Running as a form of aerobic exercise is highly popular with over 17 million road race finishers in the United States in 2015 (40). Many uninjured middle and long distance runners supplement or replace some of their running training with aerobic cross-training to increase total aerobic training volume or maintain fitness. Five to 15 days of running cessation, which may be the result of injury, can reduce time to exhaustion and maximal aerobic capacity (V̇o2max) in distance runners (11,24). Nonrunning aerobic training alone can help maintain aerobic fitness in nonrunners and runners alike (12,27,28). Including nonimpact or low-impact cross-training within running training may also allow runners to safely return to a full running volume while maintaining aerobic fitness after time off for recovery (e.g., pre or early season) or from injury.
Various types of aerobic cross-training modalities exist such as cycling (CYCLE), elliptical, swimming, water running and more recently, elliptical bikes (EBIKE). Because of the popular use of cross-training modalities among runners, it is important to consider the effects that these different cross-training methods have on running performance and injury risks. A newly designed outdoor EBIKE is an aerobic cross-training modality engineered as a hybrid between an elliptical trainer and bicycle with modifications intended to emulate the running motion (http://www.elliptigo.com/Description). Although a standard indoor elliptical (ELLIP) requires users to stand while training, it is immovable. Because of greater whole body involvement, running-like motion and side-to-side instability, the EBIKE may require greater core control and involvement (i.e., hip and pelvic musculature) compared with stationary elliptical and CYCLE. These specific characteristics of EBIKE could improve hip movement, lower limb joint mobility, muscle strength with training which may impact running performance, and injury risks in runners.
Regarding performance, running economy (RE)—rate of oxygen consumption (ml·min−1·kg−1) at a set running speed (7)—seems to have a strong association with distance running performance (26) and seems to be a better measure to differentiate runners of different performance levels compared with maximal aerobic capacity (i.e., V̇o2max) (2,19,33). Elliptical (12,21,27), CYCLE (20,21,34), and EBIKE (28) training in isolation allow runners to maintain their aerobic capacity compared with running-only (RUN) training. However, Honea (21) showed decrements in 3,000-m time trial performance after isolated CYCLE and elliptical training, decrements in RE after CYCLE training, but improvements in time trial performance for RUN training in high school cross country runners. It is therefore evident that limitations of certain cross-training modalities exist as it pertains to performance improvements. Not only have few studies assessed RE and running time trial performance before and after training using more than 2 cross-training modalities, no research has studied the effects of replacing some running volume with various cross-training modalities on these performance variables.
Regarding injury risks, training errors including too much intensity (29), high training volumes and poor recovery (39), and high rates of training load increases (18) seem to be important risk factors. Thus, replacing some runs with nonimpact or low-impact cross-training when increasing training intensity and volume may be useful in reducing injury risks. In addition, certain hip joint kinematics including hip adduction, hip internal rotation, and contralateral pelvic drop (10,36) seem to have some association with running injuries (e.g., anterior knee pain). Because reduced hip extensor and abductor muscle strength are associated with increased hip internal rotation and hip adduction excursion (42), improved hip muscle strength or neuromuscular control may help prevent excessive transverse and frontal plane hip joint motion associated with injuries in runners. Furthermore, clinical injury screening tools such as functional movement screen (FMS) may be useful in assessing injury risks in runners. Hotta et al. (23) reported that the deep squat (DS) and active straight leg raise (ASLR) tests of the FMS were predictive of injury development during a 6-month season in competitive young male runners. The ease of FMS test administration makes it useful for clinical testing by physical therapists and trainers for injury risk prediction and, potentially, prevention.
To date, the benefits of different modes of cross-training in runners on physiological, biomechanical, or injury-related parameters are not well understood, and this information would be greatly beneficial for coaches, athletes, and clinicians. Therefore, the purpose of this study was to compare movement quality, RE and performance, injury-related biomechanical variables, and hip muscle strength before and after 4 weeks of training with different cross-training modalities at the start of a cross country season in high school runners. We expected improvements in total FMS DS and ASLR for the EBIKE group but not for the other 3 groups. We also expected similar improvements in RE and 3,000-m performance for all groups, but greater improvements in hip muscle strength and reductions in the magnitude of injury-related hip kinematics for the EBIKE group but no changes in these measures for the other groups.
Experimental Approach to the Problem
Because many different aerobic cross-training modalities exist, we aimed to understand whether certain modalities were more effective than others to improve running performance and injury-related factors. We performed a within-group comparison of RE, 3,000-m performance, mobility, muscle strength, and running biomechanics in high school runners after 4 weeks of training with 4 modalities: RUN and running replacement with CYCLE, ELLIP, and outdoor EBIKE. This was performed at the beginning of a high school cross country season to understand the performance and injury risk implications of incorporating cross-training following summer months when training workloads are reduced.
Thirty-eight male high school cross country runners from the same team were recruited for the study. No a priori power analysis to assess the required sample size was performed because the sample size was limited to team roster size. Participants from the local cross country team were included in the study if they were free from any lower extremity injuries. The team's coach scheduled the cross-training sessions from this study as part of the cross country training schedule for a fall season. At the start of the study, participants attended a study information session with their parents. All participants were younger than 18 years and thus, parents or guardians were informed of testing procedures and expectations and provided written consent approved by the University of Memphis' Review Board. Subjects and parents were given an opportunity to ask for clarification of expectations and testing procedures before providing consent and throughout the duration of the study. Participants younger than 18 years also gave written consent to participate in the study.
A timeline of all testing and training procedures are presented in Figure 1. The study started 1 month into a high school cross country season. Thus, all runners had been performing regular team running practices for 1 month after returning from summer break. During summer break, the runners were instructed to only complete 3–4 easy runs at preferred speed each week but took at least 1 full week of no running. During week 1 of the study, a 3,000-m team race was first performed by all participants on their high school track during week 1 of the study. These race times were used to assess in-field performance and to stratify each training group by performance level. Participants performed the team race all at once to ensure maximal effort and completion times were recorded manually using a stopwatch.
During week 2, all participants attended a laboratory testing session. The pretraining testing sessions were completed over a 3-day period (Friday–Sunday), and each participant's testing session time was recorded. For laboratory testing days, participants were instructed not to complete a run before testing sessions and to consume their normal prerun meals or foods. Body mass and height were measured at the beginning of the study.
Following anthropometric measurements, FMS testing for only the DS and ASLR was performed using a standard FMS test kit (Functional Movement Screen, USA). Scoring for the 2 tests was based on previously described methods with possible scores between 0 and 3 (9). For the ASLR, the worst (i.e., lowest) score of the 2 legs was used in the total score (i.e., sum of DS and ASLR score (23)).
After FMS testing, hip abduction and hip extension maximal voluntary isometric contractions (MVICs) were performed using modified manual muscle testing (MMT) procedures for the right limb only, and a handheld dynamometer (Model 01165; Lafayette Instruments, Inc., Lafayette, IN, USA) was used to collect maximal isometric voluntary force (N). Hip abduction MVIC testing was first performed while participants lying supine on a therapy table (Figure 2A). Participants were instructed to rest their hands on their chest and to push maximal against the dynamometer for 4 seconds while not externally rotating their testing limb. A researcher held the dynamometer at a location of 5 cm directly above the lateral malleolus in a static position while the participants abducted their hip maximally. Another researcher stabilized the participants by placing a hand on the left side of the pelvis. Hip extension MVIC testing was then performed while participants stood on their nontesting leg with their hands resting on the strap and their testing their knee flexed to 90° with the lower leg supported by the researcher (Figure 2B). Participants then pushed maximally against the dynamometer placed posteriorly just above the knee for 4 seconds. The participants were instructed to use the strap only to stabilize and not to pull themselves forward. If participants pulled on the strap, they would become unstable and fall forward. Three MVIC trials per movement were completed on the right side, and 1-minute rest periods were provided between trials. The average force measurement from the 3 MVIC trials per movement was used for statistical analyses. Reliability of these modified MMT procedures was assessed on a subset of the participants (n = 10) 1 day apart using intraclass correlation coefficient (ICC) and Pearson's r correlation. These procedures showed strong day-to-day reliability (ICC = 0.98; r = 0.99) for hip abduction and hip extension (ICC = 0.91; r = 0.91) MVIC.
Participants then completed a 6-minute run on a treadmill (C962i; PRECOR, Woodinville, WA, USA) at a pace equivalent to 75% of their own 3,000-m average pace. An 8-camera motion analysis system (240 Hz; Qualisys AB, Göteborg, Sweden) was used to obtain three-dimensional (3D) kinematic data during the run. Before the run, retroreflective markers were placed on the right lower extremity and pelvis to obtain 3D kinematics during running (31). A 1-second static calibration trial was recorded before the start of data collection to define joint centers and segment coordinate systems and dimensions. Calibration markers were then removed, and participants were fitted with a facemask used to collect respiration gases during running using the metabolic system (ParvoMedics TrueOne, Murray, UT, USA). The gases and flow meter were calibrated before each testing session (<0.4% error). Expired gases were collected once steady-state was confirmed using the plateau of oxygen consumption. Four 30-second average V̇o2 data points were extracted during the 2-minute period and averaged together for statistical analyses. Kinematic data using the motion capture system were collected during a 10-second period, whereas oxygen consumption (ml·min−1·kg−1) was collected during the last 2 minutes of the treadmill run.
The exact same testing procedures were then repeated during week 9. Weeks 3 and 8 were reduced training load weeks that coincided with the periodization of their running training program and were used as an adaptation week to allow any potential training adaptations to occur before posttraining testing. The posttraining testing sessions were also completed over a 3-day period (Friday–Sunday), and each participant's testing session time during the day matched the pretraining testing time. Participants were reminded not to complete a run before completing the laboratory testing session and to consume their normal prerun meals or foods. The same researcher administered the FMS, MMT, and placed the reflective markers on the participant during both testing sessions.
During week 10, the 3,000-m team race was once again completed, but this time the race took place on a grass cross country course instead of a track. This was performed because the team was approaching their championship race and the coach wanted to ensure that the runners were exposed to a more cross country race-specific effort.
The training period took place during weeks 4–7 (Figure 1). A 3,000-m performance-stratified randomization was used to assign all participants to 1 of the 4 training groups. Because of sickness and family emergencies, 7 runners were unable to complete the posttraining tests and were excluded from the study. Final participant characteristics per group and number of runners per group are presented in Table 1.
During week 3, participants completed 2 practice sessions to become familiar with their respective cross-training modality. The CYCLE and EBIKE groups cycled around a 400-m track to practice starting, turning, and stopping on their respective modalities. All participants were assigned the same running training program by their coach, and the only aspect of training that was modified was the type of training performed for 2 easy runs each week (Table 2). The 2 weekly easy run sessions per week consisted of 25–35 minutes of submaximal effort equivalent to a Borg rating of perceived exertion between 11 and 13 (i.e., moderate to hard effort) (4). The RUN group completed an easy run as usual, whereas the ELLIP, CYCLE, and EBIKE groups performed exercise on their respective training modalities. For the ELLIP group, participants trained on a standard elliptical device (EFX546; PRECOR) with zero incline and were allowed to adjust the resistance and cadence to reach the prescribed effort. Participants were instructed to use the stationary arm rests for stability. For the CYCLE group, participants trained on hybrid and mountain bicycles with seat postheight adjusted to produce a knee flexion of approximately 25–30° to minimize (a) risk of knee injuries and (b) oxygen consumption (3). For the EBIKE group, participants trained on commercially available EBIKE (8C, ElliptiGO, Solana Beach, CA, USA). The CYCLE and EBIKE rode along a nearby flat paved CYCLE path along with 2 researchers who remained at the front and rear of the group to accommodate different cycle speeds among the participants.
Visual3D software (C-Motion, Inc., Germantown, MD, USA) was used to compute all biomechanical variables. A right-hand rule with a Cardan rotational sequence (xyz) was used for the 3D angular computations, where x represents the medial-lateral axis; y represents the anterior-posterior axis; and z represents the longitudinal axis. Kinematic data were interpolated using a least-squares fit of a third-order polynomial, with a 3 data point fitting and a maximum gap of 10 frames. Kinematic data were filtered using a fourth-order Butterworth low-pass filter at 8 Hz. The hip joint center positions were defined by 2 virtual markers located at 25% of the line between greater trochanter markers (43). The hip joint angular kinematic variables were expressed in the pelvis coordinate system. Spatiotemporal variables were also computed including cadence, (as a percent of leg length) step length, vertical oscillation of the whole body center of mass, and step reach (i.e., anterior position of the heel relative to the center of mass at foot strike). For all biomechanical variables, the average of all 7 stance phases for the right limb only was used in the statistical analyses. Time of foot strike was identified using the time of maximal downward pelvis velocity minus a 15-milliseconds offset (32), whereas toe-off was identified using the time of second peak knee extension (16).
Paired t-tests were used to compare dependent variables before and after training within groups. The alpha level was set at p ≤ 0.05 for all tests. Normality was assessed using the Shapiro-Wilk test. The 95% confidence intervals for mean differences were reported, and Cohen's d effect sizes were calculated to assess effect magnitudes using the interpretation of Hopkins (i.e., small: d < 0.6; moderate: 0.6 > d < 1.2; large: d > 1.2) (22).
For the FMS, the DS (p = 0.013, d = 0.87), ASLR (p = 0.035, d = 1.29), and aggregate score (p = 0.003, d = 1.36) were all improved after training only in the EBIKE group (Table 3).
For MMT, hip abduction MVIC force was unchanged for RUN, CYCLE, and EBIKE (p > 0.05), but showed small reductions for ELLIP after training (p = 0.04, d = 0.59; Table 2). Hip extension MVIC force was improved for RUN (p = 0.018, d = 1.25), CYCLE (p = 0.002, d = 1.17), and EBIKE (p = 0.043, d = 0.82), but not unchanged for ELLIP (p > 0.05) after training (Table 3). The RUN and CYCLE groups showed large effect sizes for changes in hip extension MVIC.
During treadmill running, none of the hip kinematic or spatiotemporal variables were different after training in any of the groups (Table 4).
Only the EBIKE group showed better RE (i.e., lower oxygen consumption) after training (p = 0.05, d = 0.46) (Table 4). For the 3,000-m team race, RUN (p = 0.02, d = 0.60), CYCLE (p = 0.006, d = 1.50), ELLIP (p = 0.01, d = 0.92), and EBIKE (p < 0.001, d = 1.41) all showed time to completion improvements after training (Table 4). The CYCLE and EBIKE groups showed large effect sizes for changes in 3,000-m completion times.
This study compared changes in performance and injury-related variables among different modes of aerobic cross-training in high school runners starting 1 month into their fall cross country season. The findings from this work are novel because no study has previously compared the effectiveness of substituting some running training with cross-training using 3 different modalities on performance and injury-related variables in high school runners following the same running program.
Our hypothesis that total FMS DS and ASLR score would be improved after training for the EBIKE group but not for the other 3 groups was supported. No other studies have assessed changes in FMS scores following a cross-training intervention in runners. Moderate-to-large EBIKE training effects were observed for the total (d = 1.36), DS (d = 0.87), and ASLR (d = 1.29) scores, suggesting meaningful improvements in lower limb mobility. Competitive male runners between 18 and 24 years of age can be up to 10 times more likely to develop a running injury when the DS and ASLR aggregate score was equal to or below 3 compared with scores of 4 or above (23). On average, the DS and ASLR aggregate scores were above 3 for all groups before and after training which may suggest lower chances of injury development in our high school runner population (Table 3). The 1-point improvement from 3.9 to 4.9 in the EBIKE group may have important implications to further reduce injury risks in male high school runners. Deep squat assesses ankle, hip, and thoracic spine mobility in addition to closed chain lower limb coordination, whereas ASLR assesses hamstring and plantarflexor flexibility with a stable pelvis position (8,9). The 63.5-cm step length of EBIKE is 18–23 cm longer than most stationary elliptical trainers and may require greater ankle and hip joint excursions throughout the pedal cycle. Compared with seated CYCLE where hip motion is limited to flexion (15), EBIKE and ELLIP produce flexion to extension hip range of motion while weight-bearing. Thus, the longer step length of EBIKE compared with ELLIP and the flexion to extension hip range of motion that is lacking during CYCLE may have contributed to improved mobility of the ankle, hip, and thoracic mobility. The step length of EBIKE is 25 cm shorter than a typical step length during slower running (37) and should therefore yield small joint excursions, which could imply a reduced impact on mobility improvements compared with running. However, with continued running instead of cross-training on easy running days, the most eccentric extensor muscle action during running may lead to greater muscle fiber damage (6) and as a result, reduced muscle flexibility. Furthermore, ASLR is influenced by abdominal wall muscle activity (25) and may reflect lumbar spine stability (30) and ultimately, pelvic control. Therefore, improved ASLR scores may be due in part by the fact that EBIKE is inherently more unstable in the frontal plane compared with RUN, CYCLE, and ELLIP and may require greater abdominal wall and pelvic muscle control to maintain stability while riding. To date, no studies have compared sagittal plane joint kinematics among EBIKE, CYCLE, and ELLIP and therefore, these interpretations are based on observations and extrapolations from previous findings. In the future, these data may help understand the kinematic mechanisms involved in changing lower limb mobility after a period of cross-training. Although the current findings of improved FMS DS and ASLR aggregate scores indicate a potential lower risk of injury development in high school runners, these findings cannot be generalized to other runner populations at this time. It is also important to note that longer exposures to cross-training modalities may introduce other injury risks factors (e.g., traumatic injuries from crashes and other overuse injuries).
Contrary to our hypothesis that RE would be similarly improved in all training groups, RE was only improved after training, including EBIKE. However, because the improvement only showed a small effect size (d = 0.46), the meaningfulness of this finding is questionable. Although no kinematic data were collected during training on the different modalities, we suspect that greater side-to-side motion due to instability on the EBIKE compared with CYCLE, ELLIP, and RUN may be related to the small RE improvement. The increased instability would require greater core muscle involvement to ensure whole-body stability and as a result, greater oxygen requirements during EBIKE exercise. Oxygen requirements are in fact on average 33% higher than CYCLE, but similar to running at running speeds equivalent to easy to moderate paces in our current population (i.e., 12.1–13.8·km·h−1) (14). As hypothesized, 3,000-m performance was improved for all groups, but the CYCLE (d = 1.50) and EBIKE (d = 1.41) groups showed large effect sizes, whereas RUN (d = 0.60) and ELLIP (d = 0.92) showed moderate effect sizes. Although running training would be expected to produce the largest improvements in running performance, the replacement of 2 runs per week with cross-training may allow for improved recovery between more intense running sessions allowing for a greater training adaptation over a 4-week period. The unchanged RE and improved 3,000-m performance suggest that in the early stages of the cross country season, high school runners can replace some of their easy runs with cross-training, ideally CYCLE or EBIKE, with no decrements in RE and some improvements in running performance.
These findings are similar to the findings of Flynn et al. (17) who did not observe changes in RE after 6 weeks of RUN or running with CYCLE training but reported similar improvements in 5-km performance in both groups of trained runners. Furthermore, submaximal and maximal aerobic capacity in addition to 5,000 and 1,609 m performance improved similarly after 5 weeks of RUN or running with CYCLE training in trained runners (35). It is important to note that the motion of CYCLE differs greatly from running as it places the knee and hip joints in much more flexed positions. These greater joint flexion positions during CYCLE changes the contraction lengths of muscles involved and seem to subsequently alter stride mechanics, and torso, pelvis, and hip motions immediately after CYCLE (38). Unchanged RE and maximal aerobic capacity have also been reported after 4 weeks of CYCLE-only training in recreational runners (34), whereas others found reduced 3,000-m performance in high school runners after 4 weeks of ELLIP-only training (21). Recently, Klein et al. (28) observed no changes in RE or 5,000-m time trial performance after 4 weeks of EBIKE-only and RUN training in experienced runners. These findings suggest that CYCLE-only and EBIKE-only training, but not ELLIP-only training, can be used to maintain aerobic fitness in runners who are unable to run perhaps due to injury. This may also have implications for fitness maintenance during periods of planned rest with no running. However, although weight-bearing in the case of ELLIP and EBIKE, the most concentric muscle action produced during cross-training is different from the most eccentric muscle action of joint extensors during running (44). It has been well documented that a high frequency of unaccustomed eccentric muscle actions leads to muscle damage and often, muscle soreness (5,13). Return to running after extended periods of cross-training only training may lead to lower extremity muscle damage or soreness. Building up early season running volume may be performed more safely while maintaining aerobic fitness with both running and cross-training sessions when coming back from periods of no running (e.g., injury or rest periods). Future work should assess muscles soreness and damage during volume buildup periods of cross-training with and without running.
Contrary to our hypothesis, EBIKE did not yield greater improvements in hip muscle strength or reductions in the magnitude of injury-related hip kinematics. During running, hip abductors (i.e., gluteus medius), extensors, and external rotators (i.e., gluteus maximus) work to control hip adduction and contralateral pelvic drop, and hip internal rotation, respectively. In fact, reduced hip extensor and abductor muscle strength are associated with increased hip internal rotation and hip adduction excursion in young runners (i.e., 18 years) (42). In the current study, although we observed improved hip extensor muscle strength in RUN, CYCLE, and EBIKE, no changes in hip internal rotation or hip abduction were found. This suggests that changes in hip internal rotation and hip abduction excursion may be the result of other control mechanisms. In addition, ELLIP training yielded small reductions (d = 0.59) in hip abductor strength, but no changes in frontal plane hip kinematics. This may be explained by the finding that hip abductor strength was not associated with frontal plane hip kinematics during running in young adults (i.e., 29 years) (1). Finally, the current findings of improved hip extensor strength after RUN, CYCLE, and EBIKE training may provide positive training benefits because hip extensors play an important role in joint power production as running speed increases (41) and control of hip kinematics (42). The increase in hip extensor strength may have contributed to the moderate to large improvements in 3,000-m performance in RUN, CYCLE, and EBIKE groups. However, more research is needed to understand the direct impact of increasing hip extensor muscle strength on running performance.
This study has a few limitations that must be discussed. First, the training period was short with only 4 weeks with the training period limited to 8 total training sessions (i.e., 2 per week). However, the purpose of the study was to assess how different cross-training modalities would affect the dependent variables during a period of volume buildup at the start of the cross country season and not throughout the entire cross country season. Future work could study the changes in performance and injury-related variables after longer periods of exposure to cross-training in addition to or as a replacement for some running volume of varied intensities. In addition, the second 3,000-m team race was performed on a grass course, whereas the first race was completed on a track. Although the surfaces were different, all runners completed the 3,000-m races under the same conditions, and the course distance was accurately measured for cross country races and was a commonly used training location for the cross country team. The pre- and post-training difference magnitude assessed with Cohen's d allowed for quantifications of group-specific training effects on 3,000-m race performance.
This study compared the effectiveness of substituting some easy running training during the early season (i.e., September) with cross-training using 3 different modalities on performance and injury-related variables in male runners from the same high school team. Three thousand meter performance improved after training on all modalities, but large effects were only found for the outdoor CYCLE and EBIKE groups. RUN, outdoor CYCLE, and EBIKE all showed improvements in maximal hip extensor strength, whereas the ELLIP group produced lower maximal hip abductor strength. Outdoor EBIKE training was the only modality that improved mobility after the training period. These findings suggest that high school running coaches could consider replacing 2 easy runs per week with outdoor CYCLE or EBIKE cross-training to improve running performance in runners during early season training when runners return from summer break.
The authors have no professional relationships with companies or manufacturers that might benefit from the results of the study. There is no financial support for this project, and no funds were received for this study. The results of this study do not constitute endorsement of any product by the authors or the National Strength and Conditioning Association.
1. Baggaley M, Noehren B, Clasey JL, Shapiro R, Pohl MB. Frontal plane kinematics of the hip during running: Are they related to hip anatomy and strength?. Gait Posture 42: 505–510, 2015.
2. Barnes KR. Running economy: Measurement, norms, and determining factors. Sports Med Open 1: 8, 2015.
3. Bini R, Hume PA, Croft JL. Effects of bicycle saddle height on knee injury risk and cycling performance. Sports Med 41: 463–476, 2011.
4. Borg GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc 14: 377–381, 1982.
5. Byrne C, Twist C, Eston R. Neuromuscular function after exercise-induced muscle damage: Theoretical and applied implications. Sports Med 34: 49–69, 2004.
6. Clarkson PM, Nosaka K, Braun B. Muscle function after exercise-induced muscle damage and rapid adaptation. Med Sci Sports Exerc 24: 512–520, 1992.
7. Conley DL, Krahenbuhl GS. Running economy and distance running performance of highly trained athletes. Med Sci Sports Exerc 12: 357–360, 1980.
8. Cook G, Burton L, Hoogenboom BJ, Voight M. Functional movement screening: The use of fundamental movements as an assessment of function-part 2. Int J Sports Phys Ther 9: 549–563, 2014.
9. Cook G, Burton L, Hoogenboom BJ, Voight M. Functional movement screening: The use of fundamental movements as an assessment of function—part 1. Int J Sports Phys Ther 9: 396–409, 2014.
10. Dierks TA, Manal KT, Hamill J, Davis IS. Proximal and distal influences on hip and knee kinematics in runners with patellofemoral pain during a prolonged run. J Orthop Sports Phys Ther 38: 448–456, 2008.
11. Doherty RA, Neary JP, Bhambhani YN, Wenger HA. Fifteen-day cessation of training on selected physiological and performance variables in women runners. J Strength Cond Res 17: 599–607, 2003.
12. Egana M, Donne B. Physiological changes following a 12 week gym based stair-climbing, elliptical
trainer and treadmill running program in females. J Sports Med Phys Fitness 44: 141–146, 2004.
13. Enoka RM. Eccentric contractions require unique activation strategies by the nervous system. J Appl Physiol (1985) 81: 2339–2346, 1996.
14. EPARC EaPARC-. Comparing the Energy Expenditure of Riding the ElliptiGO Elliptical
Bicycle to Conventional Cycling and Running. Online: For ElliptiGO, Inc.: University of California San Diego Solana Beach, CA: ElliptiGO, 2011.
15. Ercison MO, Nisell R, Nemeth G. Joint motions of the lower limb during ergometer cycling. J Orthop Sports Phys Ther 9: 273–278, 1988.
16. Fellin RE, Rose WC, Royer TD, Davis IS. Comparison of methods for kinematic identification of footstrike and toe-off during overground and treadmill running. J Sci Med Sport 13: 646–650, 2010.
17. Flynn MG, Carroll KK, Hall HL, Bushman BA, Brolinson PG, Weideman CA. Cross training: Indices of training stress and performance. Med Sci Sports Exerc 30: 294–300, 1998.
18. Gabbett TJ, Hulin BT, Blanch P, Whiteley R. High training workloads alone do not cause sports injuries: How you get there is the real issue. Br J Sports Med 50: 444–445, 2016.
19. Heise GD, Martin PE. Are variations in running economy in humans associated with ground reaction force characteristics?. Eur J Appl Physiol 84: 438–442, 2001.
20. Hoffmann JJ, Loy SF, Shapiro BI, Holland GJ, Vincent WJ, Shaw S, Thompson DL. Specificity effects of run versus cycle training on ventilatory threshold. Eur J Appl Physiol Occup Physiol 67: 43–47, 1993.
21. Honea DM. The impact of replacing run training with cross-training on performance of trained runners. In: Department of Health, Leisure, and Exercise Science. Boone, NC: Appalachian State University, 2012. pp. p 87.
22. Hopkins WG. Effect Statistics: A New View of Statistics Internet Society for Sport Science, 2013.
23. Hotta T, Nishiguchi S, Fukutani N, Tashiro Y, Adachi D, Morino S, Shirooka H, Nozaki Y, Hirata H, Yamaguchi M, Aoyama T. Functional movement screen for predicting running injuries in 18- to 24-year-old competitive male runners. J Strength Cond Res 29: 2808–2815, 2015.
24. Houmard JA, Hortobagyi T, Johns RA, Bruno NJ, Nute CC, Shinebarger MH, Welborn JW. Effect of short-term training cessation on performance measures in distance runners. Int J Sports Med 13: 572–576, 1992.
25. Hu H, Meijer OG, van Dieen JH, Hodges PW, Bruijn SM, Strijers RL, Nanayakkara PW, van Royen BJ, Wu W, Xia C. Muscle activity during the active straight leg raise (ASLR), and the effects of a pelvic belt on the ASLR and on treadmill walking. J Biomech 43: 532–539, 2010.
26. Jones AM. The physiology of the world record holder for the women's marathon. Int J Sports Sci Coach 1: 101–116, 2006.
27. Joubert DP, Oden GL, Estes BC. The effects of elliptical
cross training on VO2max in recently trained runners. Int J Exerc Sci 4: 4–12, 2011.
28. Klein IE, White JB, Rana SR. A comparison of physiological variables between the elliptical
bicycle and run training in experienced runners. J Strength Cond Res 30: 2998–3006, 2016.
29. Kluitenberg B, van der Worp H, Huisstede BM, Hartgens F, Diercks R, Verhagen E, van Middelkoop M. The NLstart2run study: Training-related factors associated with running-related injuries in novice runners. J Sci Med Sport 19: 642–646, 2016.
30. Liebenson C, Karpowicz AM, Brown SH, Howarth SJ, McGill SM. The active straight leg raise test and lumbar spine stability. PM R 1: 530–535, 2009.
31. McClay I, Manal K. Three-dimensional kinetic analysis of running: Significance of secondary planes of motion. Med Sci Sports Exerc 31: 1629–1637, 1999.
32. Milner CE, Paquette MR. A kinematic method to detect foot contact during running for all foot strike patterns. J Biomech 48: 3502–3505, 2015.
33. Morgan DW, Bransford DR, Costill DL, Daniels JT, Howley ET, Krahenbuhl GS. Variation in the aerobic demand of running among trained and untrained subjects. Med Sci Sports Exerc 27: 404–409, 1995.
34. Moroz DE, Houston ME. The effects of replacing endurance running training with cycling in female runners. Can J Sport Sci 12: 131–135, 1987.
35. Mutton DL, Loy SF, Rogers DM, Holland GJ, Vincent WJ, Heng M. Effect of run vs combined cycle/run training on VO2max and running performance. Med Sci Sports Exerc 25: 1393–1397, 1993.
36. Noehren B, Davis I, Hamill J. ASB clinical biomechanics award winner 2006 prospective study of the biomechanical factors associated with iliotibial band syndrome. Clin Biomech (Bristol, Avon) 22: 951–956, 2007.
37. Petersen J, Sorensen H, Nielsen RO. Cumulative loads increase at the knee joint with slow-speed running compared to faster running: A biomechanical study. J Orthop Sports Phys Ther 45: 316–322, 2015.
38. Rendos NK, Harrison BC, Dicharry JM, Sauer LD, Hart JM. Sagittal plane kinematics during the transition run in triathletes. J Sci Med Sport 16: 259–265, 2013.
39. Ristolainen L, Kettunen JA, Waller B, Heinonen A, Kujala UM. Training-related risk factors in the etiology of overuse injuries in endurance sports. J Sports Med Phys Fitness 54: 78–87, 2014.
40. Running U. 2015 U.S. Road Race Trends—By the Numbers: Running USA, 2016. Available at: http://runningusa.org
41. Schache AG, Brown NA, Pandy MG. Modulation of work and power by the human lower-limb joints with increasing steady-state locomotion speed. J Exp Biol 218: 2472–2481, 2015.
42. Taylor-Haas JA, Hugentobler JA, DiCesare CA, Hickey Lucas KC, Bates NA, Myer GD, Ford KR. Reduced hip strength is associated with increased hip motion during running in young adult and adolescent male long-distance runners. Int J Sports Phys Ther 9: 456–467, 2014.
43. Weinhandl JT, O'Connor KM. Assessment of a greater trochanter-based method of locating the hip joint center. J Biomech 43: 2633–2636, 2010.
44. Williams DS III, Green DH, Wurzinger B. Changes in lower extremity movement and power absorption during forefoot striking and barefoot running. Int J Sports Phys Ther 7: 525–532, 2012.