Secondary Logo

Journal Logo

Muscle Coordination during an Outdoor Cycling Time Trial

BLAKE, OLLIE M.; WAKELING, JAMES M.

Medicine & Science in Sports & Exercise: May 2012 - Volume 44 - Issue 5 - p 939–948
doi: 10.1249/MSS.0b013e3182404eb4
APPLIED SCIENCES
Free

Introduction/Purpose Muscle activity in cycling has primarily been studied in the laboratory; however, conclusions are limited by the ability to recreate realistic environmental conditions. The purpose of this study was to determine muscle coordination patterns in an outdoor time trial and investigate their relationships to power output (PO), total muscle activity (Itot), overall mechanical efficiency (ηO), cadence, and gradient.

Methods Surface EMG, gradient, and cycling parameters were measured while cycling 18.8 km outdoors. A principal component analysis was used to establish coordination patterns that were compared with Itot, ηO, PO, cadence, and gradient.

Results PO was positively correlated with Itot, and high PO was associated with elevated rectus femoris and vastus lateralis activity and synchronization of muscles crossing the same joint. PO and cadence demonstrated positive and negative relationships, respectively, with gradient. Relationships between muscle coordination, PO, ηO, Itot, and gradient showed that muscle coordination, PO, and ηO fluctuate during an outdoor time trial as a result of pacing and gradient. A trade-off existed between ηO and PO, and ηO was dependent on muscle activation around the top and bottom of the pedal cycle and activity in more than the knee extensor muscles. Fluctuations in muscle activity due to the changing PO, from pacing and terrain, seemed to mitigate fatigue indices seen in indoor cycling studies.

Conclusions This study provides evidence that muscle activity is dependent on the terrain aspects of the cycle course as muscle coordination changes with the altered locomotor demands. The coordination patterns significantly covaried with PO, Itot, ηO, cadence, and gradient, which highlights the importance of recording these parameters under field conditions and/or careful reproduction of outdoor environments in indoor studies.

Department of Biomedical Physiology and Kinesiology, Simon Fraser University, Burnaby, British Columbia, CANADA

Address for correspondence: Ollie M. Blake, Department of Biomedical Physiology and Kinesiology, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, Canada V5A1S6; E-mail: omb@sfu.ca.

Submitted for publication May 2011.

Accepted for publication October 2011.

Muscle activity in cycling has been well studied in the laboratory, yet there is a lack of research outdoors in a realistic cycling situation. From laboratory studies, the general muscle coordination patterns of a pedal cycle have been described where the timing and amplitude of activation underlying the patterns are dependent on cadence, workload, pedals and shoes, body position, fatigue, and training status (see the review of Hug and Dorel ([12]).

Each muscle of the lower extremity has its own function during cycling. The medial (VM) and lateral (VL) vastii are thought to be the primary power producers in cycling (21) displaying the highest amount of activity relative to maximum voluntary contractions regardless of the workload (8). In general, single-joint muscles such as the vastii, gluteus maximus (GM), and soleus (Sol) are thought to function primarily as power producers, whereas biarticulate muscles such as the rectus femoris (RF) and biceps femoris (BF) are thought to transfer the power between joints (25). The VL (2,14), RF (2,14,23), and VM (14) have shown significant relationships to power output in time trials (2,23) and incremental cycling to exhaustion (14). The VL has also displayed increases in activity during a 40-km indoor time trial on a stationary bicycle (2), whereas the same increases were not observed in the 40-km time trial or in 30-min and 100-km time trials for other muscles including the tibialis anterior (TA), medial gastrocnemius (MG), BF, and RF (2,7,23).

It is important to address muscle performance and behavior in both indoor and outdoor cycling because of possible discrepancies. Environmental conditions in outdoor cycling, such as the gradient, influence both the cadence and power output in male cycling competitions. Cyclists use lower cadences (13) and higher power outputs in mountainous versus flat stages (17,26) of male multistage cycling races. They also have higher mean power outputs in time trials versus group stages (15). It is necessary to consider the environmental influences on cadence and workload because both cadence and workload affect muscle coordination in cycling (12,30).

The pacing strategy used in a time trial also influences the muscle coordination patterns. Pacing strategies disperse the workload required to complete a cycling event in different ways (see the review of Abbiss and Laursen ([1]), which affects both the power output and fatigue levels, both of which have an effect on muscle coordination (12). There is conflicting evidence regarding the interaction between pacing and muscle activation. Hettinga et al. (10) found that VL and BF activity increased throughout a 4000-m time trial despite negative (increasing power output), positive (decreasing power output), and even (constant power output) power output pacing strategies (1). Conversely, St Clair Gibson et al. (23) found that RF activity declined in conjunction with power output in a 100-km time trial with high-intensity bouts. With only the RF measured, they speculated that other muscles could have acted in compensation, thereby altering muscle coordination.

It is important to conduct cycling research in a natural setting because conclusions and correlations from laboratory studies are limited by the ability to recreate realistic environmental conditions. For example, muscle activation research in cycling has primarily focused on the effects of progressively increasing or constant workloads on a limited number of muscles. This is in contrast to the reality of cycling where many muscles are active and workloads fluctuate according to the changing environment. This is a limitation of simulated time trial experiments where participants control the resistance because changes in cadence and power output dictated by the environment are not considered. The purpose of this study was to determine the coordination patterns in time trial cycling outdoors and investigate their relationships to power output, total muscle activity, overall mechanical efficiency (ηO), cadence, gradient, and distance.

Back to Top | Article Outline

METHODS

Cycling protocol and data acquisition

Nine competitive male cyclists (primarily competing in mass start, criterium, and 5- to 40-km time trial road cycling races) cycled four laps of approximately 4.7 km each on paved roads (on a loop all of the cyclists use regularly for training) in the shortest time possible. The participants gave their informed written consent, and all procedures were approved by the ethics committee in accordance with the Office of Research Ethics at Simon Fraser University. Each lap of the course started and finished at the highest elevation and consisted of a descent followed by rolling and flat terrain before the final climb (Fig. 1). The difference between the highest and lowest elevations was approximately 22 m. The participants were all tested at the same time of the morning within a 3-wk period, and wind and temperature measurements were taken at three different locations on the course for each participant.

FIGURE 1

FIGURE 1

Muscle activity of 10 leg muscles (TA, MG, lateral gastrocnemius (LG), Sol, VM, RF, VL, semitendinosus (ST), BF, and GM) was continuously recorded via surface EMG using Norotrode bipolar Ag/AgCl surface electrodes (10-mm diameter, 21-mm interelectrode distance; Myotronics, Kent, WA). The signals were amplified (Biovision, Wehrheim, Germany) and sampled at 2000 Hz using a 16-bit analog-to-digital converter (USB-6210; National Instruments, Austin, TX) and the LabVIEW software (National Instruments). The time-varying intensity of the EMG signals (IEMG) was calculated for each muscle by wavelet techniques using 10 wavelets (k = 1–10 [27]), which act as a band-pass filter (bandwidth approximately 11–432 Hz). Unfortunately, because of the nature of outdoor data collection, many difficulties with noise and displacement of the EMG electrodes could not be detected or fixed during the test. Pedal cycles where the highest IEMG occurred at the lowest frequency band (2–12 Hz), for any muscle, were considered too noisy from movement artifact and were removed. When all pedal cycles of a lap were deemed too noisy, the cyclist’s data were removed from the study. Consequently, data from three of the participants were eliminated because more than one muscle was missing a complete lap of data (six cyclists were analyzed: age = 36.3 ± 3.0 yr; mass = 72.4 ± 2.0 kg; height = 1.78 ± 0.18 m; distance cycled per year = 9600 ± 2015 km; mean ± SEM).

The participants cycled on their own racing bicycles equipped with SRM PowerControl and SRM PowerMeter crank arms (Schoberer Rad Meßtechnik (SRM), Jülich, Germany) that measured cycling data including power output, cadence, and speed as well as HR through a Polar T31 transmitter (Polar Electro Oy, Kempele, Finland). A Global Positioning System with a barometric altimeter (GPSMAP 60CSx; Garmin International, Inc., KS) was fastened to the cyclists to determine location, speed, and elevation profile throughout the test. The speeds recorded from the Global Positioning System and SRM units were used to synchronize the location, elevation, and cycling data. To synchronize the EMG and cycling data, cadence was measured using both the SRM and a magnetic pedal switch, through the 16-bit analog-to-digital converter. The participants performed a 20-min warm-up on their own bicycles mounted on a stationary cycle trainer (Cycleforce Swing; Tacx, Technische Industrie Tacx, Wassenaar, The Netherlands) before completing the time trial. They also used their own clipless pedals and were instructed to maintain a consistent position, seated with hands on the drop bars, and pedal continuously for the duration of the trial.

Back to Top | Article Outline

Data analysis

Mechanical power output was normalized to the mean of each participant because of the intersubject variability in the measured values. The IEMG (time-varying intensity) calculated for each muscle was interpolated into 100 evenly spaced points for each pedal cycle such that the first point occurred when the pedal was at the top dead center of the pedal rotation (TDC). The interpolated IEMG for each muscle were normalized to its mean IEMG for the trial for each participant. For each pedal cycle, the total IEMG (Itot) was calculated as the sum of the IEMG for all muscles. Using the same IEMG techniques and the same 10 leg muscles, a significant positive relationship has been established between metabolic power and Itot (r = 0.86 ([29]). Therefore, Itot was used to estimate ηO as the ratio of mechanical power output to the metabolic power on a pedal cycle–by–pedal cycle basis.

Dominant muscle coordination patterns were identified for each pedal cycle using principal component (PC) analysis on the IEMG from all muscles (30). The data from all subjects were arranged into a P × N matrix A, where P = 1000 (10 muscles per activation pattern × 100 time points per pedal cycle) and N = the total number of pedal cycles. The covariance matrix B calculated from A was used to determine the eigenvectors and eigenvalues of B. The eigenvectors represent the PC weightings (IPC,W) of each PC, and the corresponding eigenvalues denote how much of the signal is characterized by the IPC,W where the largest absolute eigenvalues correspond to the most primary PCs. The product of matrix A and the transpose of the weighting matrix produced the PC loading scores (IPC,LS) for the N pedal cycles.

Back to Top | Article Outline

Statistics

With PC1 representing the dominant coordination pattern for all trails and all participants, the contribution of each IPC,LS relative to IPC1,LS for all pedal cycles was used in the analysis by normalizing each IPC,LS to IPC1,LS (ÎPC,LS). This implies that for a ratio of one there was an equal amount of a particular IPC,LS to IPC1,LS, although the IPC,W may represent only a small percentage of the entire EMG signal. The coefficients for muscle coordination were statistically compared with the Itot, mechanical power output, cadence, speed, HR, gradient, ηO, and distance using multivariate ANCOVA. Itot, power output, cadence, gradient, ηO, and distance were tested individually as the dependent variable with subject as a random factor and the first 20 IPC,LS values (IPC1,LS and ÎPC,LS for all other PCs) as covariates using multivariate ANCOVA. Cadence was included as a covariate in each statistical analysis except where it was the dependent variable. Pairwise Pearson correlation coefficients were also determined for all factors and also included the total EMG intensity per muscle per pedal cycle. Statistical tests were considered significant at P ≤ 0.05, and values are presented as mean ± SEM.

Back to Top | Article Outline

EMG reconstructions

The sum of the vector products of the IPC,W and the IPC,LS (∑IPC,WIPC,LS) reconstructs the instantaneous activation patterns for each pedal cycle. To visualize the effect of muscle coordination on each mechanical factor (power output, Itot, or ηO), the muscle coordination patterns were reconstructed using the first 20 PCs. If the IPC,LS had no significant effect on the mechanical factor, then the mean IPC,LS from all pedal cycles was used in the reconstruction. If the IPC,LS had a significant effect on the mechanical factor, then the pedal cycles were ranked by that factor, the top and bottom sets of 200 pedal cycles were extracted, and the mean IPC,LS from each set was used in the reconstruction. The reconstructed patterns therefore highlight the primary features of muscle coordination for the highest and lowest of each mechanical factor. In addition, for ηO, the muscle coordination patterns were reconstructed in a similar way for both the 200 pedal cycles with the highest normalized power outputs (100% group) and the 200 pedal cycles around the mean normalized power output (50% group, using 100 pedal cycles above and 100 pedal cycles below the mean).

Back to Top | Article Outline

RESULTS

The mean cadence, power output, and HR were 92.7 ± 0.5 rpm, 311.2 ± 2.5 W, and 171.6 ± 0.4 bpm, respectively, and the mean power outputs for each lap were 326.6 ± 2.3, 311.0 ± 1.3, 300.9 ± 1.2, and 305.4 ± 1.6 W for laps 1, 2, 3, and 4, respectively. The wind speed was less than 5 km·h−1 for all measurements, and the mean temperature was 21.7°C ± 0.7°C. The participants each traveled 18,683.4 ± 28.7 m during the time trial depending on the line of travel on the road.

Back to Top | Article Outline

Muscle activation

The first 20 PCs explained approximately 87% of the EMG signal with the first PC explaining more than 52%. The first PC can be visualized through the IPC1,W (Fig. 2) and was highlighted by heightened peak activation of VL and RF. IPC2,W differentiated between activation in VL and RF and activation in all other muscles, most notably MG and LG, and IPC3,W uncoupled VL and RF (Fig. 2). Both IPC1,LS and ÎPC2,LS were significantly correlated with power output and Itot (r = 0.69 and 0.97 for IPC1,LS and r = 0.51 and 0.68 for ÎPC2,LS, respectively; Fig. 3), and all three knee extensor muscles monitored (VL, VM, and RF) and Sol were significantly positively correlated with both IPC1,LS and ÎPC2,LS (r = 0.77, 0.78, 0.83, and 0.72 for IPC1,LS and r = 0.69, 0.58, 0.68, and 0.50 for ÎPC2,LS, respectively).

FIGURE 2

FIGURE 2

FIGURE 3

FIGURE 3

Mean total EMG intensity per pedal cycle and timing differences for the individual muscles were observed between laps. There was a general decrease from lap 1 to lap 4 in the mean total EMG intensity per pedal cycle for ST, BF, VL, RF, and TA (Fig. 4); in contrast, there was a general increase for Sol and GM. The mean coordination patterns for each lap were similar in timing but differed in amplitude with lap 1 demonstrating emphasized VL activation, laps 2 and 3 showing increased peak MG and LG activation, and lap 4 having elevated peak RF, VM, Sol, MG, and LG activity (Fig. 5). A shift in timing for most muscles was revealed in IPC4,W and ÎPC4,LS and was significantly correlated with HR (r = 0.60).

FIGURE 4

FIGURE 4

FIGURE 5

FIGURE 5

Back to Top | Article Outline

Mechanical power output

There was a significant correlation between power output and Itot (r = 0.74; Fig. 3), and eight of the first 20 IPC,LS were significantly related to power output. The high–power output pedal cycles used in the reconstructed signal had a large contribution of the VL and RF relative to the other muscles, whereas the lower power output cycles had a larger relative contribution of the MG and LG (Fig. 6). In addition, aside from an overall increase in Itot, the reconstructed patterns for the highest power outputs revealed more IEMG before and at TDC for TA, VM, RF, and VL and later activation of Sol, ST, BF, and GM at the bottom and first parts of the upstroke when compared with the lowest power outputs (Fig. 6). At high power outputs, the peak activation of most muscles occurred within three synchronized groups: the knee extensors; followed by Sol, GM, and BF; and, finally, MG, LG, and ST. In contrast, the low power outputs showed variation in the timing of peak activation within these three groups of muscles (Fig. 6). Of the high–power output pedal cycles, 106 occurred between the start and 1000 m, and 65 per 1000 m occurred between 4000 m and the start–finish line (because the distance between 4000 m and the finish line was less than 1000 m, the value per 1000 m has been used for comparison), whereas only 46 occurred between 1000 and 4000 m (15.3 per 1000 m; Fig. 1).

FIGURE 6

FIGURE 6

Back to Top | Article Outline

Gradient

Gradient demonstrated significant positive and negative relationships with power output and cadence, respectively (Fig. 3), and seven of the first 20 IPC,LS were significantly related to gradient. Reconstructed EMG traces at high gradients showed increased GM activity during the downstroke, earlier peak activation of BF and VM, later and increased peak activation of RF, and decreased peak activation of TA and ST. Also, there was a more even distribution of activation between the knee extensors, ankle extensors, and GM for high gradients, whereas lower gradients were more dependent on MG and LG. The mean normalized power outputs for high and low gradients were 1.260 ± 0.014 and 1.13 ± 0.04 with nonnormalized power outputs of 369.5 ± 4.1 W and 330.5 ± 11.2 W, respectively.

Back to Top | Article Outline

ηO

There was a significant negative correlation between ηO and Itot (r = −0.69; Fig. 3), a significant positive correlation between Itot and power output (r = 0.74; Fig. 3), and no significant relationship between ηO and power output (Fig. 3). Twelve of the first 20 IPC,LS showed significant relationships to ηO, and ηO had significant negative relationships with IPC1,LS and ÎPC2,LS (r = −0.65 and −0.45, respectively; Fig. 3). Because IPC2,W separated RF and VL from the other muscles monitored, at high ηO, there was less RF and VL relative to the other muscles, which was reversed for low ηO pedal cycles. Changes to RF, VL, and TA across TDC, GM, and Sol at 25% of the pedal cycle and ST and BF during the downstroke and across the bottom of the pedal cycle were the primary features of the PCs showing significant relationships to ηO. Reconstructed signals for high and low ηO showed large amounts of IEMG for RF and VL relative to the other muscles for low ηO (Fig. 6). The emphasis was shifted from RF and VL to MG and LG for high ηO (Fig. 6).

Examination of the reconstructed coordination patterns for high and low ηO at similar power outputs revealed power output–dependent differences. The relative contribution of RF and VL to increased Itot (ηO = mechanical power output/Itot) was reduced at 50% compared with 100% power output. At 100% power output, there were large spikes of IEMG for RF, VL, and TA relative to all other muscles for the low ηO (Fig. 6). Also at 100% power output, the activation of ST, MG, and LG was similar for high and low ηO, whereas all other muscles had higher IEMG amplitudes for low ηO. There was an even distribution of peak activation at 100% power output among most muscles for high ηO that was not as apparent for low ηO. At 50% power output, there was more muscle activity during the upstroke, particularly in TA, LG, RF, and VL for low ηO. The mean normalized power outputs for the high- and low-ηO groups were 1.548 ± 0.003 and 1.544 ± 0.003, respectively, for 50% power output and 1.922 ± 0.015 and 1.942 ± 0.019, respectively, for 100% power output. The pedal cycles for both high and low ηO occurred at similar locations on the course for both 50% and 100% power outputs.

Back to Top | Article Outline

DISCUSSION

The main findings of this study provide evidence that muscle coordination, power output, and ηO are dependent on the distribution of power and the terrain profile in outdoor cycling time trials. Also, ηO depends on the coordination of multiple muscles, particularly synchronized activation of muscles acting across the same joint and those active at the top and bottom of the pedal cycle, and not just the primary power producers during an outdoor time trial. Similarly, high power output is dependent on coordinated recruitment of muscles acting across the same joint and elevated activation of RF and VL.

Back to Top | Article Outline

Indoor and outdoor cycling

The mean cadence was similar to values found in male multistage cycling races (13). The SD was 7.9 rpm and is indicative of the variance in cadence used by different participants in variable terrain. With cadence normalized to each participant, there remained a significant negative relationship between cadence and gradient suggesting the deviation in cadence was in part due to variations in gradient. This is not surprising because cyclists use lower cadences in hilly and mountainous stages in male cycling races (13).

The timing and duration of muscle activation in both IPC1,W and the mean coordination pattern were similar but not identical with those found in indoor studies such as that by Wakeling and Horn (30) (Fig. 2). The closest cadence-matched condition used by Wakeling and Horn (30) was 100 rpm with a resistance equivalent to approximately 70 W, which was comparable in timing for most muscles except GM. This is understandable given that 70 W is considerably lower than the mean power output in this study and GM activity is dependent on resistance and not highly active at low power outputs (9). The closest power-matched condition had a resistance of approximately 250 W where GM activation was closer to our study, but the cadence was far lower at only 60 rpm. This resulted in earlier peak VM and VL activity, which was also found at higher gradients in this outdoor trial where cadences were lower. This provides further evidence that disparities in muscle coordination can be explained by the differences in cadence and resistance (12,30). These fluctuations in cadence and resistance resulting in altered power output and muscle coordination occur naturally on outdoor terrain but are more difficult to simulate indoors on a stationary bicycle.

Back to Top | Article Outline

Power output and muscle activity

Similar to indoor studies, increased power outputs or workloads were associated with increased levels of muscle activity as shown by the significant positive relationships between power output and both Itot and IPC1,LS (2,9,14,23). Although all muscles displayed increased IEMG in IPC1,W, RF and VL were the muscles most responsible for higher power outputs because these were the only muscles showing positive increases in ÎPC2,LS with rising power outputs.

Along with increased Itot, there was more synchronized activation of the muscles acting across the same joint at high power outputs. During the downstroke in cycling, peak joint moments occur in sequence from knee to hip to ankle (25), and a similar progression occurred in this study for muscle activation. The knee extensors were active synchronously followed by GM, BF, and Sol and then MG, LG, and ST (Fig. 6). The Sol is known to be more active at higher resistances than MG and LG (30) and at a similar location in the pedal cycle to GM to stabilize the ankle to transfer power to the pedal (16,21). Also, ST, MG, and LG are biarticulate muscles that can transfer power between the joints (25); therefore, they may have acted to transfer the joint moments from the hip to the ankle.

Back to Top | Article Outline

Muscle coordination, power output, and ηO depend on pacing and gradient

The significant relationships between the PCs, power output, ηO, and gradient show that muscle coordination, power output, and ηO fluctuate during an outdoor time trial as a result of the gradient. Similar to male cycling races, there was an increasing relationship between power output and gradient (17,26). Visually, this can be seen in Figure 1, where the majority of the highest power output pedal cycles occurred on uphill sections of the course. The lower cadences prevalent at higher gradients allow for earlier onset of power production at the start of the pedal cycle and longer duration of activation. This can be seen through the reconstructed coordination patterns for high power output, which showed increased TA, VM, RF, and VL at TDC and early in the pedal cycle and more Sol, ST, BF, and GM activity later in the downstroke (Fig. 6).

VL and RF followed the same trend in activation as the pacing strategy used in the outdoor time trial with both the mean power output per lap and VL and RF IEMG resembling a reverse J-shaped pacing strategy (decreasing from lap 1 to lap 3 and lap 4 higher than lap 3; Fig. 4 ([1]). This is similar to previous findings showing significant increases in VL activity throughout a 40-km time trial with negative pacing (2), no change in RF during a 30-min time trial with even pacing (7), and decreased RF through a 100-km time trial with positive pacing (23). The 40-km time trial showed no change in RF activity, which is partly explained by the environmental conditions (2). Lower cadences and increased RF at TDC would not be found in indoor studies using a constant cadence such as the 40-km time trial because these outcomes were found on higher gradients in this outdoor study.

Of the pedal cycles with the highest ηO, the largest concentrations were located on the slight downhill sections and the transitions from uphill to downhill with few occurring on the longest uphill section at the end of each lap (Fig. 1). In contrast, the largest concentrations of pedal cycles with the lowest ηO were located in the transitions from downhill to uphill and on the uphill sections (Fig. 1). This implies that at high intensities of outdoor time trial cycling, ηO is maximized where the resistance due to gradient is decreasing and minimized where it is increasing.

Back to Top | Article Outline

Fluctuations in muscle activity mitigate muscle fatigue

The fluctuations in muscle activity that resulted from the course profile may mitigate muscle fatigue that has previously been reported from more controlled indoor cycling trials (11,20) because there were no significant relationships between the muscle coordination patterns and distance. An effect of distance (or time) on muscle coordination would be expected if fatigue had progressed over the duration of the time trial. Similar to other indoor time trial studies, typical fatigue indices, such as increased muscle activation, were not observed for most muscles in this study (2,7). In addition, the muscles known to be susceptible to fatigue such as MG and LG (3,5) had stable IEMG for each lap (no significant difference in mean lap IEMG except MG lap 1, which was significantly higher; Fig. 4). Duc et al. (7) suggested that time trials do not induce significant quadriceps fatigue in competitive cyclists, which was supported in our study. The ability of the cyclists to self-regulate the pacing was a commonality between the indoor time trial studies by Bini et al. (2) and Duc et al. (7) and this study, which was not applicable to greater controlled indoor trials displaying muscle fatigue (11,20). Cyclists in the indoor time trials of Bini et al. (2) and Duc et al. (7) regulated their pacing by adjusting the resistance, whereas fluctuating terrain also influenced resistance in the present study. The shifting activation between each muscle may have provided adequate rest to avoid performance-reducing fatigue and maintain power output for the duration of the time trial as postulated by St Clair Gibson et al. (23). Whether dictated by terrain or self-adjusted, this implies that variable resistance diminishes the potential for fatigue during a time trial.

Back to Top | Article Outline

ηO and mechanical power output

High ηO occurred at lower than maximal power outputs and may depend on specific muscle activation timing around the top and bottom of the pedal cycle and activation in more than just the knee extensor muscles. Some of the key changes in TA, RF, VL, ST, and BF IEMG related to ηO occurred at the transitions between the downstroke and upstroke (across top and bottom of the pedal cycle) when little force is applied perpendicular to the crank arms (18,19,22). Coordinated recruitment between the muscles of the left and right legs could be the most important factor because of the mechanical link between the crank arms. Disruption of the pedal cycle would result if the pedals were not coordinated where the forces are minimal. This is less of a problem where forces are highest during the downstroke because some cyclists exhibit relatively small resistive forces during the upstroke while power output remains stable (18,19,22). Regardless of the specific explanations, these transitions seem to play a key role in ηO despite minimal forces acting on the pedals.

The reconstructed signals for high and low ηO for all pedal cycles were dependent on the interplay between RF and VL activity and MG and LG activity, as shown in the reconstructed coordination patterns (Fig. 6) and the negative relationship between ηO and ÎPC2,LS (Fig. 3). The amount of vastii activation differed considerably depending on the resistance despite showing very consistent activation previously (21). Ericson et al. (9) evaluated only VM at different power outputs showing consistency with minimal change in muscle activity. VM may be consistently active regardless of the power output, whereas VL fluctuates with power and is therefore more involved at the higher power outputs associated with time trial intensities.

Despite no significant relationship between power output and ηO, the power outputs for high and low ηO were substantially different: 0.96 ± 0.02 and 1.21 ± 0.03 normalized power output, respectively. Examination of ηO at 50% and 100% of the maximum power output was used to minimize the influence of power output on the reconstructed coordination patterns because muscle coordination changes with power output. At 100% power output, high and low ηO were associated with altered RF and VL activity, whereas at 50%, changes in ηO were due to other muscles (Fig. 6). Activation of many muscles to produce high power outputs was more efficient than relying solely on the knee extensors because high ηO for 100% power output displayed an even distribution of peak activation among most muscles. When power outputs were not as high at 50% power, ηO was dependent on minimizing muscle activation during the upstroke.

Back to Top | Article Outline

Methodological considerations

There are many difficulties when conducting muscle activation studies in an outdoor environment as compared with an indoor laboratory setting. Most notable are variables such as equipment, weather, and terrain. To record EMG signals continuously requires a reliable portable power source, large amounts of portable data storage, and consistent contact between the electrodes and the skin. Unfortunately, because of adhesion problems, normally detected in real time in the laboratory, data from three of the participants had to be excluded from the analysis. To minimize the effects of weather, the participants were all tested at the same time of day because both wind resistance (4) and temperature (24) are factors known to affect cycling performance. Wind and temperature were similar for each participant with wind speeds less than 5 km·h−1 and a mean temperature of 21.7°C ± 0.7°C.

Dorel et al. (6) showed increased GM activity in an aerodynamic position compared with riding with hands on the drop bars. Cyclists were instructed to ride with their hands on the drop bars but could have adopted a more aerodynamic position by reducing the hip joint angle. Future outdoor studies should include joint angles to help control for the influence of altered body position on muscle coordination.

The estimate of ηO relies on the relationship between total EMG intensity and energy consumption. The time trial was over 18 km with mean power output and HR of 311.2 ± 2.5 W and 171.6 ± 0.4 bpm, respectively. On the basis of previous time trial studies, it is reasonable to assume the participants cycled some of the time trial above a respiration quotient of one using both aerobic and anaerobic energy sources. Participants in a 30-min time trial cycled at approximately 74% of peak power output (276.0 ± 30.6 W, mean ± SD) with mean HR of 172.8 ± 6.7 bpm (7), and participants in a 40-km time trial cycled above and below a respiration quotient of one (2). Wakeling et al. (29) found a significant increasing relationship between metabolic power and total EMG intensity. In that study, metabolic power was underestimated at the highest workloads because it was calculated using only aerobic energy sources. When considering all workloads except 90% V˙O2max (25%, 40%, 55%, 60%, and 75% V˙O2max), there was a significant linear relationship between metabolic power and total EMG intensity (r2 = 0.79, correlation r = 0.89). This indicates that using total EMG intensity as a proxy for metabolic power at workloads similar to those found in indoor time trial studies (2,7) is acceptable. Therefore, the current study provides useful information about relative ηO despite the significant role of anaerobic energy sources used during the cycling time trial when efficiency was derived using only aerobic sources.

Back to Top | Article Outline

CONCLUSIONS

This study does not compare indoor with outdoor cycling directly but provides evidence that muscle activity is dependent on the measurement conditions. Muscle coordination, power output, Itot, and ηO fluctuate during an outdoor time trial due in part to the pacing strategy and changing terrain. In addition, high power outputs were associated with elevated levels of muscle activity, yet the highest ηO were not found at the highest power outputs. The muscles acting across the top and bottom of the pedal cycle and an even distribution of peak muscle activity seemed to be main determinants to ηO. In practical terms, this may be an indication that an even pacing strategy would be superior for similar time trials to avoid large changes to any one muscle. This study also presents evidence that fluctuating coordination patterns could provide rest to the primary power-producing muscles to avoid performance reductions from fatigue. This indicates that periodic adjustments to factors that affect muscle activation such as upper body position, fore–aft positioning on the saddle, and standing to pedal may be good strategies to help mitigate fatigue during a time trial.

Finally, because muscle coordination is dependent on both resistance and cadence, which change with terrain in outdoor cycling, deductions from laboratory studies should be cautious in interpretations outside the bounds of their specific conditions. Therefore, this study highlights the importance of measuring in the field or at least careful reproduction of outdoor environments in indoor studies.

The authors thank Karen Forsman for her help with data collection, Jasper Blake for the use of the SRM system, and the Natural Sciences and Engineering Research Council of Canada for financial support to J.M.W.

There are no conflicts of interest with any of the authors of this article.

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

Back to Top | Article Outline

REFERENCES

1. Abbiss CR, Laursen PB. Describing and understanding pacing strategies during athletic competition. Sports Med. 2008; 38 (3): 239–52.
2. Bini RR, Carpes FP, Diefenthaeler F, Mota CB, Guimaraes ACS. Physiological and electromyographic responses during 40-km cycling time trial: relationship to muscle coordination and performance. J Sci Med Sport. 2008; 11 (4): 363–70.
3. Bini RR, Diefenthaeler F, Mota CB. Fatigue effects on the coordinative pattern during cycling: kinetics and kinematics evaluation. J Electromyogr Kinesiol. 2010; 20 (1): 102–7.
4. Davies CTM. Effect of air resistance on the metabolic cost and performance of cycling. Eur J Appl Physiol Occup Physiol. 1980; 45 (2–3): 245–54.
5. Dingwell JB, Joubert JE, Diefenthaeler F, Trinity JD. Changes in muscle activity and kinematics of highly trained cyclists during fatigue. IEEE Trans Biomed Eng. 2008; 55 (11): 2666–74.
6. Dorel S, Couturier A, Hug F. Influence of different racing positions on mechanical and electromyographic patterns during pedalling. Scand J Med Sci Sports. 2009; 19 (1): 44–54.
7. Duc S, Betik AC, Grappe F. EMG activity does not change during a time trial in competitive cyclists. Int J Sports Med. 2005; 26 (2): 145–50.
8. Ericson MO. On the biomechanics of cycling: a study of joint and muscle load during exercise on the bicycle ergometer. Scand J Rehabil Med Suppl. 1986; 16: 1–43.
9. Ericson MO, Nisell R, Arborelius UP, Ekholm J. Muscular activity during ergometer cycling. Scand J Rehabil Med. 1985; 17 (2): 53–61.
10. Hettinga FJ, De Koning JJ, Broersen FT, Van Geffen P, Foster C. Pacing strategy and the occurrence of fatigue in 4000-m cycling time trials. Med Sci Sports Exerc. 2006; 38 (8): 1484–91.
11. Housh TJ, Perry SR, Bull AJ, et al.. Mechanomyographic and electromyographic responses during submaximal cycle ergometry. Eur J Appl Physiol. 2000; 83 (4–5): 381–7.
12. Hug F, Dorel S. Electromyographic analysis of pedaling: a review. J Electromyogr Kinesiol. 2009; 19 (2): 182–98.
13. Lucia A, Hoyos J, Chicharro JL. Preferred pedalling cadence in professional cycling. Med Sci Sports Exerc. 2001; 33 (8): 1361–6.
14. Macdonald JH, Farina D, Marcora SM. Response of electromyographic variables during incremental and fatiguing cycling. Med Sci Sports Exerc. 2008; 40 (2): 335–44.
15. Mujika I, Padilla S. Physiological and performance characteristics of male professional road cyclists. Sports Med. 2001; 31 (7): 479–87.
16. Neptune RR, Kautz SA, Zajac FE. Muscle contributions to specific biomechanical functions do not change in forward versus backward pedaling. J Biomech. 2000; 33 (2): 155–64.
17. Padilla S, Mujika I, Orbañanos J, Santisteban J, Angulo F, José Goiriena J. Exercise intensity and load during mass-start stage races in professional road cycling. Med Sci Sports Exerc. 2001; 33 (5): 796–802.
18. Patterson RP, Moreno MI. Bicycle pedalling forces as a function of pedalling rate and power output. Med Sci Sports Exerc. 1990; 22 (4): 512–6.
19. Patterson RP, Pearson JL, Fisher SV. The influence of flywheel weight and pedalling frequency on the biomechanics and physiological responses to bicycle exercise. Ergonomics. 1983; 26 (7): 659–68.
20. Petrofsky JS. Frequency and amplitude analysis of the EMG during exercise on the bicycle ergometer. Eur J Appl Physiol Occup Physiol. 1979; 41 (1): 1–15.
21. Ryan MM, Gregor RJ. EMG profiles of lower extremity muscles during cycling at constant workload and cadence. J Electromyogr Kinesiol. 1992; 2 (2): 69–80.
22. 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. 1991; 9 (2): 191–203.
23. St Clair Gibson A, Schabort EJ, Noakes TD. Reduced neuromuscular activity and force generation during prolonged cycling. Am J Physiol Regul Integr Comp Physiol. 2001; 281 (1): R187–96.
24. Tatterson AJ, Hahn AG, Martin DT, Febbraio MA. Effects of heat stress on physiological responses and exercise performance in elite cyclists. J Sci Med Sport. 2000; 3 (2): 186–93.
25. Van Ingen Schenau GJ, Boots PJM, de Groot G, Snackers RJ, van Woensel WWLM. The constrained control of force and position in multi-joint movements. Neuroscience. 1992; 46 (1): 197–207.
26. Vogt S, Heinrich L, Schumacher YO, et al.. Power output during stage racing in professional road cycling. Med Sci Sports Exerc. 2006; 38 (1): 147–51.
27. von Tscharner V. Intensity analysis in time–frequency space of surface myoelectric signals by wavelets of specified resolution. J Electromyogr Kinesiol. 2000; 10 (6): 433–45.
28. Wakeling JM, Blake OM, Chan HK. Muscle coordination is key to the power output and mechanical efficiency of limb movements. J Exp Biol. 2010; 213 (3): 487–92.
29. Wakeling JM, Blake OM, Wong I, Rana M, Lee SSM. Movement mechanics as a determinate of muscle structure, recruitment and coordination. Philos Trans R Soc Lond B Biol Sci. 2011; 366 (1570): 1554–64.
30. Wakeling JM, Horn T. Neuromechanics of muscle synergies during cycling. J Neurophysiol. 2009; 101 (2): 843–54.
Keywords:

EMG; POWER OUTPUT; EFFICIENCY; GRADIENT

©2012The American College of Sports Medicine