Running is greatly known for its positive impact on general health and longevity, but it may also put individuals at risk of injuries. The incidence of running-related injuries per 1000 h of running ranged from 2.5 in long-distance track and field athletes to 33.0 among novice runners (Videbaek et al., 2015). Anatomical malalignment of the foot is one of the many factors associated with running-related injuries, and flat-footed runners are reported more likely to sustain soft-tissue and overuse injuries (Gijon-Nogueron and Fernandez-Villarejo, 2015). Low-arch or flat-foot is characterized by forefoot abduction, hindfoot valgus and a decrease in medial arch height (Huang et al., 1993). Flat-footed runners are susceptible to increased stress and pain experienced on the medial aspects of the lower extremity (Bandholm et al., 2008). In an attempt to reduce the risk of sustaining lower extremity injuries in runners with flat-feet, non-surgical interventions such as anti-pronation taping has been put in place to reduce foot pronation (Cheung et al., 2011).

The effects of anti-pronation taping are evident in the increase in navicular height and re-distribution of plantar pressure patterns in the medial and lateral midfoot regions (Newell et al., 2015; Nolan and Kennedy, 2009). While low-Dye taping is generally effective in reducing pronation of the foot in the literature, most studies either focused only on male participants or combined results of both sexes without reporting the outcomes from male and female participants separately (Franettovich et al., 2008; Newell et al., 2015). In a study on external ankle taping, no biomechanical differences during sprinting were found between low-Dye and sham taping in a group of young males (Moore et al., 2020). There are, however, biomechanical differences in locomotion between males and females (Chiu and Wang, 2007; Ferber et al., 2003; Hennig, 2001; Stefanyshyn et al., 2003). For example, females were found to have greater ankle motion, vertical ground reaction forces (GRF), and higher tibialis anterior muscle activation during walking compared to males (Chiu and Wang, 2007). Female runners also exhibited higher loading rate of the vertical GRF and reduced peak plantar pressure than their male counterparts (Hennig, 2001; Stefanyshyn et al., 2003). Given these known differences between sexes, the effect of taping on running biomechanics observed in males or in cases where males and females were treated as one group may not be applicable to female runners.

In the literature, there were only two studies investigating the effect of taping in females (Ator et al., 1991; Vicenzino et al., 1997). In both studies, the reported outcome was limited to a static measure of vertical navicular height which was found increased immediately after the application of tape (Ator et al., 1991; Vicenzino et al., 1997). Other dynamic variables such as GRF may offer greater insights on the loading characteristics during running to better indicate the risks of injuries. Recent study suggests that peak breaking force is associated with a significantly higher injury hazard ratio in female recreational runners (Napier et al., 2018). Furthermore, assessment of comfort level may also be useful to provide a more holistic evaluation of the taping intervention as shoe inserts that are comfortable have been shown to decrease injury frequency (Mündermann et al., 2001). In footwear research, lower perceived comfort is associated with reduction in the variability of running kinematics or a more repetitive running pattern (Mohr et al., 2017). This finding is line with the new paradigms of ‘preferred movement path’ and ‘comfort filter’ for injury prevention, suggesting that a runner intuitively selects a comfortable product that allows them to run in their preferred movement path (Nigg et al., 2015). In general, application of tapes in the foot and ankle region is often perceived as uncomfortable (Hyland et al., 2006). To better understand the effects of low-Dye taping on females, a comprehensive study on running biomechanics and comfort level is warranted.

The aim of this study was to examine the effects of low-Dye tape on GRF and comfort level in flat-footed female recreational runners. It was hypothesized that application of the low-Dye tape would be perceived as less comfortable and would decrease biomechanical loading when compared with sham taping.

This controlled laboratory study adopted as a within-subject, single-blind, randomized crossover design in which participants served as their own control. Low-Dye taping was used to reduce pronation of the foot and sham taping acted as a placebo (Newell et al., 2015). To blind participants from the experiment / placebo condition, they were informed that two different types of taping techniques would be compared. Participants were required to run at three different speeds on a treadmill under each taping condition where GRF data were collected.

This study was approved by the Nanyang Technological University Institutional Review Board (Protocol Number: IRB-2017-08-011). An a-priori power analysis for difference between two conditions (low-Dye versus sham, matched pairs) showed that at least 15 participants (α = 0.05, effect size = 0.80 large, power = 0.80) were required (Ng et al., 2020). In the present study, 15 female volunteers provided written informed consent to participate (Table 1). After filling in a background survey, they were screened for foot type using the following three methods:

      (i) Arch index (AI). Participants walked over the EMED^® M platform (Novel GmbH, Germany) using a 2-step approach. Dynamic arch index (AI) obtained through the Novel software to indicate foot posture (Murley et al., 2009). An AI of more than or equals to 0.26 was considered flat foot, reflecting a large midfoot area compared with the total foot area (Cavanagh and Rodgers, 1987). This method of classifying flat foot has been wide used in other studies (Ng et al., 2020; Tong and Kong, 2016).

      (ii) Navicular drop test. Navicular drop was measured as the change in height of the navicular tuberosity when the foot was moved from subtalar joint neutral position (non-weight bearing) to a resting calcaneal stance position (weight bearing). A navicular drop value of more than or equals to 10 mm was considered flat foot, reflecting substantial lowering of the foot arch under weight bearing condition. This technique is one of the most commonly used clinical measurement and has high intra-tester reliability and concurrent validity (Christensen et al., 2014).

      (iii) Resting calcaneal stance position (RCSP). Participants were asked to stand in a relaxed manner with equal weight on both legs and a protractor was used to measure the angle between the calcaneus bisection line and the vertical line drawn on the leg. A RCSP of more than or equals to 4˚ valgus was considered flat foot, as the standing foot posture is everted from the neutral position. This test has been shown as a reliable and acceptable technique (Sell et al., 1994).

If a foot fulfilled at least two out of the three criteria (AI ≥ 0.26, navicular drop ≥ 10 mm, RCSP ≥ 4 Ëš valgus) described above, it was considered flat. Only participants with both feet classified as flat-foot were recruited. The overall inclusion criteria were: (a) female, (b) aged between 18 and 35 years, (c) have shoe size within UK 5 and 7 (in order to fit the standardized footwear provided), (d) were recreational runners who had been running for the past 2 months, at least twice a week for more than 2 km per session, (e) bilaterally flat-footed. Participants were excluded if they (a) have previously gone through surgeries to the legs, (b) had serious injuries to their back or legs which required more than 7 days of rest in the past 6 months prior to participation in the study, or (c) had any discomfort or pain in their back or legs at the time the study was conducted.

Taping Conditions. Eligible participants underwent two different taping conditions with rigid 4-cm wide sports tapes (Strappal Rigid Tape, Vibraye, France), in a randomly assigned order. The two conditions are low-Dye taping and sham taping as a placebo treatment (Figure 1). For good consistency, application of the tape was done by the same researcher throughout this study. As for the low-Dye tape, two anchor strips were applied from the first metatarsophalangeal joint to the fifth metatarsophalangeal joint. In between the placement of the two anchor strips, about five to six stirrups (dependent on the participant’s foot size) were then applied starting from the lateral side of the anchor strip to the medial side of the anchor strip (Vicenzino et al., 1997). In order to oppose pronation of the foot, some force was applied medially. The sham taping method was adopted from a previous study (Kelly et al., 2010) in which two anchor strips were applied from the navicular tuberosity to the fifth metatarsal styloid process, posterior to the calcaneus. In between the anchor strips were about three to four stirrups (dependent on the participant’s foot size) which extended from the medial to lateral malleoli. To ensure the tape had no effect, much caution was taken such that no force was applied in the sham condition (Kelly et al., 2010). The tapes only fell off in one participant during the warm-up period. No adjustments of the tapes were needed during the actual running tests.

Upon the application of the tape, participants were allowed to walk or jog slowly around the laboratory to evaluate the comfortability of the tape. Participants were given as much time as they required. Comfort level in each taping condition was assessed using a 150-mm visual analogue scale (VAS), where the 0-mm end was “Extremely Uncomfortable” and the 150-mm end was “Extremely Comfortable” (Cheung and Ng, 2007). The participants were asked to mark their perceived level of comfort anywhere throughout the continuum by drawing a vertical line perpendicular to it, and the line was measured to the nearest millimeter. Assessing comfort using VAS has been demonstrated as a reliable method in studies on footwear and inserts (Mills et al., 2010).

Running Test. Participants were given sufficient time to warm-up and stretch. After which, they were asked to run on a split-belt force-plate instrumented treadmill sampled at 1000 Hz (Bertec, Inc., Columbus, OH) at three different speeds of 9, 10, and 11km/hr. This speed range was chosen to reflect the typical jogging speeds of recreational female runners and the speeds were comparable to those used in previous studies on females (Cheung and Ng, 2007; Ferber et al., 2011; Milgrom et al., 2003). The order of speeds was randomized. To eliminate the influence of socks and footwear, all participants were given a standardized pair of socks (Erke, Performance sock, China) and neutral shoes (Asics, Patriot 8, T669N-9001, Japan) for the running test. Each running trial lasted for three minutes, with a three-minute break in between. To allow sufficient time for the participant to reach a stable running state at each speed, GRF were recorded from 2:15 for 30 seconds in each bout of running (Kluitenberg et al., 2012).

Raw GRF data were processed using a fourth-order Butterworth digital low-pass filter using a cut-off frequency of 110 Hz determined from residual analysis. For each running trial, 60 foot steps were extracted for analysis using a customized MATLAB R2017b script (The MathWorks, Inc., Natick, MA). Only rearfoot strikes, determined based on the existence of both impact and active peaks, were included. Non-rearfoot strikes were excluded because of the absence of impact peak force and hence impact loading rate cannot be computed. In some participants, all 60 steps were rearfoot strikes. Most participants were consistent rearfoot strikers, with very few steps (less than 5) removed due to non-rearfoot striking patterns. First, the instances of touchdown and toe-off were identified from the vertical GRF data using a threshold of 20 N. Stance time was calculated as the duration from touchdown to toe-off.

Force data were then normalized to individual participants’ body weight (BW). Peak impact and active forces were identified from the vertical GRF. Loading rate was calculated from the duration of 20% to 80% between touchdown and the impact peak (Clansey et al., 2012). Next, the peak breaking, propulsive, medial, and lateral forces were determined from the anteroposterior and mediolateral GRF. To plot the mean ensemble curves, GRF data were also time-normalized to the stance phase (0 to 100%).

Statistical analyses were performed on IBM^TMSPSS^TM Statistics (Version 24.0, SPSS Inc., Chicago, IL, USA). The means, standard deviations, and 95% confidence intervals (CI) were calculated. For comfort level, normality of the VAS data was verified using the Shapiro-Wilk test. A paired t-test was performed to compare the VAS scores between the sham and low-Dye taping conditions. Statistical significance was set at p < 0.05.

For biomechanical data, there were 8 variables of interest including stance time, loading rate, and 6 peak forces (impact peak, active peak, breaking peak, propulsive peak, medial peak and lateral peak). A 2 (type of taping) X 3 (speed) Analysis of Variance (ANOVA) with repeated measure was performed for each of these variables to compare between the taping conditions and across the three speeds. In cases where Mauchley’s test showed the violation of the sphericity assumption, Greenhouse-Geisser’s epsilon adjustment was used. Bonferroni adjusted post-hoc comparisons were applied where necessary. Statistical significance was set at p < 0.05. Observed power and effect size (partial eta squared, η_p²) were also calculated.

In the comfort assessment, the VAS score was significantly lower in low-Dye taping than sham tape (low-Dye, 63.8 (24.3) mm, sham 122.0 (16.0) mm, mean difference [95% CI], -58.2 [68.2, 48.2] mm, p < 0.001).

For all biomechanical variables, there was a significant main effect of speed but no interaction effect or difference between the two taping conditions (Table 2). As running speed increased, there was a decrease in stance time (p < 0.001) and increase in loading rate (p = 0.009), impact peak (p = 0.004), active peak (p < 0.001), breaking peak (p < 0.001), propulsive peak (p < 0.001), medial peak (p < 0.001), and lateral peak (p < 0.001). Graphical comparison of the force-time histories over the stance phase of running revealed similar patterns between low-Dye and sham taping conditions (Figure 2).

This study investigated the effects of low-Dye taping on running biomechanics and comfort level in female recreational runners. The two findings were: (i) participants were less comfortable with the low-Dye taping technique when compared with sham taping, and (ii) application of low-Dye tape did not alter GRF variables during running.

As hypothesized, participants felt less comfortable when low-Dye taping was applied when compared with the sham condition. The discomfort is likely due to the restricted joint movements resulted from the tension of the tapes. The present findings are consistent with those reported by Hyland et al. (2006) that in general current taping methods are often uncomfortable. One study comparing different types of tapes found that rigid taping was ranked the least comfortable compared with barefoot and elastic taping (Smith et al., 2015). In the present study, rigid tape was used in both low-Dye and sham conditions, with variation in the application of force. The lower comfort rating in low-Dye taping may also be due to the greater surface area of the foot and force that the tape was applied when compared with the sham technique (Figure 1). It should be acknowledged that previous taping experience might influence the comfort level of taping. Through verbal confirmation before the beginning of running test, all the participants declared that they had no experience of low-Dye taping or other taping techniques. With reference to the ‘preferred movement path’ and ‘comfort filter’ paradigms for running injury prevention (Nigg et al., 2015), runners may benefit from choosing a comfortable taping method which allows them to remain their preferred running patterns. In the attempt to reduce foot pronation via taping, further development in taping material and technique may help improve the subjective comfort experienced by runners alongside anatomical and biomechanical considerations. As the optimal amount of foot pronation is yet to be established (Nigg et al., 2019), comfort perception should be an important consideration for athletes when selecting treatments to control rearfoot motion.

To our best knowledge, this study is the first to identify that applying low-Dye taping does not change three dimensional GRF variables comparing with sham taping at a range of speeds in flat-footed female runners. The hypothesis that low-Dye taping would decrease loading during running is not supported since no difference in stance time, loading rate, or peak forces in any direction were found between sham taping and low-Dye taping. These results suggest that the external load, measured by GRF, were similar regardless of the taping technique. This study also showed the increase in running speed resulted in short stance time and greater loading rate and peak forces, which are largely consistent with previous studies (De Wit et al., 2000; Grabowski and Kram, 2008).

The lack of biomechanical differences between low-Dye and sham taping is in agreement with a recent study by Moore et al. (2020) which showed that external ankle taping did not alter GRF or lower extremity kinematics and kinetics during straight sprints in young males. Focusing on females, previous studies demonstrated that low-Dye taping was effective in increasing navicular height during quiet standing immediately after the application of tape (Ator et al., 1991; Vicenzino et al., 1997). Newell et al. (2015) investigated navicular height during running and showed no differences from baseline to post-tape for navicular height in the low-Dye and no-tape conditions. When the navicular height was significantly higher immediately after application of the navicular-sling condition, it was reduced after 5 minutes of treadmill running (Newell et al., 2015). Interestingly, another study also found that the corrective effect of raising the navicular height by taping greatly diminished after 10 minutes of jogging (Ator et al., 1991). Although these earlier studies only measured static navicular height and did not evaluate the dynamic loading response (Ator et al., 1991; Newell et al., 2015; Vicenzino et al., 1997), the reported short-lasting effect of taping could help explain the lack of biomechanical differences observed in the present study. When interpreting the initial correction of foot alignment via anti-pronation taping, caution should be taken not to assume that taping will be effective in restricting joint motion or reduce loading during exercise.

In studies where the foot-ground interface was modified via different footwear properties, runners were found to adjust their kinematics to maintain constant external loads such as peak forces and loading rate (Kong et al., 2009). It is possible that the participants in the present study have adjusted their running technique in response to taping such that similar GRF profiles are maintained. For example, greater knee flexion can be a strategy to regulate impact force (Bobbert et al., 1991) in order to compensate for the potentially increased impact magnitude caused by reduced foot pronation (Gerritsen et al., 1995). On the other hand, it has been reported that sprinting biomechanics including GRF, kinematics, and kinetics did not differ between ankle taping and control groups (Moore et al., 2020). Since kinematic data were not obtained in the present study, these speculations cannot be confirmed. Given that stance time did not differ between the two taping conditions, it is unlikely that substantial kinematic adaption would have occurred. Moving forward, future studies can include joint kinematic (e.g. angles) and kinetic (e.g. moments) measurements to allow a more comprehensive biomechanical analysis. Researchers can also include a ‘no tape’ condition in additional to ‘sham taping’ (Hyland et al., 2006) to provide a clear view of how taping can influence running biomechanics. For clinical application, the current study confirmed that application of low-Dye taping in flat-footed female runners would not lead to adverse kinetic effects such as increased peak force and loading rates.

As the first attempt to examine the effect of taping on running biomechanics in females, typical GRF variables at a range of speeds were measured. While this is considered one step forward from static foot alignment during standing (Ator et al., 1991; Vicenzino et al., 1997), only kinetic data were collected from the instrumented treadmill. For a more in-depth biomechanical analysis, future studies can also include kinematics and joint kinetics to clarify whether runners have adapted their running techniques in response to taping. The second limitation was that only one type of tape and the low-Dye technique were used. For maintaining the navicular height, more rigid tape such as Leukotape may be more effective, however, caution should be taken since the comfort level might be compromised. Additional research is warranted to further investigate the possible influences of different taping materials and techniques. The third limitation was that the a-priori power analysis was performed based on a hypothetical large effect size (Ng et al., 2020) since there were no available studies on taping effects or running GRF in the literature. We may have over-estimated the effect size, resulting in recruiting too few participants and hence under-powered statistical analysis. Finally, this study focused on asymptomatic flat-footed female runners with no pre-existing condition or pain. Future studies can establish the effects of taping for treating running related conditions such as plantar fasciitis.

Compared with sham taping, the application of low-Dye taping led to a decrease in comfort level among flat-footed female runners. For running biomechanics, low-Dye tape did not alter stance time, loading rate, or peak forces across a range of running speeds from 9 to 11 km/h. Clinicians can be assured that application of low-Dye taping in flat-footed female runners would not lead to adverse kinetic effects such as increased peak force and loading rates.