Secondary Logo

Journal Logo

Original Research

Use of the Wattbike Cycle Ergometer for Attenuation of Bilateral Pedaling Asymmetry in Trained Cyclists

Kell, David T.; Greer, Beau K.

Author Information
Journal of Strength and Conditioning Research: February 2017 - Volume 31 - Issue 2 - p 468-473
doi: 10.1519/JSC.0000000000001495
  • Free



Humans tend to preferentially use one side of their body during motor acts (5). This lateral preference may result in a bilateral asymmetry during walking, running, and pedaling (2,5,19). In regards to cycling, bilateral pedaling asymmetries have been reported in force (12,22), crank torque (6,7), work (8), power output (23), and kinematics (13). For power and work output, pedaling asymmetries generally range from 5 to 20% (8,12,22), with the self-reported dominate (kicking) leg typically producing more power (6,23).

Pedaling becomes more symmetrical in regards to power output as exercise intensity increases, even to the point of nearly perfect symmetry at maximal power outputs (6,7,21). Therefore, large bilateral pedaling asymmetries are most prevalent at low-to-moderate intensities, the latter of which are the most common intensities used during both training and racing (14,16,26). Supporting this contention, Carpes et al. (6) reports that cyclists in race settings pedal with a significant degree of bilateral asymmetry.

Several studies (11,18,20) using a split-belt treadmill (a treadmill with a belt for each leg that can move at different speeds) show that individuals can adapt and acutely maintain a new style of walking which is a locomotor act similar to pedaling in reverse (24). Use of conscious control, visual feedback (VF), and instructions on how to step were all factors that led to a quick adaptation of a new walking style, but there was only a short maintenance of that adaptation when the treadmill belts were tied (going the same speed) postintervention (18). In a separate investigation, interlimb walking parameters adapted slowly but had large aftereffects when the treadmill belts were tied after the adaptation period (20). Based on these data, and because muscles during reverse pedaling and walking have similar muscle activation function and phasing (24), a new pedaling style may be able to be adapted and stored similar to those found in studies of walking. Although no study to our knowledge examines adaption of a more bilaterally symmetrical pedaling style, 1 study shows that a new pedaling style could be adopted in regards to single leg crank torque trajectory (25).

Although it has been contended that asymmetry may be the preferred method of locomotion (5), attenuating asymmetry potentially decreases the chance of injury for cyclists. Pedaling asymmetry is related to differences in extensor and flexor moments of the hip and knee joints (23). Even with bilateral pedaling symmetry for pedal force, joint and muscle torques can differ between legs leading to muscle and joint overload (13); pedaling asymmetrically would possibly exacerbate muscle and joint overload leading to an increased overuse injury risk. In regards to performance, cycling training has been shown to lower muscle activation variability (4), improve muscle synergy (9), and increase force control (3). All of these changes could lead to increased efficiency and therefore decreased metabolic cost, as has been shown for walking (15,20). As chronic asymmetrical pedaling may hinder some of these long-term adaptive mechanisms, performance-related training benefits may be attenuated as well.

Therefore, it is plausible that individuals could adopt a more bilaterally symmetrical pedaling style, especially when given VF and practice. The Wattbike cycling ergometer measures power output from each leg and provides VF to the rider, making it a potential training tool for developing a more bilaterally symmetrical pedaling style. To date, it is unknown whether the Wattbike is effective in facilitating a more symmetrical pedaling style in significantly asymmetrical pedaling cyclists, and if any improvement would be more effective than attempts to consciously control power output bilaterally. With these findings, future research may examine whether a more bilaterally symmetrical pedaling style is sustainable outside of a laboratory environment or if it would influence cycling performance or injury risk.


Experimental Approach to the Problem

This study was designed to assess whether feedback provided by the Wattbike can significantly attenuate bilateral asymmetries in regards to pedaling power output. Subjects performed 3 distinct intervals of cycling; one in which no purposeful effort was made to pedal symmetrically, one in which subjects were instructed to try to pedal symmetrically, and one in which subjects were instructed to pedal symmetrically while continuous VF displaying percent of total power output from each leg was provided. Pedaling asymmetries were determined by averaging power outputs for each leg over 5-minute periods in each condition.


Twelve (7 men, 5 women, age range = 29–59 years) subjects were recruited for this study. Sample size estimation was based on pedaling asymmetry effect sizes reported by Carpes et al. (2007). Criteria for participation included an age range of 18–60 years, at least 12 consecutive months of cycling experience, and absence of signs or symptoms suggestive of cardiovascular, pulmonary, or metabolic disease as outlined by the American College of Sports Medicine (1). Both men and women were recruited; as to our knowledge, no data exist suggesting any gender-based influences on pedaling asymmetry. All subjects provided informed consent before participating in this study, and all protocols were approved by the Sacred Heart University Institutional Review Board.


The general study design is shown in Figure 1. After subjects provided their age, cycling experience, and health history, height and weight were recorded using a standard physician's scale (Seca Model 707; Columbia, MD, USA). Subject characteristics are shown in Table 1. Subjects were allowed to adjust the Wattbike cycle ergometer (Wattbike Ltd, Nottingham, United Kingdom) seat height, seat fore/aft adjustment, handle bar height, and handle bar fore/aft adjustment to their comfort; adjustments were recorded and used during all subsequent testing procedures. No subject had previous riding experience with the Wattbike.

Figure 1.
Figure 1.:
Study design.
Table 1.
Table 1.:
Mean ± SD of subject characteristics.*

The Wattbike is an acceptably valid and reliable tool in regard to measuring power output (150–300 W), and tested using both trained and untrained subjects (17). In a trained population specifically, the reliability of the Wattbike was 2.6% (95% CI, 1.8–5.1%) (17). Subjects used their own clip-less pedals and compatible cycling shoes. Subjects were instructed to avoid prolonged or intense exercise for 48 hours before initial testing.

After a brief, self-selected warm-up (intensity <150 W), subjects performed an incremental cycling test to volitional exhaustion to determine V̇o2 peak. Subjects were instructed to maintain a cadence of 80 revolutions per minute (RPM) throughout the duration of the test. Eighty RPM were used throughout all testing procedures as it was the median cadence used in a previous study investigating effects of cadence on the biomechanics of force application during cycling (21); the median was chosen as the effect of cadence on pedaling asymmetry shows high intersubject variability (23).

Cadence was maintained with the use of VF on a computer laptop screen connected to the Wattbike monitor. The use of a laptop screen results in an approximate 3-second delay between real-time output and the display. Workload was initially set at 150 W and increased every 2 minutes by 20 W for women and 35 W for men until subjects could not achieve a cadence greater than or equal to 80 RPM within a 5-second period, or until volitional exhaustion. A ParvoMedics' Metabolic TrueOne 2400 metabolic cart (ParvoMedics, Sandy, UT, USA) was used to collect metabolic data, and V̇o2 peak was defined as the highest average recorded V̇o2 over a 30-second interval.

At least 48 hours after V̇o2 peak assessment, subjects returned to the laboratory for subsequent testing. Once again, subjects were discouraged from participation in prolonged or high-intensity exercise for 48 hours before testing. Similar to V̇o2 peak testing, subjects engaged in a self-selected, low-intensity warm-up on the Wattbike. For all the subsequent testing periods, intensity was set at a power output consistent with 60% V̇o2 peak and cadence held at 80 RPM.

For the first 10-minute baseline (BASE) testing interval, deception was used as subjects were instructed to pedal “normally” but were unaware that data were being collected for analysis. For a second 10-minute interval, subjects were assigned through random block assignment to one of the following 2 conditions: a conscious control (CC) interval, in which subjects were instructed to consciously try to pedal symmetrically, or a VF interval, in which subjects were given the same instruction but provided VF on a laptop computer screen displaying the percentage of total power each leg was contributing. For the third and final testing interval, subjects performed the alternate condition (CC or VF). A 5-minute rest period was provided between all 3 testing intervals to minimize any confounding effects that fatigue may have on results.

Data were only analyzed for the last 5 minutes of each interval to reduce the potential of a learning curve affecting the results and to eliminate any influence that endurance-oriented fatigue may have on pedaling asymmetry. Pedaling asymmetry for each interval was determined by calculating an Asymmetry Index percentage (AI%) using the following equation: AI% = [(DO-ND)/DO] ×100 in which DO is the mean dominate leg power output and ND is the mean nondominant leg power output. The dominate (DO) leg was considered whichever leg produced more power during the specific interval. This equation has previously been used to calculate asymmetry in both walking (10) and pedaling activities (6).

Statistical Analyses

All dependent variables (cadence, AI%, dominant and nondominant leg power, and dominant and nondominant % contribution to total power) were tested for normality using a Shapiro-Wilk test. If data were normally distributed, a repeated measures ANOVA with Fisher's LSD post hoc was used to detect any significant differences. If any variable within a data set was not normally distributed, a Friedman's test was used with Wilcoxon signed-rank test for post hoc analysis. Order effect for AI% was also examined to ensure order of condition did not bias statistical findings. After the preliminary analysis, a secondary analysis was performed including only subjects who had an AI% at baseline within the normal range reported in the literature (5–20%) (8,12,22). Level of significance was set a priori at p ≤ 0.05. All statistical analyses were conducted using SPSS version 23.0 (SPSS, Inc., Chicago, IL, USA).


Mean cadence for the 3 conditions were not significantly different (Χ2 (2) = 0.500, p = 0.779), suggesting that any minor variances in cadence did not affect other analyses. The mean AI% for the BASE, CC, and VF conditions were not significantly different (Χ2 (2) = 3.167, p = 0.205). Intersubject variations in AI% are shown in Figure 2. As expected, because of inclusion of both genders and consequently greater variance, no differences were found between conditions in regards to dominant leg [F (1,14) = 0.435, p > 0.05] and nondominant leg [F (1,14) = 2.70, p > 0.05] power outputs. There was no order effect present in regards to AI% (Χ 2 (2) = 1.500, p = 0.472). All data from the preliminary analysis are shown in Table 2.

Figure 2.
Figure 2.:
Asymmetry Index (individual subjects). BASE = baseline trial; CC = conscious control trial; VF = visual feedback trial. No significant differences between trials were found in the primary analysis.
Table 2.
Table 2.:
Mean ± SD for cycling data (N = 12).*

The secondary analysis revealed a statistically significant difference [F (2,12) = 5.303, p ≤ 0.05] between BASE and VF conditions, whereas cadence again did not differ between trials [F (2,12) = 0.635, p > 0.05]. There was an order effect present for the secondary analysis [F (2,12) = 4.610, p ≤ 0.05], but post hoc analysis revealed the significant difference only existing between the first trial (BASE) and the third (CC or VF). All data from the secondary analysis are shown in Table 3.

Table 3.
Table 3.:
Mean ± SD for cycling data: Secondary analysis (N = 7).*


The central finding of this study was that subjects who had an AI% within the reported normal range were able to significantly improve their bilateral pedaling symmetry with the use of VF. These findings support conclusions of alternate studies that have demonstrated the value of VF in the adoption of new pedaling (25) and walking (18) styles.

To our knowledge, this is the first investigation directly examining whether cyclists could adopt a more bilaterally symmetrical pedaling style. Although post hoc analysis did not reveal a statistically significant difference between the CC and VF trials for the secondary analysis, the 46.2% relative decrease in asymmetry when using VF suggests that the secondary analysis may have been underpowered.

No differences in asymmetry were found between trials in the primary analysis (p = 0.21) despite 75.9% and 48.3% greater relative asymmetry in the BASE and CC trials, respectively, as compared with the VF trial. Despite these large effects, large intersubject variance prevented statistical significance from being found. Ultimately, it is unsurprising that VF was unable to significantly decrease AI%s considering the numbers of subjects with lower initial AI%s than typically reported. Interestingly, the 4 subjects with the lowest AI%s (<2.0%) at baseline all displayed an increase AI% during the VF interval; although the mean increase was small by an AI% standpoint (≤2.0%), the relative increase (244%) was quite high. Consequently, the use of VF may be ergolytic in particular populations if the goal is to reduce asymmetry.

Although previous studies have shown bilateral pedaling asymmetries typically ranging from 5 to 20% (8,12,22), the range in this study was 0.1–16.0% during the BASE trial, with only 7 of the 12 subjects having an AI% ≥ 5%. As intensity and bilateral pedaling asymmetry are inversely related (6,7,21), it was expected that the moderate intensity used in this study (60% of power at V̇o2 peak) would have produced higher AI%s during the BASE trial. Trained cyclists display less variability in muscle activation during cycling, and Carpes et al. (4) (2011) suggest that this variability in muscle activation may be a cause for asymmetries. Consequently, and in addition to mean V̇o2 peaks and self-reported training history, the comparatively lower AI%s in this study strengthen the contention that the cyclists were well trained.

Now that it has been demonstrated that a more bilaterally symmetrical pedaling style is attainable in typically asymmetrical cyclists with the use of VF in a laboratory setting, subsequent studies may be conducted to examine if the new pedaling style improves cycling efficiency or decreases overuse injury risk. Carpes, Mota, and Faria contend that an asymmetrical gait may be the preferred method of locomotion as asymmetries occur at self-selected movement frequencies for both running and pedaling and disappear at higher ones (5). Consequently, it is unknown whether symmetry is always an appropriate goal.

Additionally, it also remains unknown whether a new pedaling style adopted with VF in a laboratory setting can be reproduced in an applied setting, presumably with VF cues removed. The duration of training with VF necessary for this hypothetical transfer of skill to occur also requires investigation. Likewise, it remains unknown how cycling duration will affect pedaling asymmetry as the trials in this study were only 10 minutes. Although the effects of fatigue stemming from high oxygen demands have been indirectly assessed (7), the effect of endurance-oriented fatigue on pedaling asymmetry warrants investigation.

Practical Applications

An increase in performance coupled with a decrease in overuse injury rates are typically the primary goals of practitioners working with cyclists; increasing the bilateral symmetry of cyclists' pedaling has the potential to achieve both ends. The results of this study suggest that such a pedaling style can be adopted in a controlled setting. Additionally, the results indicate that while utilizing VF via the Wattbike Cycling Ergometer is an effective tool in helping experienced cyclists adopt a more bilaterally symmetrical pedaling style, it could be counterproductive in cyclists who already have highly symmetrical pedaling styles. Consequently, cycling coaches may wish to use the Wattbike as a testing modality for athletes as to determine which may be the best candidates for training targeted toward improving bilateral symmetry. It remains unknown whether this would lead to greater cycling economy or injury prevention, so coaches must rely on anecdote until subsequent research is conducted.


The authors thank Dr. Christopher Restrepo, M.D., for his assistance in figure preparation. They report no conflict of interest, and the results of this study do not constitute endorsement of the Wattbike by the authors or the NSCA.


1. American College of Sport Medicine. ACSM’s Guidelines for Exercise Testing and Prescription (9th ed.). Philadelphia, PA: Lippincott Williams & Williams, 2013.
2. Ashton GC. Handedness: An alternative hypothesis. Behav Genet 12: 125–147, 1982.
3. Bernardi M, Felici F, Marchetti M, Montellanico F, Piacentini MF, Solomonow M. Force generation performance and motor unit recruitment strategy in muscles of contralateral limbs. J Electromyogr Kinesiol 9: 121–130, 1999.
4. Carpes FP, Diefenthaeler F, Bini RR, Stefanyshyn DJ, Faria IE, Mota CB. Influence of leg preference on bilateral muscle activation during cycling. J Sports Sci 29: 151–159, 2011.
5. Carpes FP, Mota CB, Faria IE. On the bilateral asymmetry during running and cycling - A review considering leg preference. Phys Ther Sport 11: 136–142, 2010.
6. Carpes FP, Rossato M, Faria IE, Mota CB. Bilateral pedaling asymmetry during a simulated 40-km cycling time-trial. J Sports Med Phys Fitness 47: 51–57, 2007.
7. Carpes FP, Rossato M, Faria IE, Mota CB. Influence of exercise intensity on bilateral pedaling symmetry. In: Progress in Motor Control IV (Vol. 11) (11th ed.). Duarte M, Almeida GL, eds. Sao Paulo, Brazil: Human Kinetics, 2007. pp: S54–S55.
8. Cavanagh P, Petak K, Shapiro R, Daly D. Bilateral asymmetry in work output during cycle ergometer pedalling. Med Sci Sports Exerc 6: 80–81, 1974.
9. Chapman AR, Vicenzino B, Blanch P, Hodges PW. Patterns of leg muscle recruitment vary between novice and highly trained cyclists. J Electromyogr Kinesiol 18: 359–371, 2008.
10. Chavet P, Lafortune MA, Gray JR. Asymmetry of lower extremity responses to external impact loading. Hum Mov Sci 16: 391–406, 1997.
11. Choi JT, Bastian AJ. Adaptation reveals independent control networks for human walking. Nat Neurosci 10: 1055–1062, 2007.
12. Daly DJ, Cavanagh PR. Asymmetry in bicycle ergometer pedalling. Med Sci Sports 8: 204–208, 1976.
13. Edeline O, Polin D, Tourny-Chollet C, Webber J. Effect of workload on bilateral pedaling kinematics in non-trained cyclists. J Hum Move Stud 46: 493–518, 2004.
14. Fernández-García B, Pérez-Landaluce J, Rodríguez-Alonso M, Terrados N. Intensity of exercise during road race pro-cycling competition. Med Sci Sports Exerc 32: 1002–1006, 2000.
15. Goble DJ, Marino GW, Potvin JR. The influence of horizontal velocity on interlimb symmetry in normal walking. Hum Mov Sci 22: 271–283, 2003.
16. Golich D, Broker J. SRM bicycle instrumentation and the power output of elite male cyclists during the 1994 Tour Dupont. Perform Cond Cycling 2: 6, 1996.
17. Hopker J, Myers S, Jobson SA, Bruce W, Passfield L. Validity and reliability of the Wattbike cycle ergometer. Int J Sports Med 31: 731–736, 2010.
18. Malone LA, Bastian AJ. Thinking about walking: Effects of conscious correction versus distraction on locomotor adaptation. J Neurophysiol 103: 1954–1962, 2010.
19. McCartney G, Hepper P. Development of lateralized behaviour in the human fetus from 12 to 27 weeks’ gestation. Dev Med Child Neurol 41: 83–86, 1999.
20. Reisman DS, Block HJ, Bastian AJ. Interlimb coordination during locomotion: What can be adapted and stored? J Neurophysiol 94: 2403–2415, 2005.
21. Sanderson DJ. The influence of cadence and power output on the biomechanics of force application during steady-rate cycling in competitive and recreational cyclists. J Sports Sci 9: 191–203, 1991.
22. Sargeant AJ, Davies CT. Forces applied to cranks of a bicycle ergometer during one- and two-leg cycling. J Appl Physiol Respir Environ Exerc Physiol 42: 514–518, 1977.
23. Smak W, Neptune RR, Hull ML. The influence of pedaling rate on bilateral asymmetry in cycling. J Biomech 32: 899–906, 1999.
24. Ting LH, Kautz SA, Brown DA, Zajac FE. Phase reversal of biomechanical functions and muscle activity in backward pedaling. J Neurophysiol 81: 544–551, 1991.
25. Ting LH, Raasch CC, Brown DA, Kautz SA, Zajac FE. Sensorimotor state of the contralateral leg affects ipsilateral muscle coordination of pedaling. J Neurophysiol 80: 1341–1351, 1998.
26. Vogt S, Schumacher YO, Blum A, Roecker K, Dickhuth HH, Schmid A, Heinrich L. Cycling power output produced during flat and mountain stages in the Giro d’Italia: A case study. J Sports Sci 25: 1299–1305, 2007.

asymmetry index; visual feedback

© 2016 National Strength and Conditioning Association