Common running injuries, such as patellofemoral pain, plantar fasciitis, and tibial stress fractures, have been associated with high impact forces during running (9,16,19,20,25). In particular, a high vertical impact peak in the vertical ground reaction force curve and a high rate of rise to this vertical impact peak (vertical loading rate) have been associated with these injuries (16,19,20,25). The vertical impact force transient is typically seen in runners who strike the ground first with their heel and signifies the end of the passive phase of loading (1,6). Thus, it appears that mitigating the forces associated with the passive impact transient should be the focus for reducing the risk of many running-related injuries.
Various interventions have been suggested to reduce impact forces associated with running. These include the use of cushioned shoes and insoles (4,5), as well as conscious alterations in running technique, also known as gait retraining (7,8,14,17,22,23). It has been shown that adopting a forefoot or midfoot strike pattern reduces or eliminates the vertical impact peak and reduces the vertical loading rate (7,15,23). Reduction of impact loading may also be accomplished by increasing step rate and reducing step length (14). A reduction in stride length is often accomplished as a strategy of increased knee flexion at foot strike and reduced vertical oscillation of the runner’s center of mass (14).
Recently, barefoot running has been suggested as a means to reduce impact forces (1). Habitual barefoot runners tend to strike the ground with a less dorsiflexed foot, often using a mid- or a forefoot strike pattern (1,10,12,15). It is reasoned that barefoot runners adopt a less dorsiflexed strike pattern because heelstriking while barefoot is painful (1,10,15). In fact, heelstriking while barefoot running results in a very high vertical impact peak and resultant vertical loading rate due to the collision forces between the dense calcaneus and the ground (15). For many runners, barefoot running may be impractical because of unsafe running surfaces and potential performance limitations (13). Thus, many runners opt for an intermediate option, a minimalist shoe. Minimalist shoe running may have an even lower metabolic cost compared with standard shoe running (18). Nevertheless, there is currently a wide variation of shoes that are advertised as “minimal.” All of these shoes are commercially marketed as an alternative to barefoot running.
Only two studies, to date, have examined the effects of minimalist footwear on kinematic and kinetic parameters of running. Squadrone and Gallozzi (21) examined habitual barefoot runners during three conditions: barefoot, minimalist, and standard shoe running. The minimalist shoe used was a five-toed shoe with a flexible-upper, no-arch support, a zero-drop, no-midsole and a 3-mm outersole and the standard shoe was described as a neutral shoe. Using a two-dimensional kinematic assessment and an instrumented treadmill, barefoot and minimalist shoe running both resulted in a reduction in sagittal plane ankle and knee angle just before footstrike compared with standard shoe running. In addition, they reported a reduction in strike index, stride length, and an increase in step rate (21). Finally, the vertical impact peak was found to be reduced in both barefoot and minimal footwear conditions. However, these authors did not assess vertical loading rate in this study. Further, habitual barefoot runners were the participants. Therefore, their results may be because of long-term changes in running mechanics due to barefoot running rather than any immediate effect of footwear. More recently, Bonacci et al. (3) examined the running mechanics of highly trained runners while running in standard (each runner’s regular training shoe), lightweight (Nike LunaRacer2), minimalist (Nike Free 3.0), and barefoot conditions. The lightweight and minimalist shoes were similar in the amount of cushioning they had. As a result, these two shoes resulted in very similar mechanics to each other. Runners did exhibit a reduced stride length and increased cadence in these two shoes compared to the standard shoe, but most other variables were similar to standard shoe condition. Neither of these more minimal shoes resulted in mechanics similar to barefoot running. This study underscores the idea that cushioned shoes, even when there is less cushioning than standard shoes, still encourage a heelstrike pattern. The only condition where a plantarflexed position at footstrike was noted was the barefoot one. Bonacci et al. did not report on impact-related forces between the shoe conditions. Understanding the influence of footwear on vertical impact peaks and load rates is critical because these forces have been related to running injuries.
Therefore, the purpose of this investigation was to compare running mechanics between standard neutral cushioned shoes and minimalist shoes in novice minimalist shoe runners. We also sought to determine whether runners are able to accommodate to the minimal cushioning in minimalist shoes within a single running session. We hypothesized that runners would increase step rate while decreasing stance time in the minimalist shoes, resulting in a reduction in impact forces when compared with a standard cushioned shoe. An associated decrease in ankle dorsiflexion and increase in knee flexion at footstrike would also be noted in the minimalist shoe condition. We also hypothesized that runners would increase step rate, decrease impact forces, and alter their running kinematics to an even greater extent over the course of the run as runners became accommodated to the minimalist shoe.
The data collection protocol and informed consent document were approved by the University of Delaware Human Subjects Research Board. To participate, both written and verbal informed consent forms were obtained from each volunteer. An a priori power analysis was conducted using data from a pilot work for this study. Using the variable with the highest SD, ankle dorsiflexion at footstrike, it was revealed that 12 subjects (effect size = 0.82, α = 0.05, β = 0.20) were required to adequately power this study. To be conservative, we recruited 14 male runners (31.9 ± 10.5 km·wk−1, 24.8 ± 3.2 yr) for this study. To qualify, each runner was required to be a habitual shod heelstriker, running at least 10 miles per week, between 18 and 35 yr of age, free of injury for the past 6 months. Habitual strike pattern was self-reported. Each runner was required to be comfortable with treadmill running, defined as a score of at least “8” on a visual analog scale (“0” and “10” corresponding to completely uncomfortable and completely comfortable, respectively). They also had to be comfortable running at 3.3 m·s−1. Importantly, each runner was required to be a “novel” minimalist shoe wearer, defined as never having previously run in minimalist shoes. Shoe history was self-reported. We operationally defined a “minimalist shoe” as a racing flat or advertised minimalist shoe. We chose to examine novel minimalist shoe runners to capture the true accommodation that may occur during the initial exposure to minimalist shoe running.
Shoe order was counterbalanced among subjects to avoid a fatigue and learned effect. A Nike Pegasus (Nike, Beaverton, OR) served as the standard cushioned shoe and the Nike Free 3.0 served as the minimalist shoe. The Nike Pegasus and Free shoes have heel insole heights of 36.3 and 17.6 mm, respectively, as per caliper measurement of size 47 EUR. An MTS QTest 10 Elite load frame (Cary, NC) with a 10-kN load cell was used to measure the stiffness of the heel insole. One shoe from each group was tested once. A compressive preload of 100 N was applied, then the load was increased to 1000 N with a load rate of 400 N·s−1. Stiffness values of 64.5 and 88.2 N·mm−1 were determined for the standard and minimalist shoes, respectively. Thus, the standard shoe provided 31% greater cushioning than the minimalist shoe.
Thirty-five retroreflective markers were attached to the dominant lower extremity to analyze running kinematics (VICON, Oxford, UK). Limb dominance was operationally defined as the leg used to kick a soccer ball. Marker bases were firmly attached to bony landmarks to establish the coordinate systems of the pelvis, thigh, shank, and foot. Marker placement on shoes has previously been shown to overestimate calcaneal movement during running. Therefore, shoe windows were used to facilitate marker placement directly on the calcaneus (Fig. 1).
Efforts were made to minimize any potential offsets that could be introduced to the subject models as a result of the different footwear conditions and variability in marker placement. Separate standing calibration trials were taken for each shoe condition, using identical anatomical marker placements. This was accomplished by using markers that could easily be separated from their bases, thus leaving the bases in place during the running trials. Once the first run condition was completed, the markers were reattached to their previously mounted bases for the subsequent standing calibration trial. Finally, all tracking markers, including markers mounted directly on the calcaneus, remained in place and unchanged between shoe running conditions.
For the collection of running trials, subjects ran on an instrumented treadmill (AMTI, Watertown, MA) at 3.35 m·s−1 (an 8-min-per-mile pace). To fully capture any accommodation to shoe conditions, no warm-up period was allowed. This was accomplished by accelerating the treadmill from a full stop to the test speed using an acceleration rate of 0.2 m·s−2 as the subject kept pace. This acceleration rate was modest and was deemed to be a comfortable and attainable acceleration rate by each participant. Data of five consecutive strides were analyzed in the first minute of running as soon as the test speed was reached and after 10 min of running. Kinematic and kinetic data were collected at 200 and 1000 Hz, respectively. Stance was determined using a threshold of 30 N of the vertical ground reaction force. Once data were collected after the 10th minute, the treadmill was stopped and the shoe condition was changed. Care was exercised to maintain the calcaneal markers in their original positions as shoes were fully unlaced and removed followed by replacing them with the second shoe condition. Anatomical markers were reattached to their previously mounted bases. The second testing session was then commenced, and data were recorded in exactly the same manner as described for the first shoe condition.
Data were then processed using Visual 3D (CMotion, Bethesda, MD). To eliminate any offset introduced into the ankle kinematic data due to the differences in heel height among the two shoe conditions, a virtual foot was created. This was done by subtracting the vertical height of the foot and malleoli anatomical markers in the laboratory coordinate system to construct the virtual segment. As such, the virtual foot had an inclination (dorsiflexion) angle of 0° during the standing calibration, when referenced to the laboratory coordinate system. Subsequent foot kinematic calculations were based on the virtual foot referenced to the laboratory coordinate system, and ankle kinematic calculations were based on the virtual foot referenced to the shank segment. Using this virtual foot, a positive angle of the foot inclination angle at footstrike would correspond to a dorsiflexed foot segment at foot strike. An 8- and 40-Hz, low-pass, fourth-order, zero-lag Butterworth filters were used to filter the kinematic and kinetic data, respectively. Customized software (LabVIEW 8.0, National Instruments, Austin, TX) was used to calculate the following discreet variables: vertical impact peak, average vertical loading rate, dorsiflexion at footstrike, foot inclination at footstrike, knee angle at footstrike, stance length, and step rate. The method for calculating the average vertical loading rate was done over the middle 60% of the vertical ground reaction force curve from foot strike to the vertical impact peak (Fig. 2) (16).
A series of two-way repeated-measures ANOVA (shoe × time) was used to analyze the data (α ≤ 0.05). When significant differences were found for main effects or the interaction, post hoc comparisons using paired t-tests (α < 0.05) were conducted between levels. Mauchly test of sphericity was used to determine whether the data met the assumptions of the ANOVA.
There were no significant shoe–time interactions, and therefore, only main effects were assessed. In terms of the main effect of shoe type, no differences were found for step length, step rate, and foot inclination angle at foot strike (P = 0.967, F = 0.002, P = 0.230, F = 1.586, and P = 0.332, F = 1.014, respectively) between shoes. However, runners struck the ground with a more dorsiflexed foot (P = 0.025, F = 6.379), in more knee flexion (P = 0.001, F = 13.000), and with higher vertical impact peak (P = 0.017, F = 14.902) and higher average vertical loading rate (P < 0.000, F = 23.400) (Table 1) while running in the minimalist shoe. These results did not agree with our hypotheses that runners would land in less foot inclination and with a less dorsiflexed ankle, experience lower impact forces, and reduce temporospatial measures in the minimalist shoe when compared with the standard running shoe.
There were no main effects of time for step length or step rate either (P = 0.088, F = 3.393 and P = 0.616, F = 0.265, respectively). Significant main effects for time for all kinematic and kinetic variables were found. Runners in both shoe conditions demonstrated reduced foot inclination at footstrike (P = 0.048, F = 4.763), reduced dorsiflexion (P = 0.035, F = 5.543), and increased knee flexion at foot strike (P = 0.002, F = 14.112), yet higher vertical impact peak (P = 0.002, F = 14.902) and average vertical loading rate (P < 0.000, F = 91.884) after 10 min of running.
We sought to determine whether running in a cushioned minimalist shoe results in reduced variables associated with impacts when compared with running in a standard shoe. In addition, we sought to determine whether accommodation to running in the minimalist shoe occurs over the course of a 10-min run. In contrast to our hypotheses, we found that running in the minimalist shoe failed to result in changes in temporospatial parameters, increased average vertical loading rates, and vertical impact peaks when compared with running in a standard running shoe. As hypothesized, running in the minimalist shoe resulted in a slightly more flexed knee at footstrike. The runners accommodated to the minimalist shoes in the same manner as to the standard shoes.
The minimalist shoe resulted in a more dorsiflexed ankle and more knee flexion at footstrike compared with standard shoe running. This increased knee flexion may have been a strategy to reduce the impact forces associated with a more dorsiflexed footstrike (15). However, this increase in knee flexion was small and apparently insufficient to alter the temporospatial measures or reduce impact forces during minimalist shoe running.
It is not clear why dorsiflexion would be increased in the minimal shoe, yet foot inclination was unchanged. Foot inclination was defined as the virtual foot referenced to the running surface, whereas ankle dorsiflexion was the foot referenced to the shank segment. Therefore, foot inclination is likely a better indication of foot strike pattern than ankle dorsiflexion angle. Previous work suggests that foot inclination at foot strike is strongly correlated with foot strike patterns (2). Therefore, we feel that the large (>10°) positive foot inclination angle at foot strike in both shoes indicates a defined heelstrike in the present study. As a large positive foot inclination angle at foot strike was coupled with a stiffer, less compliant minimalist shoe, there was less cushioning between the dense calcaneus and the running surface compared with the standard shoe. Thus, it is not surprising then that the minimalist shoe was associated with greater impact loading. Landing on the heel reduces the ability of the ankle to assist in attenuating the loads of impact. High-impact loading has been associated with a number of running-related injuries (9,16,19,20,25). In fact, the average vertical loading rates that we found during minimalist shoe running exceeded the values reported by Milner et al. (16) in runners with a history of tibial stress fractures (78.97 ± 24.96 body weight per second). This is amplified by the fact that the running speed was higher in the study of Milner et al. than in the present study (3.7 vs 3.3 m·s−1) because higher ground reaction forces would be expected at the faster speed. Further justification of this comparison is that there are no differences in impact forces between overground and treadmill running (24).
The greater ankle dorsiflexion seen in the minimal shoe was in contrast to the findings of both Squadrone and Gallozzi and Bonacci et al. Squadrone and Gallozzi (21) found that habitually barefoot runners in the Vibram five-finger shoe landed in plantarflexion compared to dorsiflexion in those in standard running shoes. The plantarflexion found when using the Vibram five-finger shoe was also similar to that found during barefoot condition. However, these runners were habitual barefoot runners and the minimal shoe had no cushioning at all. Bonacci et al. (3) found no difference between the Nike Free 3.0 and the runner’s standard shoes, which agrees with our findings. The Bonacci study used a similar shoe to the present study, but runners were given 10 d to accommodate to the shoes, which may explain some of the difference. The virtual foot may also at least partially be the source of this discrepancy in dorsiflexion angle between the present work and those of Squadrone and Gallozzi and Bonacci et al. While we accounted for foot angle differences between shoe types, we did not account for differences in shank orientation that may have resulted from the greater heel elevation in the standard shoe.
Our results also suggest that a 10-min accommodation period does not result in more favorable loading mechanics during running in the minimalist shoe. In fact, any changes in mechanics between minute 1 and minute 10 were nearly identical between the two shoe types. Rather than decreasing their impact loading as the runners became accommodated, an increase was found between the two time points in both shoes. While impacts have been noted to increase with fatigue (11), 10 min of running should not have led to fatigue in this group of runners. This may be explained by the fact that the Pegasus may have been an equally novel shoe for the runners as the minimalist shoe. An alternative explanation is that minimizing impact forces is not a criterion of the neuromuscular system during the initial minutes of running.
When taking the studies of Squadrone and Gallozzi and Bonacci et al. and the current study together, the following points can be made. To see a change in footstrike pattern that simulates barefoot running, the shoes may need to be as minimal as possible. The shoe in the study of Squadrone and Gallozzi was simply a thin rubber sole with a flexible upper. It was the only shoe condition of all three studies that resulted in similar mechanics as a barefoot condition. The study of Squadrone and Gallozzi used habitual barefoot runners who may also do some of their running in minimal shoes (21). Although the standard shoe condition was the one that was most unfamiliar, runners in the study of Squadrone and Gallozzi immediately increased their dorsiflexion at footstrike and increased their vertical impact peak. In both the study of Bonacci et al. and the present study, the minimal shoe used was one that still had significant cushioning and allowed for a comfortable heelstrike (3). In both cases, runners remained as rearfoot strikers in the minimal shoe condition, landing on their heel with greater impact.
Many runners use these transitional shoes as a way to adopt a barefoot-mimicking running technique, involving minimal plantarflexion and a mild forefoot strike pattern (15). However, these results suggest that runners should be careful when transitioning to cushioned minimal shoes. While there may be a belief that one will automatically adopt a forefoot or midfoot strike pattern by wearing minimal shoes, both the study of Bonacci et al. and the present study suggest otherwise. When using one of the vast array of minimal shoes with cushioning, additional gait retraining may be needed to train one to land with a midfoot or mild forefoot strike pattern. On the basis of previous studies, landing patterns can be altered (7,8,14) and with proper feedback and reinforcement, maintained up to 3 months beyond the training period (7,8). Whether this alteration is maintained in the long term when using cushioned minimal shoes is yet to be determined.
These results suggest that the minimally cushioned footwear used in the study did not induce a transition to a nonrearfoot strike pattern. As runners actually increased their impact loading, these shoes may increase the risk of injury, especially during this early phase of accommodation.
The authors thank the following funding sources: Drayer Physical Therapy Institute, Department of Defense (W911NF-05-1-0097), and National Institutes of Health (1 S10 RR022396).
There are no conflicts of interest among any of the authors of this article.
The authors also thank the subjects for their participation in this study.
Lindsay Buchenic assisted with processing of the running data. William Zaylor performed the materials testing on shoe stiffness qualities.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Altman AR, Davis IS. Barefoot running: biomechanics and implications for running injuries. Curr Sports Med Rep
. 2012; 11 (5): 244–50.
2. Altman AR, Davis IS. A kinematic method for footstrike pattern detection in barefoot and shod runners. Gait Posture
. 2012; 35 (2): 298–300.
3. Bonacci J, Saunders PU, Hicks A, Rantalainen T, Vicenzino BG, Spratford W. Running in a minimalist and lightweight shoe is not the same as running barefoot: a biomechanical study. Br J Sports Med
. 2013; 47 (6): 387–92.
4. Butler RJ, Davis IM, Laughton CM, Hughes M. Dual-function foot orthosis: effect on shock and control of rearfoot motion. Foot Ankle Int
. 2003; 24 (5): 410–4.
5. Butler RJ, Davis IS, Hamill J. Interaction of arch type and footwear on running mechanics. Am J Sports Med
. 2006; 34 (12): 1998–2005.
6. Cavanagh PR, Lafortune MA. Ground reaction forces in distance running. J Biomech
. 1980; 13 (5): 397–406.
7. Cheung RT, Davis IS. Landing pattern modification to improve patellofemoral pain in runners: a case series. J Orthop Sports Phys Ther
. 2011; 41 (12): 914–9.
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. Davis IS, Bowser B, Mullineau D. Do impacts cause running injuries? A prospective investigation. In: Proceedings of the American Society of Biomechanics Annual Meeting
. Providence, RI: August 21, 2010.
10. 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.
11. Derrick TR, Dereu D, McLean SP. Impacts and kinematic adjustments during an exhaustive run. Med Sci Sports Exerc
. 2002; 34 (6): 998–1002.
12. Divert C, Mornieux G, Baur H, Mayer F, Belli A. Mechanical comparison of barefoot and shod running. Int J Sports Med
. 2005; 26 (7): 593–8.
13. Franz JR, Wierzbinski CM, Kram R. Metabolic cost of running barefoot versus shod: is lighter better? Med Sci Sports Exerc
. 2012; 44 (8): 1519–25.
14. Heiderscheit BC, Chumanov ES, Michalski MP, Wille CM, Ryan MB. Effects of step rate manipulation on joint mechanics during running. Med Sci Sports Exerc
. 2010; 43 (2): 296–302.
15. 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.
16. 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.
17. 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.
18. Perl DP, Daoud AI, Lieberman DE. Effects of footwear and strike type on running economy. Med Sci Sports Exerc
. 2012; 44 (7): 1335–43.
19. Pohl MB, Hamill J, Davis IS. Biomechanical and anatomic factors associated with a history of plantar fasciitis in female runners. Clin J Sport Med
. 2009; 19 (5): 372–6.
20. Pohl MB, Mullineaux DR, Milner CE, Hamill J, Davis IS. Biomechanical predictors of retrospective tibial stress fractures in runners. J Biomech
. 2008; 41 (6): 1160–5.
21. Squadrone R, Gallozzi C. Biomechanical and physiological comparison of barefoot and two shod conditions in experienced barefoot runners. J Sports Med Phys Fitness
. 2009; 49 (1): 6–13.
22. Williams DS 3rd, Green DH, Wurzinger B. Changes in lower extremity movement and power absorption during forefoot striking and barefoot running. Int J Sports Phys Ther
. 2012; 7 (5): 525–32.
23. Williams DS, McClay IS, Manal KT. Lower extremity mechanics in runners with a converted forefoot strike pattern. J Appl Biomech
. 2000; 16 (2): 210–8.
24. Willy RW, Davis IS. Instrumented comparison of overground and treadmill running in healthy individuals. In: Proceedings of the 2008 Annual Meeting of the American College of Sports Medicine
. Indianapolis, IN: May 28, 2008.
25. Zadpoor AA, Nikooyan AA. The relationship between lower-extremity stress fractures and the ground reaction force: a systematic review. Clin Biomech (Bristol, Avon)
. 2011; 26 (1): 23–8.