Secondary Logo

Journal Logo

Original Research

Rocker-Bottom, Profile-Type Shoes Do Not Increase Lower Extremity Muscle Activity or Energy Cost of Walking

Santo, Antonio S.; Roper, Jenevieve L.; Dufek, Janet S.; Mercer, John A.

Author Information
Journal of Strength and Conditioning Research: September 2012 - Volume 26 - Issue 9 - p 2426-2431
doi: 10.1519/JSC.0b013e31823f8b71
  • Free



Rocker-bottom, profile-type shoes have been studied for efficacy in ameliorating symptoms associated with ailments of the lower limb (10,16). In particular, it has become increasingly standard practice to prescribe therapeutic shoes, as a prophylaxis and treatment, to patients suffering from peripheral neuropathy to relieve forefoot plantar pressure, which is associated with skin degradation in the insensate foot (4,10,19). Interestingly, one would expect that the use of rocker-bottom profile shoes would have an effect on the kinematics and kinetics of gait, posture, and ambulation owing to the instability they create; however, the few studies that have investigated these measures show little clinical significance (1,2,12–14).

Recently, a category of shoe has been introduced to the retail market that combines the rocker-bottom concept with a compliant midsole (RB). Based upon manufacturer claims, the intent of this shoe design is to activate muscles differently or to activate additional muscles than what are typically used when walking in flat-bottomed shoes. However, there is limited research on biomechanical or physiological parameters while walking in these types of shoes, and some of the research is conflicting. For example, Romkes et al. (17) observed differences in electromyography (EMG) while unstable, rocker-profile shoes were worn (i.e., Masai Barefoot Technology [MBT]) compared with traditional shoes in larger extrinsic foot muscles crossing the ankle joint complex (i.e., gastrocnemius [GA] and tibialis anterior) and in smaller extrinsic foot muscles (i.e., peroneus longus and flexor digitorum longus). In contrast, other researchers have reported no change in EMG while the subjects walked in MBT shoes, whereas Harris et al. reported that EMG patterns differ based on the profile of the RB shoe (3,8,15,18). Additionally, it was reported the metabolic cost of walking at 1.25 m·s−1 was greater while wearing MBT shoes compared with while wearing flat-bottomed shoes (5). However, the weight of the MBT shoes (1,010–1,350 g) was more than that of the flat-bottom shoes (250–1,100 g), and it is not clear if the change in the metabolic cost was because of the shoe design or because of wearing heavier shoes (5). In total, it is not clear if the rocker-bottom profile of an RB profile shoe challenges the extrinsic foot muscles such that the pattern of motor unit recruitment is altered and the energy efficiency of ambulating is reduced in an attempt to reach a homeostatic walking strategy.

Therefore, the purpose of this study was to determine if wearing rocker profile–type shoes (i.e., Skecher Shape-Ups) would alter muscle activity of primary lower extremity muscles or influence the metabolic cost of walking as compared with walking in flat-bottom shoes (W). Furthermore, the purpose was extended to determine if the weight difference between RB and W alone influenced muscle activity and metabolic cost. It was hypothesized that muscle activity and metabolic cost would be greater during walking in RB vs. W but not different compared with when walking in flat-bottom shoes with mass added (WM) to match the weight of the RB.


Experimental Approach to The Problem

The hypotheses were tested by using a within-subject experimental design. All the subjects completed 3 shod conditions that were as follows: flat-bottomed shoe (W), flat-bottomed shoe weight-matched to RB (WM), and rocker profile with compliant midsole (RB). The measurements taken during each walking condition were muscle activity and V[Combining Dot Above]O2. These data were recorded concurrently during a 10-minute walk in each footwear condition.


Twenty-eight (17 women and 11 men, age: 22.8 ± 6.6 years; height: 170 ± 6.7 cm; mass: 72 ± 20 kg; percent body fat: 23.0 ± 11.7; fat mass: 18.2 ± 17.3 kg; lean body mass: 54.8 ± 10.6 kg), apparently healthy participants completed all the conditions. The subjects were considered apparently healthy if they were free of known neurological, metabolic, cardiovascular, or pulmonary diseases and had no lower extremity injuries that prevented them from walking normally on a treadmill. This study was approved by the institutional review board of the host university, and all the subjects gave their written informed consent. After informed consent was obtained, height, weight, and percent body fat as determined by air displacement plethysmography (BodPod, Cosmed USA, Inc., Concord, CA, USA) were recorded. Physiological data from all subjects were used for analysis, whereas EMG data from the first 15 subjects (11 women, 4 men; age: 22.7 ± 5.7 years; height: 169 ± 5.8 cm; mass: 69 ± 12 kg) were used for the EMG analysis.


The participants completed 3 shod conditions walking on a level treadmill (Quinton 5000, Quinton Instrument Co., Seattle, WA, USA): (a) walking in flat-bottom shoe (Adidas) (W), (b) walking in flat-bottomed shoes weight-matched to RB (W and RB weigh differently and we wanted to account for differences in mass as a possible confounding variable), and (c) Rocker-bottom shoe with compliant midsole (Skecher Shape-Ups XF-Energy Blast) (RB) (Figure 1). Shoe sizes tested ranged from 9 to 12 for men and 7 to 9 for women for both shoe models. The RB shoes weighed 507 ± 89.7 g, whereas the W shoes weighed 318 ± 32.8 g. Before testing, the participants were shod with both pairs of shoes and given time to practice walking over ground and on the treadmill. The participants were then instrumented with electrodes on specific lower extremity muscles (right side biceps femoris [BF], rectus femoris [RF], GA, tibalis anterior [TA]) to measure surface EMG (Noraxon, Scottsdale, AZ, USA; sample rate: 1,500 Hz). Standard skin preparation procedures were used during EMG electrode application, and lead placement followed SENIAM recommendations (6,11). Oxygen consumption was measured using an ORCA metabolic cart (ORCA Cardiopulmonary System, Santa Barbara, CA, USA) that required the subjects to breathe through a mask fitted onto the mouth and nose. This mask was connected to the ORCA metabolic cart with concomitant breath-by-breath oxygen uptake (V[Combining Dot Above]O2) measured.

Figure 1
Figure 1:
Shod conditions: Typical walking shoe (W), typical walking shoe weight-matched to RB (WM), and rocker-bottom shoes (RB). (Photograph: ©; used with permission.)

The participants walked on the treadmill and were asked to adjust the treadmill speed to what they considered to be a comfortable walking speed at 0% grade while wearing W shoes. The test speed was set to 10% faster than the self-selected speed (mean test speed: 1.3 ± 0.2 m·s−1). For each shoe condition, the participants walked on the treadmill at the test speed for 10 minutes. Conditions were administered in a counterbalanced fashion to minimize any order effect, and there was about 2- to 5-minute rest between conditions for the participants to change shoes and have mass added to the shoe. The EMG data were collected at the beginning, middle, and ending of the 10-minute walk, whereas V[Combining Dot Above]O2 data were collected throughout the entire 10-minute walk. For presentation purposes only, EMG data were normalized such that the W beginning EMG was 100%.

Statistical Analyses

The EMG data were full-wave rectified, and the average (AVG) and root mean square (RMS) values were calculated over a 30-second period for each respective muscle. A 3 (shod) × 3 (time) repeated measures analysis of variance (ANOVA) was used to determine statistical significance among shod conditions at levels of time (i.e., beginning, middle, ending) of the exercise bout. A power analysis was conducted to determine the number of subjects needed to detect a 10% change in normalized muscle activity using estimates of variability between 10 and 40% with alpha set at 0.05 and beta at 0.80. This analysis yielded that at least 6 subjects (variability 10%) and as many as 76 subjects (variability = 40%) would be needed. This range of variability was selected because it reflected observed within-subject variability in the acquired data set.

The 10-minute mode, 10-minute average, and terminal 5-minute average of breath-by-breath V[Combining Dot Above]O2 (milliliters per kilogram per minute) data were calculated and used for statistical analyses. A 1-way repeated measures ANOVA was used to determine statistical significance among conditions for each V[Combining Dot Above]O2 measure of central tendency. Using a power analysis, the number of subjects required to detect a 10% change in expected V[Combining Dot Above]O2 for walking 1.3 m·s−1 with an SD of 1 ml·kg−1·min−1, alpha set at 0.05 and beta at 0.80 was 5 subjects.

A p value of ≤0.05 was considered statistically significant for all the tests.


Muscle activity for any muscle was not influenced by the interaction of shod and time (p > 0.05). There was no shod or time main effect (p > 0.05) for BF, RF, GA AVG, and RMS EMG. The TA AVG and RMS were also not different among time conditions (p > 0.05), but both parameters were different among shod conditions (p < 0.05). Using simple effects post hoc tests, it was determined that TA AVG and RMS EMG were greater during W vs. RB (p < 0.05) but not different from during W vs. WM (p > 0.05). The AVG EMG data for all muscles are presented in Figure 2 (RMS data followed the same pattern illustrated in Figure 2 and statistical results were identical to AVG EMG for each muscle). Intraclass correlation coefficients for EMG across times were all >0.900 for all muscles.

Figure 2
Figure 2:
Average electromyography (EMG) at the beginning (B), middle (M), and ending (E) of a 10-minute walk in flat-bottom shoes (W), flat-bottom shoes with added mass (WM), and rocker-bottom (RB) shoes of the biceps femoris (BF), rectus femoris (RF), gastrocnemius (GA), and tibialis anterior (TA).* Only the TA was different between shod conditions with EMG being greater during W vs. RB but no different than WM (p < 0.05).

Rate of oxygen consumption was not different among shod conditions using any measure of central tendency (i.e., 10-minute mode, 10-minute average, or terminal 5-minute average) (Figure 3; p > 0.05). Ten-minute average metabolic data are presented in Table 1 in units of METs, joules per minute, joules per kilogram per minute, kilocalories per minute for comparison with that of other studies. The intraclass correlation coefficient for V[Combining Dot Above]O2 (10-minute average) was 0.910.

Table 1
Table 1:
Average (±standard error) rate of oxygen consumption (V[Combining Dot Above]O2) during walking (1.3 ± 0.2·m·s−1) in W, WM, and RB.*†
Figure 3
Figure 3:
Oxygen consumption among shod conditions using 3 methods of central tendency (p > 0.05). Mode, 10-minute average and terminal 5 minutes were determined using 30 minutes of data from the subjects walking on a treadmill at 0% grade and 10% faster than a self-selected pace.


The main finding of this study was that RF, BF, and GA muscle activity and metabolic cost were unaltered while controlling walking speed in healthy adults wearing flat-bottomed shoes, rocker-bottom profile-type shoes, and weight-matched flat-bottomed shoes. Furthermore, TA activity was lower during walking in rocker-bottom, profile-type shoes vs. flat-bottomed shoes. Despite the dramatic change in shoe construction between flat-bottom and rocker-bottom shoes, it appears that there are no additional muscle activation or metabolic cost benefits of wearing RB beyond the typical, flat-bottomed shoe.

There is limited research on muscle activity response while walking in RB shoes. Overall, it seems that when the speed is controlled, muscle activity of lower extremity muscles does not seem to be different while walking in RB vs. flat-bottom shoes (3,15). However, there may be changes in muscle activity within a gait cycle. For example, Romkes et al. (17) observed lower TA muscle activity during the loading phase of stance and elevated muscle activity during the swing phase in the RB vs. flat-bottom shoes. In any case, the lack of change or minimal change in muscle activity during walking in RB vs. flat-bottom shoes among studies is in line with the observations of either no or minimal change in joint kinematics and kinetics (3,15,17). However, the RB shoe does seem to influence ankle kinematics and kinetics more so than other joints, albeit subtly (3,15,17).

A challenge in comparing data between and among studies is that there are several variations of rocker-profile shoes. For example, the location of the apex of the rocker-bottom may be different in shoes used in different studies. Wang and Hansen (21) reported that the shape of the rocker bottom influenced ankle kinematics during walking, and Harris et al. (8) reported that EMG patterns were influenced by the RB profile. Furthermore, it is not clear how the midsole stiffness varies between shoes–this would influence the stability of the surface (in essence). Nevertheless, taken together, the lack of change in EMG for the RF, BF, and GA observed in this study is consistent with the findings of other research as is the lower TA EMG during RB vs. W (3,15,17).

Rocker-profile shoes limit sagittal plane motion in the joints of the foot to create a rigid platform that helps maintain a consistent ankle-foot rollover shape, which subsequently forces the foot to roll forward when body mass passes over the fulcrum (10,20,21). Owing to the consistent rollover shape, studies have shown no difference in metabolic cost of wearing rocker-profile shoes compared with that of wearing flat-bottomed shoes and weight-matched flat-bottomed shoes (7,20). This study is in agreement with these data in that we did not observe any changes in metabolic cost despite accounting for differences in shoe weight and methods for determining central tendency. However, Van Engelen et al. (5) reported that the metabolic cost of walking in a rocker-profile shoe (MBT) was higher when compared with that of walking in flat-bottomed shoes or barefoot (2.33 vs. 2.08 and 2.01 J·kg−1·m−1, respectively). Differences among studies may be because of the shoe construction of the RB used. For example, the MBT shoe weighs more than the shoe used in this study and likely has a much higher compliance in the heel and less in the forefoot sections (7).

In our study, we required that the subjects walk at the same speed during each shoe condition. However, we do not know if the subjects would naturally change walking speed for each shoe condition. Romkes et al. (17) provide some insight into this because they allowed speed to vary between shoe conditions and reported that subjects walked slower while wearing MBT vs. typical walking shoes. A lower walking speed would seem to reduce the metabolic demand for maintaining the speed and reduce muscle activity. However, a lower walking speed may increase the energetic cost of locomotion (vs. rate of oxygen consumption) to cover a specific walking distance because there is a nonlinear relationship between energetic cost of locomotion and walking speed (9). Further research is needed to determine how RB shoes influence the energetic cost of locomotion and muscle activity while walking at self-selected speeds vs. matched speed between shoes. Based on our results, when speed is controlled, the energetic cost of locomotion is not different among shoe conditions. However, if walking speed was allowed to vary, it may be that a reduction in speed while wearing RB vs. flat-bottom shoes could influence energetic cost of locomotion.

The RB shoe condition was novel to subjects. Our experimental protocol allowed the subjects to wear the RB shoes before testing. Because the shoe condition is rather novel, we were concerned about potential changes to gait during the 10-minute test walk. Therefore, we sampled EMG data at 3 points (i.e., beginning, middle, and ending) during each condition. The lack of influence of time on any muscles is an indication that little or no gait adaptations were made during the 10-minute walk. Stöggl et al. (18) investigated the adaptation to walking in MBT shoes over a 10-week period of time. They reported no influence of time on EMG but did identify some kinematic adaptations. If there were kinematic changes during the 10-minute walk in our study, they were not significant enough to influence muscle activity. That being said, the subjects often anecdotally remarked that walking in the RB shoes felt different from walking in flat-bottom shoes. This observation may be related to the large variability (i.e., larger SDs) in EMG for the RF and BF muscles vs. the TA and GA muscles. If individual subjects responded differently to the shoe conditions, there is the risk of detecting a shoe effect. However, the source of the high variability seems to be outlier responses by 2 subjects. Therefore, we reanalyzed the data without those 2 subjects, and there was no change in outcome of the study. This is an indication that the general response was that the shoe condition did not influence muscle activity for RF, BF, and GA muscles but did influence TA muscle activity.

The subjects in our study were apparently healthy and had no injury that influenced the ability to walk. The RB profile shoe is used to assist gait for some injury conditions (e.g., plantarfacitis), and we do not know how people with lower extremity injuries respond to walking in RB vs. flat-bottom shoes. Furthermore, it is necessary to determine whether rocker-bottom, profile shoes have an effect on metabolic cost in older, overweight, and obese populations and whether walking speed, grade, and shoe accommodation confound the effect of RB.

Practical Applications

The use of rocker-profile shoes to treat ailments such as foot ulcers, lower back pain, arthritis, neck problems, and osteoporosis may be of clinical significance. We observed that there was either no change in muscle activity (RF, BF, GA) or lower muscle activity (TA) for healthy subjects while walking in the RB vs. W concurrent with no change in V[Combining Dot Above]O2. Therefore, the consumer who is healthy and without lower extremity injury should not expect additional benefits by wearing an RB shoe in muscle activation (i.e., training effect) and metabolic cost beyond that of a typical, flat-bottomed shoe. It should be noted, however, that if the purchase and use of these types of shoes motivate an individual to be more physically active, then the cost of investment may be substantiated.


1. Bogart JJV, Long JT, Klein JP, Wertsch JJ, Janisse DJ, Harris GF. Effects of the toe-only rocker on gait kinematics and kinetics in able-bodied persons. IEEE Trans Neur Syst Rehabil Eng 13: 542–550, 2005.
2. Brown D, Wertsch JJ, Harris GF, Klein J, Janisse D. Effect of rocker soles on plantar pressures. Arch Phys Med Rehabil 85: 81–86, 2004.
3. Buchecker M, Wagner H, Pfusterschmied J, Stoggl TL, Muller E. Lower extremity joint loading during level walking with Masai barefoot technology shoes in overweight males. Scand J Med Sci Sports 17: 1–9, 2010.
4. Busch K, Chantelau E. Effectiveness of a new brand of “stock” diabetic shoes to protect against diabetic foot ulcer relapse. A prospective cohort study. Diabet Med 20: 665–669, 2003.
5. Van Engelen SJPM, Wajer QE, Plaat LW, Van Der Doets HC, Van Dijk CN, Houdijk H. Metabolic cost and mechanical work during walking after tibiotalar arthrodesis and the influence of footwear. Clin Biomech 25: 809–815, 2010.
6. Frericks B, Hermens HJ. European Recommendations for Electromyography: Results of the SENIAM Project. The Netherlands: Roessingh Research and Development, 1999.
7. Hansen AH, Wang CC. Effect of rocker shoe radius on oxygen consumption rate in young able-bodied persons. J Biomech 44: 1021–1024, 2011.
8. Harris GF, Klein JP, Janisse D, Brown D, Shu Y, Wertsch J. Effect of rocker soles on lower extremity dynamic EMG patterns. Gait Posture 11: 157–158, 2000.
9. Holt KG, Hamill J, Andres RO. Predicting the minimal energy costs of human walking. Med Sci Sports Exerc 23: 491–498, 1991.
10. Hutchins S, Bowker P, Geary N, Richards J. The biomechanics and clinical efficacy of footwear adapted with rocker profiles—Evidence in the literature. Foot 19: 165–170, 2009.
11. Konrad P. The ABC of EMG: A Practical Introduction to Kinesiological Electromyography. Scottsdale, AZ: Noraxon, Inc., 2005.
12. Long JT, Klein JP, Sirota NM, Wertsch JJ, Janisse D, Harris GF. Biomechanics of the double rocker sole shoe: Gait kinematics and kinetics. J Biomech 40: 2882–2890, 2007.
13. Long JT, Sirota NM, Klein JP, Wertsch JJ, Janisse D, Harris GF. Biomechanics of the double rocker sole shoe: Gait kinematics and kinetics. Conf Proc IEEE Eng Med Biol Soc 2: 5107–5110, 2004.
14. Myers KA, Long JT, Klein JP, Wertsch JJ, Janisse D, Harris GF. Biomechanical implications of the negative heel rocker sole shoe: Gait kinematics and kinetics. Gait Posture 24: 323–330, 2006.
15. Nigg B, Hintzen S, Ferber R. Effect of an unstable shoe construction on lower extremity gait characteristics. Clinic Biomech 21: 82–88, 2006.
16. Nigg BM, Emery C, Hiemstra LA. Unstable shoe construction and reduction of pain in osteoarthritis patients. Med Sci Sports Exerc 38: 1701–1708, 2006.
17. Romkes J, Rudmann C, Brunner R. Changes in gait and EMG when walking with the Masai barefoot technique. Clin Biomech 21: 75–81, 2006.
18. Stöggl T, Haudum A, Birklbauer J, Murrer M, Müller E. Short and long term adaptation of variability during walking using unstable (Mbt) shoes. Clin Biomech (Bristol, Avon) 25: 816–822, 2010.
19. Uccioli L, Faglia E, Monticone C, Faveles F, Dunla L, Aldeghi A, Quarantiello A, Calia P, Menzinger G. Manufactured shoes in the prevention of diabetic foot ulcers. Diabetes Care 18: 1376–1377, 1995.
20. Vanderpool MT, Collins SH, Kuo AD. Ankle fixation need not increase the energetic cost of human walking. Gait Posture 28: 427–433, 2008.
21. Wang CC, Hansen AH. Response of able-bodied persons to changes in shoe rocker radius during walking: Changes in ankle kinematics to maintain a consistent roll-over shape. J Biomech 43: 2288–2293, 2010.

shod; EMG; metabolic cost; oxygen consumption; footwear

© 2012 National Strength and Conditioning Association