Influence of saddle height and exercise intensity on pedalling asymmetries in cyclists

Pedaling asymmetries quantified during stationary cycling, when cyclist body positioning and intensity remain unchanged, may not fully reproduce the training and competition situations, in which cyclists experience different intensities and may opt for different saddle positioning aiming at power output optimization. Previous studies showed that torque and power can be asymmetric in cyclists. It is not clear whether changes in saddle height and exercise intensity may affect asymmetries. The aim of the present study was to determine pedaling asymmetries during cycling at different saddle heights and different exercise intensities. Twelve competitive cyclists performed an incremental maximal test, a constant-load (“heavy” intensity domain), and a Wingate test. Constant load and the Wingate tests were repeated using three different saddle heights (reference and lower or higher by 2.5% of the distance from the pubic symphysis to the ground). Crank torque was recorded throughout the pedaling cycle. Asymmetry (higher torque for the preferred limb) was found in all saddle heights (p<0.001) in both intensities. Asymmetry index was similar across the saddle positions (p>0.05) in both intensities. Our results suggest that asymmetric cyclists present a consistent pattern regardless of small changes in the saddle height or in exercise intensity. For practical implication, cyclists producing asymmetric torque may be adapted to this condition so they are continuously exposed to asymmetric effort and overload on the lower limbs.


INTRODUCTION
Studies on pedaling asymmetries consider a constant body positioning sustained during short bouts of exercise mostly at fixed intensity.Although such investigations made a major contribution to the study of leg asymmetries in cycling, they did not consider the influence of changing saddle position as representative of small adjusts that cyclists do to optimize power output during a training or competition [1][2][3][4][5][6][7] .Small changes in saddle position are observed during cycling practice.During cycling to exhaustion, cyclists change their sitting position (i.e. more forward position) when performing at high levels of effort 8,9 .When exercising at supra-maximal efforts, altering the saddle position affects mean power output 10,11 , which is not observed at sub-maximal intensities 12 .Another source of deviations in performance is the influence of changing saddle height on force-length muscle relations and therefore force output 13,14 .One could argue that specific saddle positions could result in specific adaptation in length muscle relations, as observed for rectus femoris torque in response to different range of motions experienced by cyclists and runners 15 .Surprisingly, the question whether changes in saddle height position could influence the magnitude of pedaling asymmetries seems not to be addressed by previous studies.Pedaling asymmetry is often analyzed in a single saddle position, which limits application of this information for a specific race, for example, when cyclists are free to vary, even if by small amounts, the saddle position (i.e. during climbing and time-trials).
To address the effects of changing saddle height and exercise intensity on the magnitude of pedaling asymmetries, trained cyclists were evaluated assuming different saddle heights.The intensity of the exercise was also varied to determine effects of sub-maximal and supra-maximal intensity on pedaling asymmetries.Since cycling is an activity in which the motor pattern of pedaling is extensively repeated, the consistency of asymmetries during altered saddle positioning could suggest the occurrence of asymmetric performance in training and competition.

Participants
Twelve cyclists currently competing at regional and national level took part in this investigation.The mean and standard deviation age, height, body mass, body mass percentage, cycling experience, cycling training volume, maximal oxygen uptake (VO 2 max), and maximal power were 31.7 ± 5.9 years, 176.1 ± 6.1 cm, 73.8 ± 6.6 kg, 10.8 ± 2.9 %, 3.9 ± 4.1 years, 217.5 ± 103.2 km/week, 56.8 ± 3.8 ml•kg -1 •min -1 , and 316.4 ± 35.6 W, respectively.All participants were injury-free at the time of participation and signed an informed consent form before the start of the study, which was approved by the local Committee for Ethics in Research with Humans (protocol number 065/06) and in accordance with the Helsinki declaration.

Experimental Procedures
Tests were performed in four different days, with 48 h interval between each.On the first day, participants completed anthropometric assessments, performed a maximal incremental exercise test, and a familiarization trial to the Wingate test.The incremental test was used to determine parameters related to aerobic capacity of the athletes.The expired respiratory gases were collected and analyzed using a Quark PFTergo metabolic system (Cosmed, Rome, Italy) calibrated in accordance with the manufacturers' instructions.VO 2 max was defined as the highest oxygen uptake (VO 2 ) recorded over a 30 s period.Ventilatory threshold (VT) was determined using the ventilatory equivalent method 16 .Familiarization with Wingate included a 15 s maximal trial.On testing days 2-4, participants performed a 6 min constant-load test followed by a repetition of the Wingate anaerobic test.Constant load and Wingate tests were repeated at different saddle heights (one height per day).The order of the saddle heights tested was randomized.The standard saddle position was defined as the individual preferential position assumed by the participants during training and competitions 14 .This measure was taken from the own cyclist's bicycle.The two additional saddle heights tested were obtained by changing saddle up or down by 2.5% of the distance from the pubic symphysis to the ground 17 .All tests were performed in an electromagnetic braked cycle ergometer (Excalibur Sport ® , Lode, Netherlands).Sitting position was individually adjusted for each participant to correspond to their own bicycle with regards to horizontal position, saddle and handlebar heights.

Incremental test
The incremental test started at 100 W, with 30 W increments every 3 min until exhaustion 18 .Preferred cycling cadence was determined during the two initial stages of the test and thereafter participants were instructed to maintain that cadence throughout the test.Exhaustion was defined as the moment in which the participant could not maintain the cycling cadence anymore 18 .All participants were equally and strongly encouraged throughout the tests to perform to the best of their ability.

Constant-load test
The intensity of the constant-load test was determined using physiological parameters according to equation 1, and ensuring that all participants were exercising in the "heavy" intensity domain 19 . (1) Where: ∆50% corresponding to 50% of the difference between maximum oxygen uptake (VO 2 max) and oxygen uptake at ventilatory threshold (VO 2 @VT).
Before the start of the constant-load test, cyclists warmed up for 4 min at 30 W. Intensity was subsequently adjusted to ∆50% and sustained for 6 min.An active recovery interval of 4 min cycling at 30 W followed the constant-load test.Participants cycled at their preferred cadence throughout the test.After the test, participants were allowed to rest for 10 min before start the Wingate test 20 .

30-s Wingate anaerobic test
For the Wingate test, athletes were instructed to remain seated on the saddle and to perform at maximal effort throughout the test 11 .The load (resistance) used during the Wingate test was equivalent to 7.5% of the participant individual body mass 21 .Participants were verbally encouraged to perform at maximal effort during the entire test.The test was followed by 3 min of active recovery cycling at 50 W.

Torque analyses
The preferred lower limb was verified using the revised version of Waterloo Inventory 22 .Crank torque was measured every two degrees throughout the pedaling cycle (0-360º) using instrumented crank-arms (LEM -Excalibur Sport ® , Lode, Netherlands).Peak torque for each limb was defined as the highest value measured between 0º and 180º (propulsive phase) of the pedaling cycle.The asymmetry index was calculated using equation 2 23 , which provides the magnitude and direction of bilateral asymmetry in relation to the preferred limb. (2) Where: AI means the percent asymmetry index considering the ratio between peak torque measurements taken for the preferred (P) and non-preferred (NP) limbs.

Statistics
Data are presented as mean and standard deviation.Data normality, sphericity and homogeneity of variances were tested using the Shapiro-Wilk, Mauchly and Levene tests, respectively.Peak crank torques were compared between lower limbs and saddle heights using analysis of variance (ANOVA) for mixed linear models (2 lower limbs x 3 saddle heights) for each intensity test (constant load and Wingate).Bonferroni correction for multiple comparisons was performed when necessary.For a saddle height effect and repeated-measures ANOVA with Bonferroni post-hoc was used to compare the different saddle heights.
The asymmetry index was compared between the three saddle heights by repeated-measures ANOVA and Bonferroni post-hoc.Student t-test was applied to compare the asymmetry index between the constant-load and Wingate tests.All statistical tests were performed using SPSS for Windows (SPSS 17.0, USA) and a significance level of α = 0.05 was adopted.Effect size (ES) was calculated as Cohen's d to compare the magnitude of differences.The criteria to interpret the magnitude of the ES were 0.0-0.2trivial, 0.2-0.6 small, 0.6-1.2moderate, 1.2-2.0large and >2.0 very large 24 .

DISCUSSION
Our main findings support the concept that pedaling asymmetry can be consistent across different saddle heights when cyclists performed at two different intensities.Our results are in accordance with previous reports showing asymmetries in crank torque are in favor of the preferred limb 3,6,7 .The asymmetries observed across the different exercise intensities tested also elicited similar asymmetry indexes between trials with different saddle heights.From a practical perspective, our data suggest that cyclists may be subject to asymmetric performance and loading when performing at preferred saddle position, and also when small deviations from the reference saddle height are experienced.
The results suggested that pedaling asymmetry is present regardless of the saddle position tested.Even though former investigations suggest that changing saddle position might affect peak and mean power output 10,11 , few information is available concerning torque and torque asymmetry.Our results suggest that cyclists performing asymmetrically are highly consistent within their own pattern, which reinforces the possibility that they are also asymmetrical during training and competition.
The asymmetry index was similar between saddle heights at different exercise intensities, which suggest an asymmetric performance when pedaling at the reference position and even when small deviations from this position.Although this would suggest neuromuscular adaptation for force production, the magnitude of muscle activation when cycling at different conditions and intensities was symmetric in both cyclist and non-cyclists 25,26 .
A similar pedaling asymmetry was observed during both sub-maximal and maximal cycling trials.It has been suggested that torque asymmetry can be inversely associated with intensity during simulated time-trial competition 6 .However the results of the present study do not support the intensity effect reported.The differences in the outcomes between our study and the previous one 6 may rely on the fact that while we used a constant sub-maximal intensity, Carpes et al. 6 performed a simulated competition, in which participants were free to vary intensity during the trial.
An important question rising from the current results concerns the potential neuromuscular adaptations that might take place in asymmetric cyclists.To answer this question, a prospective or longitudinal study would be required.Furthermore, our results raise the question of whether correcting pedaling asymmetry can influence performance.Although not clear concerning an effect on performance, it was previously suggested that, among runners, any change in movement technique aiming at improve symmetry should be carefully conducted, since mechanical overload to muscles and joints increase injury risk through increased stress to these tissues 27,28 .For cyclists, training the muscles that cross the hip joint could contribute to decreases or even avoids lower limb asymmetries 5 .It is important to mention that quantification of the index of effectiveness was not possible in our study, but such variable could significantly help to discuss implications of pedaling technique in regard of the asymmetries observed.

CONCLUSION
The results of the current study demonstrate that pedaling asymmetry in trained road cyclists can be consistent across different saddle heights and even at different cycling intensities.From a practical point of view, our data suggest that cyclists producing asymmetric torque may be adapted to this condition so they are continuously exposed to asymmetric effort and overload on the lower limbs.

Figure 1 .
Figure 1.Mean and standard-deviation peak torque for preferred and non-preferred lower limb for saddle in the reference, downward, and upward position in the constant-load cycling.*Difference between limbs (p<0.05).

Figure 2 .
Figure 2. Mean and standard-deviation peak torque for preferred and non-preferred lower limb for saddle in the reference, downward, and upward position in the Wingate test.*Difference between limbs (p<0.05).