. While it was beyond the current scope, one athlete from this study (SUB #9; Table 4.2) wished to correct her cycling asymmetries (Figure 4.5). Through a series of technical instruction cues, that athlete was able to discover a cycling technique that elicited symmetrical cycling patterns (data not reported). In other words, if the central nervous system can freely alter the available degrees of freedom within the limbs to perform the task (Raasch & Zajac, 1999), it may be possible to improve outcomes at the time of RTP by retraining the central nervous system into using a better locomotive coordination strategy. In fact, this could be done using the first whole-leg dynamic exercise that is indicated after surgery. References Abrams, G. D., Harris, J. D., Gupta, A. K., McCormick, F. M., Bush-Joseph, C. A., Verma, N. N., ... Bach Jr, B. R. (2014). Functional performance testing after anterior cruciate ligament reconstruction: A systematic review. Orthopaedic Journal of Sports Medicine, 2(1), 1-10. Adams, D., Logerstedt, D., Hunter-Giordano, A., Axe, M. J., & Snyder-Mackler, L. (2012). Current concepts for anterior cruciate ligament reconstruction: A criterion-based rehabilitation progression. Journal of Orthopaedic & Sports Physical Therapy, 42(7), 601-614. Barbosa, D., Santos, C. P., & Martins, M. (2015). The application of cycling and cycling combined with feedback in the rehabilitation of stroke patients: A review. Journal of Stroke and Cerebrovascular Diseases, 24(2), 253-273. 116 Barber, S. D., Noyes, F. R., Mangine, R. E., & Hartman, W. (1990). Quantitative assessment of functional limitations in normal and anterior cruciate ligamentdeficient knees. Clinical Orthopaedics and Related Research, 255, 204-214. Broker, J. P., & Gregor, R. J. (1994). Mechanical energy management in cycling: Source relations and energy expenditure. Medicine and Science in Sports and Exercise, 26(1), 64-74. Brophy, R., Silvers, H. J., Gonzales, T., & Mandelbaum, B. R. (2010). Gender influences: The role of leg dominance in ACL injury among soccer players. British Journal of Sports Medicine, 44(10), 694-697. Carpes, F. P., Mota, C. B., & Faria, I. E. (2010). On the bilateral asymmetry during running and cycling—A review considering leg preference. Physical Therapy in Sport, 11(4), 136-142. de Leva, P. (1996). Adjustments to Zatsiorsky-Seluyanov's segment inertia parameters. Journal of Biomechanics, 29(9), 1223-1230. Devita, P., Hortobagyi, T., & Barrier, J. (1998). Gait biomechanics are not normal after anterior cruciate ligament reconstruction and accelerated rehabilitation. Medicine and Science in Sports and Exercise, 30(10), 1481-1488. Di Stasi, S., Hartigan, E. H., & Snyder-Mackler, L. (2015). Sex-specific gait adaptations prior to and up to 6 months after anterior cruciate ligament reconstruction. Journal of Orthopaedic & Sports Physical Therapy, 45(3), 207-214. Elftman, H. (1939). Forces and energy changes in the leg during walking. American Journal of Physiology—Legacy Content, 125(2), 339-356. Elmer, S., Barratt, P. R., Korff, T., & Martin, J. C. (2011). Joint-specific power production during submaximal and maximal cycling. Medicine and Science in Sports and Exercise, 43, 1940-1947. Ericson, M. O. (1988). Mechanical muscular power output and work during ergometer cycling at different work loads and speeds. European Journal of Applied Physiology and Occupational Physiology, 57(4), 382-387. Escamilla, R. F., MacLeod, T. D., Wilk, K. E., Paulos, L., & Andrews, J. R. (2012). ACL strain and tensile forces for weight-bearing and non-weight-bearing exercises after ACL reconstruction: A guide to exercise selection. Journal of Orthopaedic & Sports Physical Therapy, 42(3), 208-220. Fleming, B. C., Beynnon, B. D., Renstrom, P. A., Peura, G. D., Nichols, C. E., & Johnson, R. J. (1998). The strain behavior of the anterior cruciate ligament during bicycling: An in vivo study. The American Journal of Sports Medicine, 26(1), 109-118. 117 Gabbard, C., & Iteya, M. (1996). Foot laterality in children, adolescents, and adults. Laterality: Asymmetries of Body, Brain and Cognition, 1(3), 199-206. Gardner, A. S., Stephens, S., Martin, D. T., Lawton, E., Lee, H., & Jenkins, D. (2004). Accuracy of SRM and power tap power monitoring systems for bicycling. Medicine and Science in Sports and Exercise, 36(7), 1252-1258. Grindem, H., Snyder-Mackler, L., Moksnes, H., Engebretsen, L., & Risberg, M. A. (2016). Simple decision rules can reduce reinjury risk by 84% after ACL reconstruction: The Delaware-Oslo ACL cohort study. British Journal of Sports Medicine, 50(13), 804-808. Grooms, D., Appelbaum, G., & Onate, J. (2015). Neuroplasticity following anterior cruciate ligament injury: A framework for visual-motor training approaches in rehabilitation. Journal of Orthopaedic & Sports Physical Therapy, 45(5), 381393. Hart, J. M., Pietrosimone, B., Hertel, J., & Ingersoll, C. D. (2010). Quadriceps activation following knee injuries: A systematic review. Journal of Athletic Training, 45(1), 87-97. Hewett, T. E., Paterno, M. V., & Myer, G. D. (2002). Strategies for enhancing proprioception and neuromuscular control of the knee. Clinical Orthopaedics and Related Research, 402, 76-94. Hoes, M. J., Binkhorst, R. A., Smeekes-Kuyl, A. E., & Vissers, A. C. (1968). Measurement of forces exerted on pedal and crank during work on a bicycle ergometer at different loads. European Journal of Applied Physiology and Occupational Physiology, 26(1), 33-42. Hunt, M. A., Sanderson, D. J., Moffet, H., & Inglis, J. T. (2003). Biomechanical changes elicited by an anterior cruciate ligament deficiency during steady rate cycling. Clinical Biomechanics, 18(5), 393-400. Hunt, M. A., Sanderson, D. J., Moffet, H., & Inglis, J. T. (2004). Interlimb asymmetry in persons with and without an anterior cruciate ligament deficiency during stationary cycling. Archives of Physical Medicine and Rehabilitation, 85(9), 1475-1478. Impellizzeri, F. M., Rampinini, E., Maffiuletti, N., & Marcora, S. M. (2007). A vertical jump force test for assessing bilateral strength asymmetry in athletes. Medicine and Science in Sports and Exercise, 39(11), 2044. Kapreli, E., & Athanasopoulos, S. (2006). The anterior cruciate ligament deficiency as a model of brain plasticity. Medical Hypotheses, 67(3), 645-650. Keller, T. S., Weisberger, A. M., Ray, J. L., Hasan, S. S., Shiavi, R. G., & Spengler, D. M. (1996). Relationship between vertical ground reaction force and speed during 118 walking, slow jogging, and running. Clinical Biomechanics, 11(5), 253-259. Korff, T., Romer, L. M., Mayhew, I. A. N., & Martin, J. C. (2007). Effect of pedaling technique on mechanical effectiveness and efficiency in cyclists. Medicine and Science in Sports and Exercise, 39(6), 991-995. Martin, J. C., Elmer, S. J., Horscroft, R. D., Brown, N. A., & Shultz, B. B. (2007). A lowcost instrumented spatial linkage accurately determines ASIS position during cycle ergometry. Journal of Applied Biomechanics, 23(3), 224-229. Martin, J. C., & Brown, N. A. (2009). Joint-specific power production and fatigue during maximal cycling. Journal of Biomechanics, 42(4), 474-479. Martin, J. C., Milliken, D. L., Cobb, J. E., McFadden, K. L., & Coggan, A. R. (1998). Validation of a mathematical model for road cycling power. Journal of Applied Biomechanics, 14, 276-291. McDaniel, J., Behjani, N. S., Elmer, S. J., Brown, N. A., & Martin, J. C. (2014). Jointspecific power-pedaling rate relationships during maximal cycling. Journal of Applied Biomechanics, 30(3), 423-430. Myer, G. D., Schmitt, L. C., Brent, J. L., Ford, K. R., Barber Foss, K. D., Scherer, B. J., ... Hewett, T. E. (2011). Utilization of modified NFL combine testing to identify functional deficits in athletes following ACL reconstruction. Journal of Orthopaedic & Sports Physical Therapy, 41(6), 377-387. Neptune, R. R., & Hull, M. L. (1995). Accuracy assessment of methods for determining hip movement in seated cycling. Journal of Biomechanics, 28(4), 423427-425437. Nyland, J., Wera, J., Klein, S., & Caborn, D. N. (2014). Lower extremity neuromuscular compensations during instrumented single leg hop testing 2–10 years following ACL reconstruction. The Knee, 21(6), 1191-1197. Otzel, D. M., Chow, J. W., & Tillman, M. D. (2015). Long-term deficits in quadriceps strength and activation following anterior cruciate ligament reconstruction. Physical Therapy in Sport, 16(1), 22-28. Palmieri-Smith, R. M., & Lepley, L. K. (2015). Quadriceps strength asymmetry after anterior cruciate ligament reconstruction alters knee joint biomechanics and functional performance at time of return to activity. The American Journal of Sports Medicine, 43(7), 1662-1669. Paterno, M. V., Rauh, M. J., Schmitt, L. C., Ford, K. R., & Hewett, T. E. (2014). Incidence of second ACL injuries 2 years after primary ACL reconstruction and return to sport. The American Journal of Sports Medicine, 42(7), 1567-1573. Paterno, M. V., Schmitt, L. C., Ford, K. R., Rauh, M. J., Myer, G. D., Huang, B., & Hewett, T. E. (2010). Biomechanical measures during landing and postural 119 stability predict second anterior cruciate ligament injury after anterior cruciate ligament reconstruction and return to sport. The American Journal of Sports Medicine, 38(10), 1968-1978. Patras, K., Ziogas, G., Ristanis, S., Tsepis, E., Stergiou, N., & Georgoulis, A. D. (2010). ACL reconstructed patients with a BPTB graft present an impaired vastus lateralis neuromuscular response during high intensity running. Journal of Science and Medicine in Sport, 13(6), 573-577. Raasch, C. C., & Zajac, F. E. (1999). Locomotor strategy for pedaling: Muscle groups and biomechanical functions. Journal of Neurophysiology, 82(2), 515-525. Rimer, E. G., Elmer, S. J., & Martin, J. C. (2014). Joint-specific cycling biomechanics of a patellar-deficient leg: A case study. Medicine and Science in Sports and Exercise, 46(5S), 421. Rimer, E. G., Marshall, C. S., Wehmanen, K. W., Farley, J. R., & Martin, J. C. (2012). A simple instruction increases relative contribution of hip extension action during concentric cycling. Medicine and Science in Sports and Exercise, 44, 471. Ruedl, G., Webhofer, M., Helle, K., Strobl, M., Schranz, A., Fink, C., ... Burtscher, M. (2012). Leg dominance is a risk factor for noncontact anterior cruciate ligament injuries in female recreational skiers. The American Journal of Sports Medicine, 40(6), 1269-1273. Schmitt, L. C., Paterno, M. V., & Hewett, T. E. (2012). The impact of quadriceps femoris strength asymmetry on functional performance at return to sport following anterior cruciate ligament reconstruction. Journal of Orthopaedic & Sports Physical Therapy, 42(9), 750-759. 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. The Journal of Bone and Joint Surgery. British, 83(8), 1104-1110. van Grinsven, S., Van Cingel, R. E. H., Holla, C. J. M., & Van Loon, C. J. M. (2010). Evidence-based rehabilitation following anterior cruciate ligament reconstruction. Knee Surgery, Sports Traumatology, Arthroscopy, 18(8), 11281144. van Ingen Schenau, G. J., Van Woensel, W. W. L. M., Boots, P. J. M., Snackers, R. W., & De Groot, G. (1990). Determination and interpretation of mechanical power in human movement: Application to ergometer cycling. European Journal of Applied Physiology and Occupational Physiology, 61(1), 11-19. Wiggins, A. J., Grandhi, R. K., Schneider, D. K., Stanfield, D., Webster, K. E., & Myer, G. D. (2016). Risk of secondary injury in younger athletes after anterior cruciate ligament reconstruction: A systematic review and meta-analysis. The American 120 Journal of Sports Medicine, 44(7), 1861-1876. Zehr, E. P., & Duysens, J. (2004). Regulation of arm and leg movement during human locomotion. The Neuroscientist, 10(4), 347-361. CHAPTER 5 REFLECTIONS OF AN APPLIED SPORT SCIENTIST Opening Remarks The variety of topics presented in this dissertation exemplify the interdisciplinary potential of applied sport science while providing novel insights and applications for sports performance and rehabilitation. I propose that the successful sport scientist is capable of addressing questions and developing applications across various disciplines in order to improve outcomes in sport. For my dissertation, I chose breadth over depth in order to broaden my knowledge across various topics relevant to sport. In the following paragraphs, I will briefly summarize each study in light of the history that brought them to life. Nobel Laureate Sir Peter Medawar suggested that the scientific paper is fraudulent because it ignores the actual thought processes and events that gave rise to the actual work (Medawar, 1999). By providing the history behind each study, it will become clear that they were inspired by real-life challenges we faced in the sports performance setting. Indeed, there is greater translational value when sport science research is driven by the experiential knowledge and problems faced by those working in the field (Greenwood, Davids, & Renshaw, 2012). Compared to basic science, which does not immediately translate into practice, applied sport science can provide an immediate benefit to coaches, support staff, and athletes (Drust & Green, 2013; Haff, 2010). And by 122 meeting their needs, the applied sport scientist effectively achieves the ultimate goal of sport science, which is to improve sports performance (Bishop, 2008; Haff, 2010). Historical Perspectives Repeated-sprint ability The key finding from my first study was that peak power and critical power can predict repeated-sprint ability among trained athletes. The lead up to that study started in 2014 when my colleagues with the University of Utah Football Team expressed how difficult it was to stop the University of Oregon Football Team's trendsetting high-tempo, no-huddle offense. Compared to "traditional" play, which involves series of short bursts of physical activity (mean: 5.2 s; SD: 1.6 s) interspersed by longer intervals (mean: 35.2 s; SD: 7.7 s; Iosia & Bishop, 2008), Oregon's offense had a much greater tempo because their players hustled back to their positions after each play in order to start the next play as quickly as possible. Many coaches believed that the high-tempo offense was so effective because players on opposing defenses lacked the conditioning necessary for that tempo of play. To assist our strength and conditioning coaches with the challenge we faced, I reviewed the literature for information related to this exercise activity pattern. Research on repeated-sprint exercise dated back over 30 years (Wootton & Williams, 1983), but I had just discovered the topic for the first time. From what was described in the literature (Iosia & Bishop, 2008), traditional American football activity already met the definition of repeated-sprint activity (Girard, Mendez-Villanueva, & Bishop, 2011). Understanding repeated-sprint exercise became even more critical knowing we would face a higher tempo of play against several other copycat opponents who adopted their own high-tempo offensive schemes. As I embarked upon my first 123 study, to investigate if peak power and critical power could predict repeated-sprint ability, the Utah Football Team could not wait for answers. The research was fairly unanimous that single-sprint performance is the primary predictor of repeated-sprint ability (Bishop, Lawrence, & Spencer, 2003; Bishop & Spencer, 2004; Dawson, Fitzsimons, & Ward, 1993). Moreover, while no research had investigated the relationship between critical power and repeated-sprint ability, a wealth of research (see Bishop, Girard, & Mendez-Villanueva, 2011, for a review) promoted training that would improve the ability to recover between sprints. According to others’ recommendations (Bishop et al., 2011; Buchheit & Laursen, 2013; Stone & Kilding, 2009), my colleagues introduced high-intensity interval training to the 2015 off-season training regimen in hopes that it could improve their players’ repeated-sprint ability. To gain buy-in from the coaches, we told them this training would improve stamina so that we could beat Oregon—That season, the Utah Football Team formed a goal to beat Oregon, and this was just one element that helped Utah hand Oregon their greatest defeat in decades. Today, we have learned from my research that peak power and critical power interplay to form a continuum for repeated-sprint ability. More research with a larger sample size is needed to confirm the presence of this continuum. If it truly exists, then further research could examine how different training programs may sufficiently cater to the individual needs of athletes seeking greater repeated-sprint ability. Single-leg cycling The main finding from my second study was that short-term, high-intensity single-leg cycling interval training improved repeated-sprint ability and endurance 124 running performance among ground-based team-sport athletes. My committee chair Dr. Jim Martin introduced single-leg cycling to me shortly after he coauthored a paper (Abbiss et al., 2011) reporting that training using single-leg cycling was superior to double-leg cycling. The finding in that study that immediately caught my attention was that short-term high-intensity single-leg cycling interval training doubled muscle glucose transporter protein 4 (GLUT-4) content (Abbiss et al., 2011). GLUT-4 is responsible for glucose uptake into the muscle (Kawanaka, 2012), whereas greater GLUT-4 protein concentration can increase uptake during exercise (Howlett, Andrikopoulos, Proietto, & Hargreaves, 2013), and immediately after muscle contraction in the presence (Dela, Handberg, Mikines, Vinten, & Galbo, 1993) and absence (Kawanaka, 2012) of insulin. Furthermore, GLUT-4 abundance is associated with fatigue resistance during intense continuous exercise (Gorselink et al., 2002). The findings of Abbiss and colleagues (2011) were vital at the time (circa 2012) because I served the United States Women's Alpine Ski Team as their strength and conditioning specialist. My fieldwork with the team revealed very high blood lactate concentrations (unpublished data) after single runs lasting 45–60 s, suggesting that muscle glycogen was a primary fuel source during intense alpine skiing. As such, singleleg cycling training appeared to be a promising strategy that could improve the quality of on-snow training. During their 2012 off-season, there were two concentrated strength and conditioning training periods when we implemented high-intensity single-leg cycling training with the national team racers using the same protocol described in Chapter 3. I did not conduct a carefully designed study but casually observed that most of the racers I coached had improved their performance during an assessment we referred to as 125 the 30-30-30. The 30-30-30 involved two series of 30 consecutive maximal jumps interspersed by 30 s of standing. At the end of that off-season, I did a more careful analysis and confirmed a systematic improvement in fatigue resistance across the first and second series of 30 jumps. That was the first time in six off-seasons that I had observed a team-wide improvement in the 30-30-30. It was also the first clue that singleleg cycling training may be an efficacious cross-training modality that could improve performance during fatiguing ground-based exercise because it was the only new training strategy I implemented that year—It was also the first, and only year to date, that seven different women from the US Ski Team stood on a World Cup race podium during a single season. Shortly after that season, I made the transition to the University of Utah Athletics Department. Beyond football, my review of literature revealed that several other sports within Utah Athletics had components of repeated-sprint exercise, including tennis (Kovacs, 2006), soccer (Carling, Le Gall, & Dupont, 2012), basketball (Matthew & Delextrat, 2009), and (as of 2017) lacrosse (Polley, Cormack, Gabbett, & Polglase, 2015). Several of our strength and conditioning coaches started assessing repeated-sprint ability within their teams. In line with previous recommendations (Bishop et al., 2011; Buchheit & Laursen, 2013; Stone & Kilding, 2009), we refined our methods in order to implement effective high-intensity interval training programs with large groups of athletes who have strict time demands. However, we agreed with others (Buchheit & Laursen, 2013) who have cautioned that high-intensity interval training can increase overall musculoskeletal strain when implemented concurrently with other activities common to the team-sport training 126 process (Stone & Kilding, 2009). We considered alternative ways to improve conditioning while maintaining lower running volumes, and cycling was an immediate option. Cycling is a nonweight-bearing task that can improve high-intensity running performance (Etxebarria, Anson, Pyne, & Ferguson, 2014). Compared to double-leg cycling, Abbiss and colleagues (2011) also reported that single-leg cycling training elicited significant peripheral adaptations that would theoretically facilitate greater repeated-sprint ability (see Girard et al., 2011, for a review). Despite my own anecdotal evidence, no one had ever tested whether or not single-leg cycling training could improve measures of high-intensity ground-based exercise performance. The study was initiated, and we have learned that single-leg cycling is an effective cross-training modality that can improve repeated-sprint ability and endurance running performance. Beyond the main findings of my study, there is plenty of work that still needs to be done. Firstly, more research should investigate the extent by which central and/or peripheral adaptations contribute to improvements in repeated-sprint ability and endurance running performance after single-leg cycling training. Furthermore, experienced lacrosse players with recreational strength and conditioning training experience participated in my study. There is need to find out if athletes with greater training status could also benefit from single-leg cycling training. Finally, another limitation in the study was the fact that we don’t know the smallest worthwhile changes in fitness that would translate to improvements in sport performance. If further research could truly define those smallest worthwhile differences, then it may be possible to consider the extent by which single-leg cycling training—and other types of training— 127 can benefit sports performance. Anterior cruciate ligament reconstruction The main finding from my final study was that most athletes demonstrate either interlimb or intralimb compensatory patterns after ACLR before return to participation. While this was the final study of my dissertation, the history behind it dates back the furthest. My interest in ACLR rehabilitation started in 2010, during my time with the US Women's Alpine Ski Team. One racer on the team tore her ACL in 2007, and then suffered three more significant time-loss injuries over the next 3 years, two of which were subsequent ACL ruptures. A case-study of that athlete initiated a process that eventually led to a functional assessment battery using a force plate to assess unilateral strength and jump symmetry in order to make decisions about our racers' readiness to return-to-snow after ACLR. By 2013, our entire organization adopted what we thought was a robust return-tosnow assessment battery. We hoped the protocol would eliminate functional asymmetries at the time of return-to-snow in order to mitigate risk of secondary ACL injury (Paterno et al., 2010). Many athletes met our strict criteria, but some still seemed to have something wrong. We as trained professionals recognized that athletes did not move as naturally when they performed the assessments with their surgical leg, but it was not possible to truly define the movement flaws with the naked eye. Unfortunately, many of the racers suffered further complications including secondary ACL injuries. In discussing this problem (circa 2013) with Dr. Martin, he proposed to assess athletes after ACLR using cycling biomechanics. Beyond whole-leg dynamic assessments using force plates, Dr. Martin's cycling biomechanics system (Elmer, 128 Barratt, Korff, & Martin, 2011; Martin & Brown, 2009; Martin, Elmer, Horscroft, Brown, & Shultz, 2007) presented a unique opportunity to study athletes much earlier in their rehabilitation process when maximal whole-leg dynamic actions cannot be performed (Adams, Logerstedt, Hunter-Giordano, Axe, & Snyder-Mackler, 2012; van Grinsven, van Cingel, Holla, & van Loon, 2010). More importantly, his protocol was capable of determining each leg’s relative contribution to the double-leg effort, as well as the relative contributions of joint-specific actions to the whole-leg mechanical work (McDaniel, Behjani, Elmer, Brown, & Martin, 2014). Amazingly, no previous research had investigated intralimb and interlimb compensatory patterns after ACLR using submaximal cycling. We provided proof-ofconcept for the project using a case-study with a participant who lost his right patella after a snowboarding accident (Rimer, Elmer, & Martin, 2014). That was convenient at the time because our cycle ergometer was only instrumented on the right side. As expected, the relative patellar-deficient knee extension contribution to the whole-leg work was 2.3 SD less than that demonstrated by trained cyclists (Elmer et al., 2011). The absence of knee extension action in his patellar-deficient leg was compensated by a larger than usual (3.3 SD) relative contribution of the ipsilateral hip extension action. There was also overcompensation by the contralateral leg, which produced 69% of the net doubleleg power (Rimer et al., 2014). While it was just one participant, that case-study provided evidence that we could successfully detect interlimb and intralimb compensatory patterns using submaximal cycling biomechanics. Gratefully, colleagues at Brunel University loaned us a second force pedal which allowed us to instrument both sides of the ergometer. Their generosity 129 made it possible for us to demonstrate that submaximal cycling can safely detect protective compensatory patterns during early stages of ACLR rehabilitation. We are the first to accomplish this. In doing so, we readily embraced participants with different confounding factors because we wanted to explore leads for future research. My study provides data to support future research investigating side and/or sex specific relationships with the magnitude of interlimb compensatory pattern during submaximal cycling after ACLR. Perhaps one of the most novel observations was that when compensatory patterns progressed in magnitude with greater cycling intensities, the progression actually happened bilaterally. Overall, these initial observations deserve greater attention because it is possible that individuals who demonstrate protective compensatory patterns before return to participation make-up a greater portion of the ~25% of athletes who suffer another ACL injury after return to participation (Wiggins et al., 2016). Indeed, it is our goal to pursue research that can improve outcomes after ACLR. Because the central nervous system can freely alter the available degrees of freedom in the two-leg system during cycling (Raasch & Zajac, 1999), it may be possible to retrain the central nervous system to adopt a better locomotive coordination strategy. Previous work has indicated that by following basic instructions, individuals can easily change their cycling biomechanics (Korff, Romer, Mayhew, & Martin, 2007; Rimer, Marshall, Wehmanen, Farley, & Martin, 2012). In fact, one participant from my study visited the laboratory for this purpose. During a single session, I successfully helped the participant restore symmetrical cycling biomechanics through a series of simple instructional cues. Data from that participant (not presented herein) has helped us secure funding for future 130 research which will determine if real-time feedback during submaximal cycling can be added to the ACLR rehabilitation protocol in order to improve outcomes after return to participation. In other words, sometimes the unintentional outcomes end up providing the greatest value. Final Thoughts The Ph.D. experience was far more transformative than what was offered by the three studies presented herien. As the letters translate, this experience has helped me establish a philosophy for my approach as an applied sport scientist. It is my goal to operate as an interdisciplinary hybrid who crosses disciplinary boundaries in order to facilitate greater collaboration that can improve sports performance (Glazier, 2015). Indeed, other industries have praised the value of the “jack of all trades.” They possess uncanny ability to facilitate interdisciplinary collaboration (Boehm & Hogan, 2014; Carey & Smith, 2007; Gohar et al., 2019), by which professionals with various combinations of knowledge and skill approach common problems as full partners and interchangeable leaders (Balagué, Torrents, Hristovski, & Kelso, 2017; Rosenfield, 1992). Interdisciplinary practice can provide a profoundly beneficial impact (Gohar et al., 2019), but perhaps transdisciplinary practice is the upper echelon of sports performance management (Balagué et al., 2017). In transdisciplinary collaboration, disciplinary boundaries of each professional quickly fade as team members become sufficiently familiar and skilled with the concepts and techniques unique to the other disciplines (Rosenfield, 1992). In essence, a sort of crosspollination occurs. As I advance through my career, I hope to become a leader who will foster an environment that blurs 131 disciplinary boundaries while fading hierarchies. Together, we will address problems with broader and deeper analyses in order to generate new knowledge, strategies, and a unified language that transcends our former disciplinary silos . References Abbiss, C. R., Karagounis, L. G., Laursen, P. B., Peiffer, J. J., Martin, D. T., Hawley, J. A., … Martin, J. C. (2011). Single-leg cycle training is superior to double-leg cycling in improving the oxidative potential and metabolic profile of trained skeletal muscle. Journal of Applied Physiology, 110(5), 1248-1255. Adams, D., Logerstedt, D., Hunter-Giordano, A., Axe, M. J., & Snyder-Mackler, L. (2012). Current concepts for anterior cruciate ligament reconstruction: A criterion-based rehabilitation progression. Journal of Orthopaedic & Sports Physical Therapy, 42(7), 601-614. Balagué, N., Torrents, C., Hristovski, R., & Kelso, J. A. S. (2017). Sport science integration: An evolutionary synthesis. European Journal of Sport Science, 17(1), 51-62. Bishop, D. (2008). An applied research model for the sport sciences. Sports Medicine, 38(3), 253-263 Bishop, D., Girard, O., & Mendez-Villanueva, A. (2011). Repeated-sprint ability—Part II. Sports Medicine, 41(9), 741-756. Bishop, D., Lawrence, S., & Spencer, M. (2003). Predictors of repeated-sprint ability in elite female hockey players. Journal of Science and Medicine in Sport, 6(2), 199209. Bishop, D., & Spencer, M. (2004). Determinants of repeated-sprint ability in well-trained team-sport athletes and endurance-trained athletes. Journal of Sports Medicine and Physical Fitness, 44(1), 1-7. Boehm, D. N., & Hogan, T. (2014). ‘A jack of all trades’: The role of PIs in the establishment and management of collaborative networks in scientific knowledge commercialisation. The Journal of Technology Transfer, 39(1), 134-149. Buchheit, M., & Laursen, P. B. (2013). High-intensity interval training: Solutions to the programming puzzle. Sports Medicine, 43(10), 927-954. Carey, G. E., & Smith, J. A. (2007). Jack-of-all-trades, master of none: Postgraduate perspectives on interdisciplinary health research in Australia. BMC health services research, 7(1), 48. 132 Carling, C., Le Gall, F., & Dupont, G. (2012). Analysis of repeated high-intensity running performance in professional soccer. Journal of Sports Sciences, 30(4), 325-336. Dawson, B., Fitzsimons, M., & Ward, D. (1993). The relationship of repeated sprint ability to aerobic power and performance measures of anaerobic work capacity and power. Australian Journal of Science and Medicine in Sport, 25, 88-93. Dela, F., Handberg, A., Mikines, K. J., Vinten, J., & Galbo, H. (1993). GLUT 4 and insulin receptor binding and kinase activity in trained human muscle. The Journal of Physiology, 469(1), 615-624. Drust, B., & Green, M. (2013). Science and football: Evaluating the influence of science on performance. Journal of Sports Sciences, 31(13), 1377-1382. Elmer, S., Barratt, P. R., Korff, T., & Martin, J. C. (2011). Joint-specific power production during submaximal and maximal cycling. Medicine and Science in Sports and Exercise, 43, 1940-1947. Etxebarria, N., Anson, J. M., Pyne, D. B., & Ferguson, R. A. (2014). High-intensity cycle interval training improves cycling and running performance in triathletes. European Journal of Sport Science, 14(6), 521-529. Girard, O., Mendez-Villanueva, A., & Bishop, D. (2011). Repeated-sprint ability—Part I. Sports Medicine, 41(8), 673-694. Glazier, P. S. (2015). Towards a Grand Unified Theory of sports performance. Human Movement Science, 56, 139-156. Gohar, F., Maschmeyer, P., Mfarrej, B., Lemaire, M., Roncarolo, M. G., Wedderburn, L. R., & van Royen, A. (2019). Driving medical innovation through interdisciplinarity: Unique opportunities and challenges. Frontiers in Medicine, 6(35), 1-6. Greenwood, D., Davids, K., & Renshaw, I. (2012). How elite coaches' experiential knowledge might enhance empirical research on sport performance. International Journal of Sports Science & Coaching, 7(2), 411-422. Gorselink, M., Drost, M. R., De Brouwer, K. F. J., Schaart, G., Van Kranenburg, G. P. J., Roemen, T. H. M., ... Van Der Vusse, G. J. (2002). Increased muscle fatigability in GLUT-4-deficient mice. American Journal of Physiology-Endocrinology And Metabolism, 282(2), E348-E354. Haff, G. (2010). Roundtable discussion: Sport science. Strength & Conditioning Journal, 32(2), 33-45. Howlett, K. F., Andrikopoulos, S., Proietto, J., & Hargreaves, M. (2013). Exerciseinduced muscle glucose uptake in mice with graded, muscle-specific GLUT-4 133 deletion. Physiological Reports, 1(3). Iosia, M. F., & Bishop, P. A. (2008). Analysis of exercise-to-rest ratios during division IA televised football competition. Journal of Strength & Conditioning Research, 22(2), 332-340. Kawanaka, K. (2012). Regulation of glucose transport in skeletal muscle during and after exercise. The Journal of Physical Fitness and Sports Medicine, 1(4), 563-572. Korff, T., Romer, L. M., Mayhew, I. A. N., & Martin, J. C. (2007). Effect of pedaling technique on mechanical effectiveness and efficiency in cyclists. Medicine and Science in Sports and Exercise, 39(6), 991-995. Kovacs, M. S. (2006). Applied physiology of tennis performance. British Journal of Sports Medicine, 40(5), 381-386. Matthew, D., & Delextrat, A. (2009). Heart rate, blood lactate concentration, and timemotion analysis of female basketball players during competition. Journal of Sports Sciences, 27(8), 813-821. Martin, J. C., & Brown, N. A. (2009). Joint-specific power production and fatigue during maximal cycling. Journal of Biomechanics, 42(4), 474-479. Martin, J. C., Elmer, S. J., Horscroft, R. D., Brown, N. A., & Shultz, B. B. (2007). A lowcost instrumented spatial linkage accurately determines ASIS position during cycle ergometry. Journal of Applied Biomechanics, 23(3), 224-229. McDaniel, J., Behjani, N. S., Elmer, S. J., Brown, N. A., & Martin, J. C. (2014). Jointspecific power-pedaling rate relationships during maximal cycling. Journal of Applied Biomechanics, 30(3), 423-430. Medawar, P. (1999). Is the scientific paper a fraud? In E. Scanlon, R. Hill, & K. Junker (Eds.), Communicating science: Professional contexts (27-31), New York, New York: Routledge. Paterno, M. V., Schmitt, L. C., Ford, K. R., Rauh, M. J., Myer, G. D., Huang, B., & Hewett, T. E. (2010). Biomechanical measures during landing and postural stability predict second anterior cruciate ligament injury after anterior cruciate ligament reconstruction and return to sport. The American Journal of Sports Medicine, 38(10), 1968-1978. Polley, C. S., Cormack, S. J., Gabbett, T. J., & Polglaze, T. (2015). Activity profile of high-level Australian lacrosse players. Journal of Strength & Conditioning Research, 29(1), 126-136. Raasch, C. C., & Zajac, F. E. (1999). Locomotor strategy for pedaling: Muscle groups and biomechanical functions. Journal of Neurophysiology, 82(2), 515-525. 134 Rimer, E. G., Elmer, S. J., & Martin, J. C. (2014). Joint-specific cycling biomechanics of a patellar-deficient leg: A case study. Medicine and Science in Sports and Exercise, 46(5S), 421. Rimer, E. G., Marshall, C. S., Wehmanen, K. W., Farley, J. R., & Martin, J. C. (2012). A simple instruction increases relative contribution of hip extension action during concentric cycling. Medicine and Science in Sports and Exercise, 44, 471. Rosenfield, P. L. (1992). The potential of transdisciplinary research for sustaining and extending linkages between the health and social sciences. Social Science & Medicine, 35(11), 1343-1357. Stone, N. M., & Kilding, A. E. (2009). Aerobic conditioning for team sport athletes. Sports Medicine, 39(8), 615-642. van Grinsven, S., Van Cingel, R. E. H., Holla, C. J. M., & Van Loon, C. J. M. (2010). Evidence-based rehabilitation following anterior cruciate ligament reconstruction. Knee Surgery, Sports Traumatology, Arthroscopy, 18(8), 11281144. Wiggins, A. J., Grandhi, R. K., Schneider, D. K., Stanfield, D., Webster, K. E., & Myer, G. D. (2016). Risk of secondary injury in younger athletes after anterior cruciate ligament reconstruction: A systematic review and meta-analysis. The American Journal of Sports Medicine, 44(7), 1861-1876. Wootton, S., & Williams, C. (1983). The influence of recovery duration on repeated maximal sprints. In H. G. Knuttgen, J. A. Vogel, & J. Poortmans (Eds.), Biochemistry of exercise, international series in sports science (Vol. 8, pp. 269273). Champaign, Illinois: Human Kinetics.