Secondary Logo

Pedaling Technique and Energy Cost in Cycling


Medicine & Science in Sports & Exercise: April 2011 - Volume 43 - Issue 4 - p 701-705
doi: 10.1249/MSS.0b013e3181f6b7ea

Purpose: Because cycling is an extreme endurance sport, energy saving and therefore efficiency is of importance for performance. It is generally believed that gross efficiency (GE) is affected by pedaling technique. A measurement of pedaling technique has traditionally been done using force effectiveness ratio (FE; ratio of effective force and total force). The aim of the present study was to investigate the relationship among GE, FE, and a new technique parameter, dead center (DC) size in competitive cyclists.

Method: Twenty-one competitive cyclists cycled for 10 min at approximately 80% V˙O2max at a freely chosen cadence (FCC). GE, FE ratio, and DC size were calculated from oxygen consumption and propulsive force recordings.

Results: Mean work rate was 279 W, mean FCC was 93.1 rpm, and mean GE was 21.7%. FE was 0.47 and 0.79 after correction for inertial forces; DC was 27.3% and 25.7%, respectively. DC size correlated better with GE (r = 0.75) than with the FE ratio (r = 0.50). Multiple regressions revealed that DC sizewas the only significant (P = 0.001) predictor for GE. Interestingly, DC size and FE ratio did not correlate with each other.

Conclusions: DC size is a pedaling technique parameter that is closely related to energy consumption. To generate power evenly around the whole pedal, revolution may be an important energy-saving trait.

Human Movement Science Programme, Norwegian University of Science and Technology, Trondheim, NORWAY

Address for correspondence: Stig Leirdal, M.Sc., Human Movement Science Programme, Norwegian University of Science and Technology, Dragvoll Idrettssenter 3 etg., 7491 Trondheim, Norway; E-mail:

Submitted for publication May 2010.

Accepted for publication August 2010.

Success in competitive cycling is dependent on a high muscular and aerobic power as well as an effective application of this power to the crank system of the bicycle. The energetic effectiveness (ability to save energy) in cycling is probably best quantified by gross efficiency (GE) (7). The pedaling technique by which this is established is still an issue of debate, but the force effectiveness (FE) ratio is often used to quantify at least one dimension of the quality of cycling technique (2,4,5,15,18,20,21,23). It is generally believed that proper technique is energy saving and thus gives high GE.

FE is defined as the ratio between the force component perpendicular to the crank arm (i.e., producing a moment of force and thereby work on the crank) and the total force generated on the pedals, or in other words, the sine of the angle between pedal force and crank. From an energetic-mechanical point of view, this should be the best "technique" parameter because the energy used for producing the ineffective static component of force will not contribute to the external power. However, it is not necessarily so that humans are able to produce the highest FE at the lowest metabolic cost: the coordinative challenge of generating power while creating a rotation of the crank by extending the lower extremity may require additional, apparently ineffective, energy expenditure. Only moderate relationships between GE and FE have been found (2,23). Furthermore, Coyle etal. (4) did find elite cyclists to have lower FE than subelite cyclists. Korff et al. (15) found GE to decrease with technique instructions that increased FE. In practice, FE never obtains its theoretical maximum, i.e., unity, because it is simply impossible to generate a high and perpendicular force on the cranks throughout the pedal revolution. Furthermore, FE and GE are strongly negatively influenced by cadence (1-3,9,10,17-20,22) and strongly positively influenced by work rate (5,7,16,20).

Another measure that might describe an aspect of cycling technique better is the (top and bottom) dead center size (DC). DC is defined as the minimal pedal work rate compared with the average work rate through the pedal cycle. The DC event is referred to in some literature and occurs at or close to the top and bottom position of the crank cycle, where both legs produce relatively low force with low effectiveness (e.g., Patterson and Pearson (18) and Sanderson (20)). It has not been used previously as a parameter to qualify technique. FE is a measure averaged over the entire pedal revolution. However, because very little (effective) force is produced in the upstroke (17), this is merely a measure for the downstroke only. The DC measure indicates to what extent a cyclist is able to overcome a technical challenge, i.e., producing a high effective force at a position where this seemingly is very difficult (top and bottom DC). Because it is a parameter based on work rate at the DC divided by the average work rate through the crank cycle, it is describing the evenness of work rate generation. The more even the crank cycle is, the lower the repetitive accelerations and decelerations of the cyclists inertia, thus likely resulting in a lower energy cost.

We hypothesize that effectiveness during the major propulsive parts of the pedal revolution, and thereby the FE for the entire crank cycle, does not discriminate the better from the worse cycling techniques. On the other hand, we propose that the force production and the effectiveness thereof during the DC parts of the revolution may be a better measure for quality of technique because this part of the crank cycle can be regarded as a serious challenge in cycling.

Therefore, the aim of the present study was to investigate the relationship among GE, FE, and DC during cycling. We used highly trained competitive cyclists, cycling at a freely chosen cadence (FCC), using a substantial but submaximal work rate to obtain close to purely aerobic metabolic conditions to calculate GE.

Back to Top | Article Outline



In the present study, 21 (19 male and 2 female) competitive cyclists participated. They were all well trained cyclists and competed at a regional, national, and some at an international level. Table 1 shows the demographics and physical characteristics of the subjects. The study was approved by the local ethics committee, and all subjects signed an informed written consent before participation.



Back to Top | Article Outline

Protocol and analysis.

Both an incremental maximal test (V˙O2max) and a submaximal pedaling technique cycling test were performed at FCC on a competition racer bike mounted on an electromagnetic roller (Tacx; I-Magic, Rotterdam, The Netherlands). The bike had 175-mm crank arms and gear ratios reaching from 39/24 to 53/12. This setup allowed subjects to use both cadence and gears to increase work rate as done during outdoor cycling. Seat and handlebar position were adjusted according to the subjects' individual preference. All subjects wore cycling shoes, and they were not allowed to do standing cycling during testing.

All subjects met in the lab on two occasions, separated by no less than 48 h. Subjects were told to avoid hard (>85%HRpeak) and long (>2 h·d−1) exercise the last 2 d before the V˙O2max test and the main experiment. On the first day, they performed the incremental V˙O2max test at FCC. After a warm-up at 150 W for 10 min, the test started at 100 W, and there was a 25-W increase every 2 min until exhaustion. Subjects were blinded to their cadence, and the FCC for each 2-min step was noted. All subjects met established criteria for attaining V˙O2max (13), and the highest 1-min average V˙O2 obtained during the test was subsequently used to determine their submaximal exercise intensity. Maximal HR (HRpeak) was defined as the highest value that was attained in average during a 5-s period at the final stage of the protocol.

On the second occasion, the subjects performed a submaximal test, starting with a 10-min warm-up at 50% of maximal aerobic power (MAP). Subsequently, they performed 5-min unloaded cycling (0 W, the chain was removed) at their FCC for the work rate corresponding to 80% V˙O2max, based on the results from the incremental V˙O2max test. The unloaded cycling was performed to account for inertial effects (14,17).

The main test consisted of 10 min of cycling at the last work rate in the V˙O2max test before the RER reached 1.00 (approximately 70% MAP, 80% V˙O2max, and 85%HRpeak). This work rate was chosen to obtain a work rate close to racing intensity and still not elicit to considerable anaerobic metabolism. During this test, the cadence was fixed at the subjects' FCC from the work rate corresponding to 80% V˙O2max at the V˙O2max test. Thus, subjects used the same gear ratio throughout the test and observed their cadence and work rate. They were requested not to vary cadence and work rate. During the last 5 min of the main part of the test, pedal forces and kinematics were recorded five times for 10 s at the start of each minute.

The average gas exchange values for these last 5 min were used for determining GE. GE was defined as the ratio work rate over metabolic cost rate as calculated from V˙O2 and RER values. Because the main test was done submaximally at RER values below 1.0, anaerobic contribution to metabolic cost was not included.

Gas exchange values were measured by open-circuit indirect calorimetry using an Oxycon Pro apparatus (Jaeger GmbH, Hoechberg, Germany). Before each measurement, the V˙O2 and V˙CO2 gas analyzers were calibrated using high-precision gases (16.0% ± 0.04% O2 and 5.0% ± 0.1% CO2; Riessner-Gase GmbH & Co., Lichtenfels, Germany). The flow meter was calibrated with a 3-L volume syringe (Hans Rudolph, Inc., Kansas City, MO). HR was measured with an HR monitor (Polar S610; Polar Electro Oy, Kempele, Finland), using a 5-s interval for data storage.

Crank and pedal kinematics were recorded using a ProReflex(Qualisys, Gothenburg, Sweden) three-dimensional motion capture system with eight cameras, in the same way as describedby Ettema et al. (8). Two spherical reflective markers were placed on extensions of both pedals in the sagittal plane of cyclist and bicycle. The positions of these markers were used to determine pedal orientation and crank angle. Both pedals were equipped, as presented previously (8,17), with two force cells (Revere model 9363, 2450 N capacity; Breda, The Netherlands), detecting pedal normal and shear forces. The pedals were calibrated by applying full normal forces and full shear forces of known magnitude. A constant proportional cross-talk between the normal and shear forces of a single pedal was detected (<3%) and taken into account by building a gain matrix.

All data were recorded using the QTM software (Qualisys) at a sample rate of 500 Hz and further processed in MATLAB (MathWorks, Natick, MA). All data were low-pass-filtered (10 Hz, eighth-order, zero-lag Butterworth). After correction for acceleration artifacts (6), pedal normal and shear forces were transformed to crank shear and normal forces by rotation of the coordination system from pedal to crank using the angle between pedal and crank as calculated from the kinematical data. The vector sum of right and left pedal forces (in the crank coordinate system) was used for further analysis.

The ratio of normal force over total force, measured over the entire crank cycle, was defined as FE (17). Power was calculated as the product of effective (normal) crank force and crank velocity. Continuous crank velocity was calculated from crank angles using a five-point differentiating filter. The average crank cycle for the 5 × 10-s measurement periods (for all variables) was calculated by interpolation of the crank angle-variable data to 360 samples, i.e., 1 sample per degree crank angle (8).

The lowest work rate (average of top and bottom DC work rate) divided by the average work rate was defined as DC. Thus, this is a parameter describing the evenness of work rate generation; 100% means a perfect circular work rate generation, while 0% indicates that the work rate at the DC equals zero.

The forces that were recorded during unloaded cycling are considered a good estimate of inertial forces (17). These were subtracted from the forces during loaded cycling to obtain an estimate of the muscular component of the pedal forces (17). Thus, the calculations on all force components are called gross values; those based on the muscular component are called net values.

Back to Top | Article Outline


Statistics were computed using Statistical Package for Social Sciences 15.0 (SPSS, Inc., Chicago, IL) using Pearson correlation coefficient. A multiple regression analysis was used for indicating the technique variables that were significant predictors for GE.

Back to Top | Article Outline


The submaximal work rate just below RER = 1.00 amounted to 279 ± 37 W at FCC of 93.1 ± 5.7 rpm. This work rate corresponded to approximately 70% MAP, 80% V˙O2max, and approximately 85%HRpeak. The average GE at this work rate and cadence was 21.7% ± 1.2%. The average FE at this work rate and cadence was 0.47 ± 0.05, and the average DC at this work rate and cadence was 27.3% ± 11.8%. When inertial forces were subtracted, the average FE was 0.79 ± 0.04 and the DC was 25.7% ± 10.0%.

We found a highly significant correlation between GE and both gross and net values of the DC r = 0.75 and r = 0.74, respectively (P < 0.01; Fig. 1). We also found a significant relationship between GE and FE, gross value, r = 0.50 (P = 0.022), but not when inertial forces were subtracted r = 0.28 (P = 0.22; Fig. 2).





Using a multiple stepwise regression analysis, DC seemed to be the only significant variable predicting GE. The total model (both DC and FE) yielded the following standardized coefficients (gross: DC β = 0.656, P = 0.001; FE β = 0.228, P = 0.177; net: DC β = 0.710, P < 0.001; FE β = 0.190, P = 0.235).

We found no significant relationships between FE and DC neither for gross values (r = 0.41, P = 0.067) nor for net values (r = 0.13, P = 0.582). The correlations between the gross and net values for both FE and DC are a high (r = 0.87, P < 0.01; r = 0.74, P < 0.01). It is noteworthy that DC is hardly affected by inertial forces (Fig. 3).



Back to Top | Article Outline


Our main results show a significant relationship between GE and DC and a more moderate relationship between GE and FE. However, multiple regression analysis showed DC to be a much stronger predictor GE than FE. Interestingly, there was no significant relationship between FE and DC.

The lack of a significant relationship between FE and DC indicates that FE and DC are two different and independent variables of cycling technique. Whereas DC describes the evenness of work rate generation during the entire crank, the FE ratio is more dependent on the force direction during the periods of high force, mainly the downstroke. Thus, in cycling, DC might be the superior "technique" parameter. It seems to be an indicator reflecting performance during the entire crank cycle rather than one focusing mostly on the downstroke as FE.

The FE during the pedal revolution seems to have limited effect on energy expenditure. There are several explanations for this finding. One explanation may be that the noneffective force component is "isometric" in nature (14), only producing compression or tension on the crank arm. Because isometric muscle contractions have a relatively low energy cost, it may have only a small effect on GE very much. Also, the FE is highly dependent on the downstroke because there is little effective force in the upstroke (17). If FE should be a "technique" parameter describing the quality of pedaling, it should differentiate between elite and nonelite cyclists. However, the downstroke work rate generation is fairly simple, and to date, no differences in FE between elite and subelite cyclists have been found. On the contrary, Coyle et al. (4) found the nonelite cyclists to have higher FE than elite cyclists. However, Coyle et al. (4) studied the different cyclists at a relative percentage of individual maximum. This means that the nonelite cyclists were cycling at a lower work rate than the elite cyclists. Because FE increases with work rate (5,20), one would expect a considerable difference between elite and nonelite cyclists and opposite of what Coyle et al. (4) found. The little between-subject variation in FE in the present study strengthens the notion that FE does not discriminate between different levels of cycling technique.

The FE ratio in the present study is well within the range found in other studies (2,4,15,18,20,21,23). The correlation between GE and FE ratio in the present study is low compared with other studies (2,23). However, in the present study, only FCC was examined, which may explain this contradiction.

The strong relationship between GE and DC indicates that evenness of power generation during the entire crank cycle is an important trait of pedaling about energy saving. By overcoming the DC, where it is difficult to generate power, a smooth work rate throughout the crank cycle is produced. Thus, fluctuations in work rate and force during the downstroke may be reduced while maintaining work rate. This may help the cyclists to maintain blood flow to the exercising muscles. Also, it might help the cyclists to share the work between more muscles or compartments of the active muscles. This might reduce local fatigue. Furthermore, an uneven work rate generation will enhance acceleration and deceleration periods through the crank cycle. From a power balance standpoint, the athlete does not fully maintain the amount of external kinetic energy (i.e., energy related to the velocity relative to the environment) by generating the same power that is lost due to external resistance. The athlete rather allows small reductions by producing a work rate less than the resistance work at the DC, to make up for this in the downstroke. It may well be that, for physiological reasons, this comes at an extra cost. However, simultaneously, the internal kinetic energy (i.e., related to rotating body segments) will also fluctuate more by an uneven power cycle. This internal kinetic energy may be used for propulsion (i.e., transformed as external power), but some of it is likely wasted as heat (7). It is therefore also likely that the amount of energy that is lost in this transformation is related to the amount of fluctuations and thereby the DC size. In the present study, the power was recorded at the pedal. Thus, the fluctuations in internal kinetic energy mostly relate to the motion of the lower limbs. In road cycling, these fluctuations would relate to the entire system, including crank inertia (11,12). Thus, the effect noted in this study on cycling on a roller bar may be larger in road cycling.

An interesting additional finding was that the subtraction of inertial forces has a substantial effect on the FE values but relatively little on the DC. Intersubject differences in FE become very small, and the relationship with GE becomes nonexistent. Thus, FE may merely be a measure related to the effect of inertial (nonmuscular) forces and not to the coordinative aspects of cycling. DC is hardly influenced by the inertial forces. Data for gross and net values lie closely around the line of identity for DC (Fig. 3) but are far more scattered for FE. These become even more apparent if the DC data point that could be considered an "outlier" is omitted (Fig. 3). Thus, the DC may be considered a parameter that represents the muscular component of the crank cycle. This is also supported by the fact that the relationship between FE and GE disappeared when we looked at the net FE. It should be noted that the force pedals introduce some extra inertia and likely inertial forces. Thus, the inertial effects in this study may slightly overestimate true inertial effects when using normal pedals.

We conclude that DC is a parameter more closely related to energy costs than the FE when cycling at FCC. To generate power evenly around the whole pedal revolution is an energy-saving trait and is highly related to GE. Future research should be directed toward the mechanisms behind DC, how DC is affected by cadence and work rate, and if it differentiates between cyclists at different performance levels.

This study received no external funding.

The authors thank the athletes who participated as subjects in the present study.

The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Belli A, Hintzy F. Influence of pedaling rate on the energy cost of cycling in humans. Eur J Appl Physiol. 2002;88:158-62.
2. Candotti CT, Ribeiro J, Soares DP, DeOliveira AR, Loss JF, Guimaraes ACS. Effective force and economy of triathletes and cyclists. Sport Biomech. 2007;6(1):31-43.
3. Coast JR, Welch HG. Linear increase in optimal pedal rate with increased power output in cycle ergometry. Eur J Appl Physiol. 1985;53:339-42.
4. Coyle EF, Feltner ME, Kautz SA, et al. Physiological and biomechanical factors associated with elite endurance cycling performance. Med Sci Sports Exerc. 1991;23(1):93-107.
5. Ericson MO, Nisell R. Efficiency of pedal forces during ergometer cycling. Int J Sports Med. 1988;9:118-22.
6. Ettema G, Huijing PA. Frequency response of rat gastrocnemius medialis in small amplitude vibrations. J Biomech. 1994;27:1015-22.
7. Ettema G, Lorås H. Efficiency in cycling: a review. Eur J Appl Physiol. 2009;106(1):1-14.
8. Ettema G, Lorås H, Leirdal S. The effects of cycling cadence on the phases of joint power, crank power, force and force effectiveness. J Electromyogr Kinesiol. 2009;19(2):94-101.
9. Foss Ø, Hallen J. The most economical cadence increases with increasing workload. Eur J Appl Physiol. 2004;92:443-51.
10. Foss Ø, Hallen J. Cadence and performance in elite cyclists. Eur J Appl Physiol. 2005;93:453-62.
11. Fregly BJ, Zajac FE, Dairaghi CA. Crank inertial load has little effect on steady-state pedaling coordination. J Biomech. 1996;29:1559-67.
12. Hansen EA, Jørgensen LV, Jensen K, Fregly BJ, Sjøgaard G. Crank inertial load affects freely chosen pedal rate during cycling. J Biomech. 2002;35:277-85.
13. Howley E. V˙O2max and the plateau-needed or not? Med Sci Sport Exerc. 2007;39(1):101-2.
14. Kautz SA, Hull ML. A theoretical basis for interpreting the force applied to the pedal in cycling. J Biomech. 1993;26(2):155-65.
15. Korff T, Romer LM, Mayhew I, Martin JC. Effect of pedaling technique on mechanical effectiveness and efficiency in cyclists. Med Sci Sports Exerc. 2007;39(6):991-5.
16. Leirdal S, Ettema G. Freely chosen pedal rate during free cycling on a roller and ergometer cycling. Eur J Appl Physiol. 2009;106(6):799-805.
17. Lorås H, Leirdal S, Ettema G. Force effectiveness during cycling at different pedaling rates. J Appl Biomech. 2009;25:85-92.
18. Patterson RP, Pearson JL. The influence of flywheel weight and pedalling frequency on the biomechanics and physiological response to bicycle exercise. Ergonomics. 1983;26(7):659-68.
19. Samozino P, Nicolas H, Hintzy F. Interactions between cadence and power output effects on mechanical efficiency during sub maximal cycling exercises. Eur J Appl Physiol. 2006;97:133-9.
20. 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 Sport Sci. 1991;9:191-203.
21. Sanderson DJ, Black A. The effect of prolonged cycling on pedal forces. J Sport Sci. 2003;21:191-9.
22. Seaury JJ, Adams WC, Ramey MR. Influence of pedalling rate and power output on energy expenditure during bicycle ergometry. Ergonomics. 1977;20(5):491-8.
23. Zameziati K, Mornieux DR, Belli A. Relationship between the increase of effectiveness indexes and the increase of muscular efficiency with cycling power. Eur J Appl Physiol. 2006;96:274-81.


©2011The American College of Sports Medicine