Tibial stress fractures are a serious and common overuse injury problem among running and military populations (36,39). Unfortunately, when individuals sustain a tibial stress fracture, the rehabilitation period to make a full recovery is typically around 6–12 wk (23). This prolonged disruption of training time associated with tibial stress fracture can lead to reductions in physical fitness and frustration (2). It has been recognized that the development of tibial stress fracture is a result of the failure of bone to adapt to repetitive subthreshold loads associated with running (28). This failure occurs when individuals do not allow for sufficient bone remodeling to repair the microcracks accumulated from the repetitive loading (19). Stress fractures are a result of an overload in mechanical force, and researchers believe stress fractures are related to the magnitude of strain and loads applied to the bone (27). Edwards et al. (16) explain that when bone load magnitudes are low, microdamage accumulation is limited, therefore allowing tissues adequate time to repair and remodel. However, if the load magnitudes are higher, then there is an increased likelihood that the micro damage will overwhelm the repair process of the bone, causing a greater stress fracture risk.
Studies that have examined external impact loading (such as ground reaction forces and shank acceleration measures over the underlying tibia) during running have found these variables to be linked with increased tibial stress fracture risk. For example, a prospective report showed that runners who sustained a tibial stress fracture elicited greater vertical force average and instantaneous loading rates (vertical average loading rate [VALR] and vertical instantaneous loading rate [VILR]) and higher peak tibial axial acceleration (PTA) during stance compared with runners not sustaining a tibial stress fracture injury (10). Similarly, retrospective reports have also shown greater load rates, increased vertical force impact peaks (IP), and PTA in runners who had a previous history of a tibial stress fracture and other overuse injuries (24,30). The strong association between high impact loading and tibial stress fracture injury potential has encouraged researchers to find effective methods of reducing these loads during running. Reports have found significant reductions in impact loading when altering shoe characteristics during running (4), whereas others have reported similar reductions from shoe insoles (34) and orthotics (15). With the search to find more effective ways of reducing these impact loads linked with tibial stress fracture risk, researchers have recently implemented the use of real-time feedback (RTF) interventions to clinically gait retrain runners (7,8,33). In a preliminary study, Crowell et al. (8) reported significant reductions in impact loading variables when five runners completed a single RTF gait retraining session. The gait retraining administered involved a visual presentation of each subject’s tibial accelerations on a monitor in front of the treadmill. Runners were instructed to run below a horizontal line that represented 50% of their PTA values recorded preintervention. In a follow-up study, these researchers conducted a 2-wk intervention (eight sessions) using the same RTF (7). The main objective of the study was to investigate the retention of any gait adaptations by assessing running performance 1 month postintervention. Their results showed that the feedback was successful at significantly reducing impact loads 1 month postintervention. However, these studies are significantly limited by not using a control group and not investigating the underlying mechanical strategies used by runners to help explain the reductions observed in impact loading. In addition, little is known about the influence of this type of RTF gait retraining on running economy because research in this area has primarily concerned itself with injury reduction or prevention while neglecting performance-related outcomes. It could be hypothesized that gait adaptations leading to greater attenuation of impact shock may be linked with a concomitant increase in metabolic cost. Therefore, to gain a more complete understanding of the overall efficacy of this RTF approach, any influence on metabolic cost must also be investigated.
The aim of the study was to determine the influence of RTF training on impact loading and running economy. It was hypothesized that impact loading would significantly reduce after the RTF training, and these reductions would be maintained 1 month postintervention, whereas the control group would remain consistent throughout. Similarly, we hypothesized that running economy would worsen in response to RTF training, whereas the control group would be unaffected. Several reports have identified the important influence of landing mechanics in the control of impact loading during running (5,13,20,21,25). Giandolini et al. (21) showed significant reductions in impact loading when runners altered their foot strike pattern from a rearfoot to a midfoot strike. However, others showed reductions in impact loading with the increased knee flexion angle at initial contact (13) along with lower heel vertical velocity at initial contact (20). Given this evidence, another aim of this study was to investigate the mechanical strategies, that is, joint mechanics at initial contact responsible for the modifications in impact loading variables. It was hypothesized that the RTF-grouped runners would show mechanical changes because of the RTF training. Subsequently, it was expected that the control group would show consistent gait mechanics throughout testing.
METHODS
Participants
Thirty-six recreational rearfoot striking male runners volunteered as participants for the study. All participants were free from musculoskeletal injury at the time of testing and signed a written informed consent form as approved by the ethics review board of the university. Subjects were excluded if they had a history of any major musculoskeletal gait modifying injuries (such as an anterior cruciate ligament tear) and cardiovascular pathology. All subjects recruited for this study had to be typically running at least 30 km·wk−1 and have been free from injury 6 months before testing. Before runners could participate in the study’s training phase, they were required to meet a specific inclusion criterion similar to the prescreening previously used by Crowell et al. (7,8). Runners who elicited average PTA value greater than 9g during the prescreening assessments were allowed to enter into the training phase of the study. The prescreening assessments consisted of six overground running trials at 3.7 m·s−1 (±5%) over a 15-m runway. This baseline value of 9g was selected based on a prospective report (10), which showed that runners who elicited PTA values greater than 9g were deemed more likely to sustain a tibial stress fracture. Subsequently, out of the 36 runners screened, 29 proceeded into the next training phase. Participants were randomly assigned into two groups, an RTF training group (n = 15) and a control (CON) group (n = 14). During the testing period, seven participants dropped out of the study because of personal complications, and so a total of 22 participants (Table 1) completed the full study testing protocol.
;)
TABLE 1: Participants’ group characteristics (mean ± SD).
Research design
The two groups who completed the testing protocol all performed a pretraining (Pre), posttraining (Post, within 1–2 d), and 1-month posttraining screening assessments. These screening assessments consisted of a three-dimensional gait analysis along with a 5-min treadmill (Power H-P Cosmos, Germany) run for measurements of running economy.
Gait analysis
A triaxial (2.8g, sensitivity range of ± 0.67 V·g−1) accelerometer (2400T G2, Noraxon, Arizona, USA) was securely attached over each participant’s right distal anteromedial aspect of the tibia bone. One of the three axes of the accelerometer was aligned with the long axis of the bone. To minimize any unwanted overshoot oscillations, the skin was stretched with water repellent adhesive tape (kinesiology tape; Vivomed, UK) before the accelerometer was mounted (5). Markers were placed over skin, tight-fitting clothing, and participant’s shoes. All participants wore tight-fitting lycra shorts to reflect the motion of the underlying segments. The acceleration data were sampled at 1500 Hz. Kinematic data were collected using a 12 infrared camera motion capture system (Oqus 3; Qualisys, Sweden) operating at 240 Hz. A retroreflective passive marker set was applied to the right lower extremity, including the pelvis segment of each subject. Anatomical markers were attached at iliac crests, anterior superior iliac spines, and the greater trochanters to define and track the pelvic segment. To define the thigh, shank, and foot segments, the following anatomical markers were attached to the medial and lateral femoral condyles, medial and lateral malleoli, and the first and fifth metatarsal heads; four noncollinear tracking markers on molded thermoplastic shells were secured on the lateral aspects of both thigh and shank segments. Three additional foot tracking markers were placed over the heel and the first and fifth metatarsal bases. All markers and accelerometer attachments were conducted by the same experimenter.
A 1-s static standing trial was recorded. In this trial, the participant stood with feet approximately shoulder width apart, knees fully extended, hips neutral, and trunk vertically upright. During the dynamic trials, the greater trochanters medial and lateral femoral condyles and medial and lateral malleoli markers were all removed leaving the remaining markers for tracking. The functional method approach was used to determine hip joint centers as previously described (3). A force platform (Kistler Instruments Limited, Switzerland) synchronized with motion and accelerometer systems was used to collect kinetic data at 1200 Hz. Subject’s running velocities were monitored by two pairs of photo gates (Brower Timing Systems, UK) positioned in the filming volume 2 m apart. Participants performed six acceptable overground running trials at 3.7 m·s–1 (±5%) along a 15-m runway. Trials were discarded if participants appeared to target the force platform. Before the overground running, participants performed a 5-min self-selected warm-up jog along with practice trials to familiarize themselves with the required running velocity. Participants wore their own shoes throughout the testing period and were asked to refrain from running for 24 h before testing.
Markers were all tracked and labeled in Qualisys Track Manager (Qualisys) and then exported to Visual3D (C-motion Inc., USA) for further processing and analysis. A six-degrees-of-freedom model of the lower limb extremities and pelvis was generated to calculate three-dimensional bilateral segment poses (position and orientation) for the pelvis, thigh, shank, and foot. All the lower extremity segment masses were based on Dempster’s (12) data, and each segment’s moments of inertia were represented as a frustum of cones. A local segment coordinate system was determined for each segment. Cardan Euler angles were calculated using an X–Y–Z (flexion–extension, abduction–adduction, and internal–external rotations) rotation order to describe the motion of the distal segments relative to the proximal. All joint angles, velocities, and moments were calculated using a right-hand rule. The static standing trial was used to define joint angles as 0°. On the basis of residual analyses, kinematic data were filtered using a low-pass, fourth-order Butterworth filter at 12 Hz (43), whereas force and acceleration data were filtered at 70 and 60 Hz, respectively. Joint moments were all normalized to body mass. Stance phase was defined as the period when the vertical force was higher than 10 N. Stance phase curves were time normalized to 100% using cubic spline interpolation.
A custom written LabVIEW program (National Instruments, UK) was used to extract IP, VALR, VILR, and PTA variables. The filtered PTA was defined to be the peak positive axial acceleration component during the stance phase of running. Vertical loading rates were calculated between 20% and 80% of initial contact to IP (30). If runners did not display an IP during stance, loading rates were calculated between 20% and 80% in the first 50 ms of stance. This period is defined as the impact phase of stance (32). Foot strike angle was determined using the kinematic method outlined by Altman and Davis (1). Rearfoot strike pattern ≥ 8.0°, midfoot strike ≥ −1.6° and <8.0°, and forefoot strike ≤ −1.6°. Descriptive foot to ground angle curves were calculated from initial contact to 50% of stance. 0° corresponded with a flat foot (horizontal to the ground). Sagittal plane joint kinematics were determined at initial contact. In addition, heel vertical velocity (HVV) at initial contact was also determined. Positive sagittal plane joint angles were defined in the flexion/dorsiflexion direction, whereas those in extension/plantarflexion were defined as negative.
RE assessment
After completion of the gait analysis assessments, participants performed a 5-min treadmill run (1% gradient) at 3.7 m·s–1. Before the commencement of this test, each subject conducted a 5- to 6-min familiarization run to ensure that they felt comfortable wearing a tightly fitted facemask along with achieving a natural running style (25). During the run, measures of the rate of oxygen consumption (V˙O2) were taken continuously for every 2 s using an online gas analysis system (Pneumotech Quark cPET, Cosmed Italy). The averaged V˙O2 values for the last min were used to calculate running economy (as runners were deemed at a steady state of running at this stage). The units of running economy were expressed in milliliters per kilogram per kilometer as outlined by Winter (44). Once the test was completed, all participants performed a 3-min self-paced cooldown.
Training intervention
The RTF training consisted of six sessions over a 3-wk period (two sessions per week). Participants were allowed to run outside the training sessions but were not allowed to run 24 h before or on test days. This allowance was to ensure participants retained their typical weekly training regime (three to four runs per week) and to prevent the influence of fatigue on results. All included participants completed all six sessions. Each session lasted approximately 35 min. At the start of testing, an accelerometer was attached to subject’s anteromedial distal aspect of the tibia (as previously described). In each session, participants ran on a treadmill for 20 min at 3.7 m·s–1 with continuous RTF. Before and after every run participants completed a no-feedback 5-min warm-up and cooldown self-paced jog. The feedback variable was based on participants PTA values during running. A peak detection algorithm was used to identify the PTA during stance. The algorithm accepted a peak if it was higher than 3g, and the new streamed data values were less than −1g. However, if these conditions were not met, then the algorithm would ignore the current peak and search for the next available peak within the real-time data stream. Once an acceptable PTA value was detected, it was stored in a continuous loop, which contained five previous PTA values. The feedback was calculated on a moving average of five previous PTA values and was presented to each participant every fifth consecutive stride during the run. The decision to provide feedback on every fifth stance period was made based on research indicating benefits to motor skill learning with reduced feedback frequency (38). This approach was used to shift their dependence from external to internal cues and thus optimize retention (42).
A Qualisys Track Manager LabVIEW client plug-in (Qualisys) was used to stream the analog data into a custom-written LabVIEW feedback program. The visual feedback was projected onto a large screen 5 m in front of the treadmill while auditory feedback was delivered via external speakers from the main laboratory PC. Both visual and auditory stimuli were simultaneously provided. The baseline PTA values were used as a criterion to define the three feedback modalities of high shock, medium shock, and acceptable shock levels. The high shock level indicated PTA values greater than 75% of baseline PTA values, and between 50% and 75% of baseline PTA values represented the medium shock range, whereas baseline PTA values less than 50% indicated the acceptable shock range. The visual feedback format was set up using traffic light symbology (Fig. 1). When participants ran in the high shock range, a red light and a high-pitched sound was presented. For the medium shock range, an amber light and a low-pitched sound was delivered. A green light but with no sound indicated participants were running at the acceptable shock range. In addition, participants were visually presented with their averaged PTA values (adjacent to traffic lights) at the same time as other visual and auditory feedback modalities. At the start of the feedback training, participants were instructed to find a “strategy or a way” to run within the acceptable shock range. No instructions were provided on how to do this, as the experimenter wanted runners to elicit self-discovery strategies to achieve this outcome (26).
FIGURE 1: A participant during an RTF gait retraining session. Visual and audio feedback were presented to participants in front of the treadmill while they ran.
The control group all performed the same running protocol, that is, six sessions of 20 min treadmill running for a 3-wk period, as the RTF training group but without feedback or instruction. Testing for the control group was conducted in the same location and with the same equipment as the RTF group. This meant participants performed on the same treadmill, along with the comparable laboratory setup as RTF group, that is, projector screen in front of treadmill. During the period between the Post and 1-month posttests, all participants were instructed to maintain their typical weekly training regime, that is, three to four runs per week (≥30 km·wk−1).
Data analysis and statistics
For each subject, all kinematic and kinetic variables were taken from individual trials and then averaged across the six trials, creating an individual subject mean. All key kinematic and kinetic variables were analyzed using a 2 × 3 (group × time) mixed general linear model, where the within-subject factor of time was set across three levels (Pre, Post, and 1-month posttraining) and the between-subject factors were set across two levels (RTF group = RTF and control group = CON). To detect any violations of the sphericity assumption within the data, Mauchly’s test of sphericity was carried out, and where this test was significant, a Greenhouse–Geisser correction was used. After significant group-by-time interactions, post hoc independent-sample t-test comparisons were conducted (two-tailed) between the groups in their change (Δ) values Pre and Post and again between Pre and 1-month posttest. For all analyses, the α level was set at P < 0.05 throughout. The magnitude of differences between Pre and Post and from Pre to 1-month posttest for gait variables were also represented using Cohen’s d effect sizes (ES) (6). The ES was defined as small, <0.40; moderate, 0.40–0.79; and large, ≥0.8.
RESULTS
Impact loading
Out of all the impact loading variables, PTA, VALR, and VILR all showed a statistically significant group-by-time interaction effect (P < 0.05) (Table 2).This interaction was a result of the RTF group showing a decrease in these variables, while the CON group remained relatively consistent throughout. The group-by-time interaction from Pre to Post tests showed that the RTF group had significant reductions (P < 0.05) in PTA (−31%), VALR (−18%), and VILR (−19%) compared with the CON group. In addition, the RTF group showed moderate to large ES (0.75–1.50) in these variables from Pre to Post tests. Although PTA showed a significant reduction (P < 0.001) from Pre to 1-month posttests, this was not evident in VALR and VILR. This lack of change could be explained by VALR and VILR values returning toward the prescreening baseline values during the month after feedback cessation, with increases of 12% and 10%, respectively. There were no significant changes in IP across time and groups.
TABLE 2: Impact loading variables for RTF and control groups (mean ± SD).
Mechanical strategies
Statistically significant group-by-time effect interactions (P = 0.0033, P = 0.030) were found in the ankle angle at initial contact and foot strike angle (Table 3). The ankle angle at initial contact showed a significant change (P = 0.04) to a more plantarflexed position from Pre to 1-month posttest for the RTF group compared with the control group. This increase in plantarflexion in early stance after the gait retraining is illustrated in Figure 2 (A). Similarly, foot strike angle significantly changed (P = 0.04) from a rearfoot strike pattern to a midfoot strike pattern Pre to 1-month posttests in the RTF group compared with the control group. Despite the observed changes in the ankle and foot posture, both hip and knee joints showed no significant changes because of the RTF training. A statistically significant group-by-time interaction effect (P = 0.013) was found in HVV at initial contact. The RTF group significantly reduced (P = 0.008) their HVV Pre to Post tests compared with the control group who remained consistent. However, no significant change in HVV at initial contact was found between Pre and 1-month posttests. In summary, the variables that were significantly modified with RTF training were ankle angle at initial contact and HVV at initial contact. These gait mechanical strategies were all associated with ankle and foot modifications and not at the more proximal knee and hip joints.
TABLE 3: Lower limb joint angles, foot strike angle, and HVV at initial contact for RTF and control groups (mean ± SD).
FIGURE 2: A. Mean ankle (sagittal plane) joint angle curves of the RTF group. B. Mean ankle (sagittal plane) joint moment curves of the RTF group. Pre (black line), Post (gray line), and 1-month posttests (dotted line). C. Mean foot to ground angles of the RTF group at Pre (black line), Post (gray line), and 1-month posttest (dotted line) assessments (foot to ground angle is zoomed to 50% of stance phase). The RTF group showed a kinematic adaptation from a rearfoot strike pattern in the Pre to a less pronounced rearfoot strike pattern in the Post and 1-month posttests.
Running economy
No statistically significant differences were reported in running economy across time or group-by-time interactions (Fig. 3). Considering the observed gait mechanical changes in response to the RTF training, running economy was shown to be unaffected in the RTF group (Pre = 216.55 ± 17.84 mL·kg−1·km−1, Post = 213.39 ±12.92 mL·kg−1·km−1, 1-month posttest = 206.58 ± 34.09 mL·kg−1·km−1).
FIGURE 3: Mean (SD) RTF and CON group RE values for Pre, Post, and 1-month posttest assessments.
DISCUSSION
The primary purpose of this study was to determine whether gait retraining, by the use of auditory and visual RTF, would reduce impact loading during running and if these reductions would persist for one month. The results of this study showed significant reductions in PTA, VALR, and VILR variables after (Post) the RTF training compared with the control group who remained consistent. These reductions, however, only persisted in PTA as VALR and VILR increased toward baseline levels 1 month after RTF training. Previous studies reported similar reductions in PTA, VALR, and VILR after RTF gait retraining (7,33). However, the previously reported reductions in impact loading with gait retraining only demonstrated improvements within the intervention group and were not compared with a control group. Moore (31) has shown how gait mechanical strategies can self-optimize in a 10-wk running program without feedback. Therefore, previously reported running gait adaptations may have been a result of the running intervention and not necessarily the specific feedback used. Considering the lack of randomized control groups in previous gait retraining study designs, the current findings are perhaps a more robust appraisal of the effectiveness of RTF gait retraining on reducing impact loads during running.
It has been established that an indicator of whether motor learning has taken place is based on the retention beyond the training period (37). Crowell and Davis (7) showed greater signs of retention (Pre to 1-month posttests) of the gait retraining interventions (geared toward reduced impact loading) compared with the present findings. The results of the present study showed a clear return toward baseline levels in loading rates 1 month after the gait retraining. Although Crowell et al. (7) suggested that all of their participants had natural new gait patterns by the end of the sixth session, runners in the present study may require further gait retraining sessions or a refresher “session” to promote the retention of reduced impact loading during running. Moreover, the poorer retention levels observed in the present study may be related to differences in frequency of the feedback provided and run durations between studies. Crowell et al. (7) used a faded feedback approach (decreasing the feedback over the last three sessions) along with increasing the run duration gradually over the gait retraining intervention, whereas the current study provided feedback continuously during constant run durations (i.e., 6 × 20-min runs). Because the motor learning literature has shown evidence that gradually removing feedback encourages individuals to develop their own task-intrinsic feedback system from self-discovery rather than being reliant on augmented feedback (41), it could be that using a faded feedback approach along with gradually increasing run duration is a more beneficial gait retraining schedule for promoting better skill retention in runners. In support of this theory, Lintern et al. (25) suggested that individuals should try to focus on task-intrinsic feedback aspects rather than rely solely on the augmented feedback, so when the feedback is removed, they can perform better without it. Another explanation for the improved retention observed in the study of Crowell et al. (7) but not in the present study may be related to the differences in verbal instructions given to participants at the start of the gait retraining. Crowell et al. (7) instructed participants to “run softer” and make footfalls lighter, whereas the present investigator instructed participants to find a strategy to reduce PTA values.
On examining the mechanical strategies used by runners in response to the RTF gait retraining, it was apparent that runners adjusted their running style primarily by altering their ankle joint mechanics and foot strike pattern. This finding is in support of the hypothesis as mechanical adaptations were only observed in the RTF group and not in the control group. Results showed that runners in the RTF group had significantly changed from a dorsiflexion position at initial contact to a more plantarflexed position after the training (Fig. 2A). In addition, they also changed their foot strike pattern to a midfoot strike from a rearfoot strike. This significant adaptation in response to the gait retraining suggests that runners adopted an increased plantarflexed foot position at initial contact in a strategy to reduce impact loading during running. With no reported changes in foot to ground angles in the control group, Figure 2C illustrates that the RTF-grouped runners were adopting a less pronounced rearfoot strike pattern posttraining. In support, Giandolini et al. (21) reported significant decreases in impact loading rates when they promoted runners to adopt a midfoot strike pattern as compared with habitual rearfoot strike pattern. The possible reason for the reductions in impact loading with the observed ankle and foot posture adaptations may be related to greater range of motion (increased dorsiflexion during early stance) about the joint assisting with the absorption of impact loads through rotational energy (Fig. 2A). Interestingly, Derrick et al. (14) found that when the line of action of the force vector was increased from the joint axis of rotation, this condition can potentially increase the amount of impact energy that is absorbed at the joint. Derrick et al. (14) then go on to explain that this mechanism allows the muscles to act eccentrically, improving the system’s capacity for impact attenuation during running. Furthermore, researchers claim that the lower impact loads associated with forefoot striking are linked to the reduction in the effective mass of the lower limbs (13,25), as this decrease in effective mass through increased joint flexion (or in this case plantarflexion) after initial contact allows the body to be in a better position for absorbing impact forces during running.
Aside from the observed ankle mechanical adaptations in the RTF group, it was apparent that another strategy adopted by the RTF-grouped runners to reduce impact loading was the lowering of their HVV at initial contact. The importance of HVV at initial contact on impact loading was postulated by the modeling work of Gerritsen et al. (20), who predicted a strong association with reductions in impact loading rates and lower HVV at initial contact. Interestingly, although the Crowell et al. (7) study did not report kinematic strategies, it is plausible that instructing runners to “run softer with lighter footfalls” could have been associated with a strategy of lowering heel touchdown velocity to reduce impact loading. In support, De Wit et al. (11) showed that barefoot runners adopted a flatter foot placement and a reduced heel velocity on landing to reduce the localized force pressure measurements and pain under the heel. Therefore, the notion of a reduced heel velocity being associated with a softer landing is reasonable. Nevertheless, although this lower initial contact velocity strategy appeared to be used in the posttest, it was apparent that runners could not maintain this strategy 1 month after the RTF training as vertical loading rates returned toward baseline levels.
Although the present findings would seem to indicate a reduced injury potential in running after the RTF, researchers acknowledge that the reorganization of altering foot strike patterns may lead to a shift in the type of injury rather than a guaranteed reduction of injury potential (9,22). Coincidentally, responses from runners in the RTF group who altered their foot strike patterns did, however, comment on their calf muscles being tight and stiff during training and posttraining. Giandolini et al. (21) also observed signs of delayed onset of muscular pains in the plantarflexors when runners adopted a midfoot strike pattern from their habitual rearfoot strike pattern. Because forefoot striking has been found to generate higher ankle joint moments (40), it could be that this type of running demands stronger calf muscles due to the eccentric or isometric contractions needed to control ankle dorsiflexion during the loading response part of stance (9). Evidence of this greater ankle moment in the early part of stance is illustrated in Figure 2 (B). It is therefore recommended that any gait retraining in this regard should adhere to the appropriate overload, recovery, adaptation, and progression principles of training.
The present findings showed that running economy was not affected by the RTF training intervention. This finding rejects the hypothesis in which we expected running economy would worsen in the RTF group. Considering the reductions in impact loading severity because of mechanical adaptations, it was apparent that these modifications did not affect the runners’ running economy. In support, Perl et al. (35) reported no significant differences in running economy between different foot strike patterns in shod running. Similarly, Messier and Cirillo (29) also showed running economy was resistant to change with modifications in running mechanics after runners completed a 5-wk verbal and visual feedback training intervention. The possible reason for running economy remaining unaffected could be that the observed modifications in running mechanics because of the RTF training were not global enough to elicit an effect on runner’s metabolic costs, that is, running economy being insensitive the change. In support of this claim, a previous report found that with major changes in running mechanics, that is, arms behind back and increased vertical oscillations of the center of mass, running economy was significantly affected, but the magnitude of effect was surprisingly low considering these changes (17). The variability in running economy among the current group of runners is very similar to a previous report (31). However, the variability does increase in the 1-month posttest, and this could reduce the chance of finding significant group effects. Some individuals demonstrated improved running economy 1 month posttest, whereas others had detriments in running economy. This could be partly explained by the lack of strict control of individual training regimes and body weight changes during the final 1-month period.
The technological development of including auditory feedback within the present gait retraining system showed itself to be a promising development in the area of RTF gait retraining. Runners commented on the benefits of auditory feedback as it allowed them to concentrate on other aspects of the environment, that is, their limb movements. This was also reported in a previous gait training report (18) as they pointed out that vision is not occupied by staring at a screen with auditory feedback, but instead, it enables the runner to use other visual cues from the environment. Inevitably, these comments should further encourage researchers to consider using auditory feedback over visual feedback systems because of its superior capability of being transferred into an outdoor setting via a portable headphone system. This is because testing subjects in a controlled laboratory setting may not always be truly representative of how subjects behave in their outdoor running environment. Hence, one should always consider and question the ecological validity of RTF gait retraining studies conducted in a laboratory environment. Therefore, developing a fully integrated portable gait retraining system for runners will not only help progress this area of research but also hopefully provide a greater insight into the effectiveness of these types of systems for reducing or preventing the risk of tibial stress fracture injury in runners.
One limitation of this study is that only the immediate and short-term (1 month post) effects of RTF gait retraining on reducing impact loading were reported. Therefore, future longer duration prospective studies, possibly incorporating brief feedback refresher sessions, are warranted to establish whether RTF gait retraining is an effective long-term intervention for reducing the risk of impact-related injuries in runners. In addition, the current study is limited as it did not assess runners in a free-running, nonlaboratory environment to investigate the transfer of laboratory training to the runners’ more natural running environments.
CONCLUSIONS
The structured RTF training program provided a potential influential method to reduce impact loading during running without having detrimental effects on a runner’s economy. As hypothesized, the feedback was effective in lowering the magnitude of tibial shock at the Post and 1-month posttraining stages; however, loading rates had returned toward pretraining rates by 1-month posttraining. Contrary to our hypothesis, running economy did not change in either group over the training/posttraining stages. With some variables showing returns toward pretraining values during the 1-month posttraining period, it remains to be established whether significant adaptations that reduce running impact severity can be retained in the long-term. Finally, as hypothesized, this study shows significant kinematic changes in the RTF group with no kinematic changes in the control group. However, the significant changes only related to ankle and foot mechanics, indicating strong consistency in hip and knee kinematics while undertaking the feedback intervention.
The authors acknowledge the Department for Education and Learning Northern Ireland for their funding support of this research project. There are no conflicts of interest with the present study findings. The results of the present study do not constitute endorsement by the American College of Sports Medicine.
REFERENCES
1. Altman AR, Davis IS. A kinematic method for footstrike pattern detection in barefoot and shod runners.
Gait Posture. 2012; 35 (2): 298–300.
2. Beck BR. Tibial stress injuries—an aetiological review for the purposes of guiding management.
Sports Med. 1998; 26 (4): 265–79.
3. Begon M, Monnet T, Lacouture P. Effects of movement for estimating the hip joint centre.
Gait Posture. 2007; 25 (3): 353–9.
4. Butler RJ, Davis IS, Hamill J. Interaction of arch type and footwear on running mechanics.
Am J Sports Med. 2006; 34 (12): 1998–2005.
5. Clansey AC, Hanlon M, Wallace ES, Lake MJ. Effects of fatigue on running mechanics associated with tibial Stress fracture risk.
Med Sci Sports Exerc. 2012; 44 (10): 1917–23.
6. Cohen J. A power primer.
Psychol Bull. 1992; 112 (1): 155.
7. Crowell HP, Davis IS. Gait retraining to reduce lower extremity loading in runners.
Clin Biomech. 2011; 26 (1): 78–83.
8. Crowell HP, Milner CE, Hamill J, Davis IS. Reducing impact loading during running with the use of real-time visual feedback.
J Orthop Sports Phys Ther. 2010; 40 (4): 206–13.
9. Daoud AI, Geissler GJ, Wang F, Saretsky J, Daoud YA, Lieberman DE. Foot strike and injury rates in endurance runners: a retrospective study.
Med Sci Sports Exerc. 2012; 44 (7): 1325–34.
10. Davis I, Milner CE, Hamill J, Davis IS. Does increased loading during running lead to tibial stress fractures? A prospective study.
Med Sci Sports Exerc. 2004; 36 (5 suppl): S58.
11. De Wit B, De Clercq D, Aerts P. Biomechanical analysis of the stance phase during barefoot and shod running.
J Biomech. 2000; 33 (3): 269–78.
12. Dempster WT. Space requirements of the seated operator.
WADC Tech Rep. 1955; 55–159.
13. Derrick TR. The effects of knee contact angle on impact forces and accelerations.
Med Sci Sports Exerc. 2004; 36 (5): 832–7.
14. Derrick TR, Hamill J, Caldwell GE. Energy absorption of impacts during running at various stride lengths.
Med Sci Sports Exerc. 1998; 30 (1): 128–35.
15. Dixon SJ. Influence of a commercially available orthotic device on rearfoot eversion and vertical ground reaction force when running in military footwear.
Mil Med. 2007; 172 (4): 446–50.
16. Edwards BW, Taylor D, Rudolphi TJ, Gillette JC, Derrick TR. Effects of running speed on a probabilistic stress fracture model.
Clin Biomech. 2010; 25 (4): 372–7.
17. Egbuonu M, Cavanagh P, Miller T. Degradation of running economy through changes in running mechanics.
Med Sci Sports Exerc. 1990; 22 (2): S17.
18. Eriksson M, Halvorsen KA, Gullstrand L. Immediate effect of visual and auditory feedback to control the running mechanics of well-trained athletes.
J Sports Sci. 2011; 29 (3): 253–62.
19. Frost HM. A brief review for orthopedic surgeons: fatigue damage (microdamage) in bone (its determinants and clinical implications).
J Orthop Sci. 1998; 3 (5): 272–81.
20. Gerritsen KGM, van den Bogert AJ, Nigg BM. Direct dynamics simulation of the impact phase in heel-toe running.
J Biomech. 1995; 28 (6): 661–8.
21. Giandolini M, Arnal PJ, Millet GY, et al. Impact reduction during running: efficiency of simple acute interventions in recreational runners.
Eur J Appl Physiol. 2013; 113 (3): 599–609.
22. Hamill J, Gruber AH, Derrick TR. Lower extremity joint stiffness characteristics during running with different footfall patterns.
Eur J Sport Sci. 2012; 1 (7): 1–7.
23. Harrast MA, Colonno D. Stress fractures in runners.
The Runner. 2010; 29 (3): 399–416.
24. Hreljac A, Marshall RN, Hume PA. Evaluation of lower extremity overuse injury potential in runners.
Med Sci Sports Exerc. 2000; 32 (9): 1635–41.
25. Lieberman DE, Venkadesan M, Werbel WA, et al. Foot strike patterns and collision forces in habitually barefoot versus shod runners.
Nature. 2010; 463 (7280): 531–5.
26. Lintern G, Roscoe SN, Sivier JE. Display principles, control dynamics, and environmental factors in pilot training and transfer.
Hum Factors. 1990; 32 (3): 299–317.
27. Martin AD, McCulloch RG. Bone dynamics: stress, strain and fracture.
J Sports Sci. 1987; 5 (2): 155–63.
28. McBryde AM. Stress fractures in runners.
Clin Sports Med. 1985; 4 (4): 737–52.
29. Messier SP, Cirillo KJ. Effects of a verbal and visual feedback system on running technique, perceived exertion and running economy in female novice runners.
J Sports Sci. 1989; 7 (2): 113–26.
30. Milner CE, Ferber R, Pollard CD, Hamill J, Davis IS. Biomechanical factors associated with tibial stress fracture in female runners.
Med Sci Sports Exerc. 2006; 38 (2): 323–8.
31. Moore IS, Jones AM, Dixon SJ. Mechanisms for improved running economy in beginner runners.
Med Sci Sports Exerc. 2012; 44 (9): 1756–63.
32. Nigg BM. Impact forces in running.
Curr Opin Orthop. 1997; 8 (6): 43–7.
33. Noehren B, Scholz J, Davis I. The effect of real-time gait retraining on hip kinematics, pain and function in subjects with patellofemoral pain syndrome.
Br J Sports Med. 2011; 45 (9): 691–6.
34. O’Leary K, Vorpahl KA, Heiderscheit B. Effect of cushioned insoles on impact forces during running.
J Am Podiatr Med Assoc. 2008; 98 (1): 36–41.
35. Perl DP, Daoud AI, Lieberman DE. Effects of footwear and strike type on running economy.
Med Sci Sports Exerc. 2012; 44 (7): 1335–43.
36. Rauh MJ, Macera CA, Trone DW, Shaffer RA, Brodine SK. Epidemiology of stress fracture and lower-extremity overuse injury in female recruits.
Med Sci Sports Exerc. 2006; 38 (9): 1571–7.
37. Salmoni AW, Schmidt RA, Walter CB. Knowledge of results and motor learning: a review and critical reappraisal.
Psychol Bull. 1984; 95 (3): 355.
38. Schmidt RA, Lange C, Young DE. Optimizing summary knowledge of results for skill learning.
Hum Mov Sci. 1990; 9 (3–5): 325–48.
39. Taunton JE, Ryan MB, Clement DB, McKenzie DC, Lloyd-Smith DR, Zumbo BD. A retrospective case-control analysis of 2002 running injuries.
Br J Sports Med. 2002; 36 (2): 95.
40. Williams DS, McClay IS, Manal KT. Lower Extremity mechanics in runners with a converted forefoot strike pattern.
J Appl Biomech. 2000; 16 (2): 210.
41. Winstein CJ. Knowledge of results and motor learning—implications for physical therapy.
Phys Ther. 1991; 71 (2): 140–9.
42. Winstein CJ, Schmidt RA. Reduced frequency of knowledge of results enhances motor skill learning.
J Exp Psychol Learn Mem Cogn. 1990; 16 (4): 677.
43. Winter DA.
Biomechanics and Motor Control of Human Movement. John Wiley & Sons Inc; 2009. pp. 70–3.
44. Winter EM.
Sport and Exercise Physiology Testing Guidelines: The British Association of Sport and Exercise Sciences Guide. London: Routledge; 2006. pp. 147–55.
