The propulsive power that is required for regular cycling is being generated entirely by leg propulsion, whereas in many other locomotive sports, such as rowing, cross-country skiing and Nordic walking, arm propulsion also provides an essential part of the total propulsive power. If arm propulsion would be excluded, the propulsive power would decrease substantially in these sports, which would directly affect the performance (8, 11, 17, 22). Previous remarks give rise to the question whether this mechanism could be extended to cycling; would adding arm propulsion to regular cycling increase the propulsive power and therefore the cycling performance?
Several studies (4, 9, 18, 21) investigated the influence of additional arm propulsion on the maximal power output (Pmax), which is suggested as the best indicator of cycling performance (8, 11, 17, 22). Pmax can be restricted by the cycling efficiency (CE), which reflects the amount of oxygen (in millilitres) required to perform a certain workload (8, 16), as well as central (i.e. cardiorespiratory) and peripheral limitations. Central restrictions refer to limitations of the cardiorespiratory system, which is not able to provide sufficient oxygenated blood to continue a certain level of power generation by the active muscles (2, 20). In case of peripheral restrictions, the cardiorespiratory system is able to provide more oxygenated blood than the active muscles require, the muscles itself reach its limit of power generation due to lactate accumulation or inadequate mitochondrial oxygen extraction from the blood (2, 20). Hence, the effects of additional arm propulsion on the Pmax are dependent on central factors such as maximal oxygen uptake (VO2max), maximal heart frequency (Hfmax) and pulmonary ventilation (VE), as well as blood lactate accumulation and mitochondrial oxygen extraction as peripheral factors (3, 4, 15, 18). Previous studies (4, 9, 18, 21) that compared regular cycling with cycling with combined arm and leg propulsion (CCALP) in a maximal cycling test reported a higher Pmax in CCALP, ranging from 12.3 to 18.0 percent. Additional to this, three studies (9, 18, 21) also reported a higher VO2max in CCALP, which indicates a higher capacity of the cardiorespiratory system in CCALP compared to regular cycling. None of the studies that reported an increased VO2max (9, 18, 21) found differences in Hfmax, moreover Nagle et al. (18) reported an increased VE in CCALP. This factor indicates an increased exchange of air between the atmosphere and the lungs in CCALP compared to regular cycling, which is a precondition to increase VO2max (2, 23). It can be suggested that the increased VE is at least partly responsible for the increase in VO2max, although latter increase can also be due to an increased stroke volume (2). The findings in previous studies (9, 18, 21) indicate that the subjects can utilize a higher part of their cardiorespiratory capacity in CCALP than in regular cycling, which resulted in a higher Pmax. This finding suggests that the limitations in Pmax during regular cycling are merely peripheral (20), hence Pmax can possibly be increased by adding muscle mass to regular cycling by means of additional arm propulsion.
Additional arm propulsion in cycling could be useful in situations in which extra propulsive power is required, for instance in cycling uphill or with headwind, or in a sprint final. Additional arm propulsion could also support leg propulsion in people with impairment of the lower extremity, as their leg propulsive power can be insufficient for cycling (7). Furthermore, commuters could use CCALP as a faster and less strenuous method of transportation than regular cycling, and CCALP could be a way to train the entire body on one device.
Previous studies that investigated the effects of CCALP (4, 9, 18, 21) deployed arm propulsion in a rotating movement. This method causes sway on the front wheel, hence impedes the applicability of the bicycle in daily practice. Recently, a prototype of an arm and leg propelled bicycle, named â€˜Powerbikeâ€™, was developed that deploys additional arm propulsion by a pushing movement of the arms, which should avoid sway of the front wheel and therefore secure the stability of the bicycle during cycling. Another feature of the Powerbike is that the arm propulsion can be used optionally, which permits cyclists to generate arm propulsion only in strenuous circumstances. The optional use of the arm propulsion mechanism allows cyclists to use the regular handlebars in curves and in hectic situations.
Although previous studies (4, 9, 18, 21) showed beneficial effects of CCALP compared to regular cycling, the use of another arm propulsion system in current study requires additional research on the specific effects of CCALP with the Powerbike on cycling performance. Besides of the use of other equipment, none of the previous studies (4, 9, 18, 21) investigated the CE of CCALP, although this is an important parameter for the cycling performance (8, 16). Therefore the primary aim of this study is to investigate differences in Pmax (Watt), time till exhaustion (TTE) and CE (work/VO2 in joules/ml O2) between regular cycling and CCALP with the Powerbike. In addition, the current study also aims to reveal whether the limitations in Pmax in regular cycling are the result of either central or peripheral restrictions. Hence, the secondary aims of this study are to investigate differences in VO2max (ml/kg/min.), Hfmax (BPM), and muscle activity (% maximal voluntary contraction) between regular cycling and CCALP with the Powerbike. Lastly, previous studies (4, 9, 18, 21) merely focused on the effects of additional arm propulsion at maximal intensity, although the largest sections of cycling stages are performed at sub-maximal intensities (14, 17). Therefore, the current study will focus on the effects of CCALP at maximal as well as sub-maximal intensity. According to previous studies (4, 9, 18, 21), it is hypothesized that the Pmax, as well as maximal and sub-maximal VO2 and Hf will be higher in CCALP than in regular cycling, CE and EMG muscle activity of the major power generating muscles of the legs are expected to be lower at sub-maximal intensities.â€ƒ
This randomized controlled cross-over trial was carried out according to the Declaration of Helsinki and was approved by the Medical Ethical Committee azM/UM, Maastricht, The Netherlands, and the Board of Directors of academically hospital Maastricht (azM). The study consisted of three visits for each subject. During the first visit, subjects were screened for eligibility based on general medical and physical testing. Subsequently, the included subjects performed a familiarization trial on the Powerbike. This trial was performed at varying resistance and lasted as long as the subject needed to become familiar with the concept of CCALP, the duration of the familiarization trial varied between five and twenty minutes. During both the second and third visit, a maximal cycling test was performed, either a regular cycling test or a test for CCALP. The second and third visit were separated by one week to ensure an identical situation, the sequence of both tests was randomly determined (www.randomizer.org). To prevent any bias due to nutritional status, subjects were asked to standardize their nutritional intake twenty-four hours prior to both cycling tests, which was documented by the researchers retrospectively.
After written informed consent, eighteen healthy, young subjects volunteered to participate in this study, which subsequently passed medical and physical screening. The total group of eighteen subjects consisted of eleven men and seven women with a mean age of 24.8 years (Â± 3.0), mean length of 1.74 metres (Â± 0.07) and a mean weight of 70.7 kilograms (Â± 9.3). All subjects had a healthy Body Mass Index (19 < BMI < 25), performed moderate to vigorous physical activity at least two hours per week, and none of them had a history of cardiovascular or respiratory disease, or related impairments that could influence the cycling performance.â€ƒ
Both cycling tests were performed on the Powerbike, which is a regular race bicycle that was equipped with an arm propulsion mechanism in front of the regular handle-bar, from the cyclistâ€™s point of view, which is depicted in Figure 1. The arm propulsion mechanism could be used optionally on top of the regular leg propulsion by alternately pushing the right and left lever down. The generated power was transferred onto the regular crank by means of cables, and integrated with the regular leg propulsion on the back wheel. In case the cyclists did not use the arm propulsion mechanism, both levers were fixed in place. During the cycling tests, the Powerbike was mounted in a Tacx cycling ergometer and connected to the Tacx Trainer software 4 on a laptop (i-Genius T2020, Tacx, Wassenaar, The Netherlands). The Tacx enabled a stable position of the subject on the bicycle, the possibility to inflict a certain regimen to the subject, and measured the power output with 1 Hz. Muscle activity was measured with surface electromyogram (EMG) at 1926 Hz (TrignoTM Wireless EMG, Delsys, Natick, United States), the EMG sensors were placed according to SENIAM guidelines (www.seniam.org) by using 4-slot Adhesive Skin Interfaces (Delsys, Natick, United States). Oxygen consumption was measured with 0.2 Hz by indirect calorimetry (Omnical, Instrument Development Engineering & Evaluation (IDEE), Maastricht, The Netherlands). Heart frequency was measured with 1 Hz by a heart rate monitor (heart rate belt smart T1994, Tacx, Wassenaar, The Netherlands) connected to the Tacx Trainer software 4.
To study possible differences in muscle activity between regular cycling and CCALP, electromyogram (EMG) of the major power generating muscles of the arms â€“ triceps brachialis and deltoideus pars clavicularis â€“ and legs â€“ soleus, gastrocnemius lateralis, gluteus maximus, vastus medialis and lateralis, and rectus femoris â€“ were recorded during both maximal cycling tests. To normalize the EMG signals, five MVC tests were performed, as described in â€˜The ABC of EMGâ€™ (13). Resistance to elicit an MVC was exerted manually by the researchers. The MVC tests were followed by a ten minute warm-up with regular cycling or CCALP at a resistance of 100 Watt, synchronized with the subsequent test of that day. Both maximal cycling tests were performed at the same incremental regimen, starting at 120 Watt and increasing every minute with 20 Watt. The subjects were instructed to cycle at a cadence of 80 rotations per minute (RPM), the test was terminated if the cadence dropped below 70 RPM. During the entire test, power output, VO2, Hf and muscle activity were measured.
To render results comparable between CCALP and regular cycling, sub-maximal data of the CE, VO2, Hf and EMG muscle activity were normalized to the TTE of the regular cycling test; TTE during the regular cycling test was considered 100 percent. Subsequently, sub-maximal data were analyzed and averaged over all steps of ten percent, which resulted in one value per step of ten percent of TTE, for each variable. As the inflicted resistance level of the Tacx fluctuated during the first step of ten percent, this step was left out of the results. The fluctuations disappeared after the first step.
The raw EMG signals were judged graphically prior to any analysis, recordings that contained signals that could not originate from the tests were excluded from analysis, based on â€˜The ABC of EMGâ€™ (13): e.g. some recordings showed conspicuously decreasing activity of the leg muscles during progress of the tests, which was probably caused by inadequate placement of the sensors, and other recordings showed abundant noise complexes, which were probably caused by interference of power hum (13). Subsequently, the raw EMG files were analyzed with a tailored MATLAB program (MathWorks, Natick, United States). The EMG signals were filtered with a Butterworth high pass filter of 20 Hz and a Butterworth low pass filter of 150 Hz, to minimize noise. Root Mean Square was applied to the filtered data. Subsequently, MVCs of all included muscles were determined by the highest peak of two MVC tests per muscle, this value was assumed as 100 percent of the MVC. All EMG muscle activity that exceeded baseline activity plus two standard deviations (SDs) (10, 13), in that particular recording, was considered as muscle activity due to the cycling test. Subsequently, this activity was averaged per step of ten percent of TTE and divided by the MVC, which led to a percentage of MVC as muscle activity per step. These normalized values could be compared between both maximal cycling tests (10). Sub-maximal VO2, Hf and CE were averaged per step of ten percent of TTE, hence these could be compared between both cycling tests. CE was calculated by the ratio of power output to VO2.
The sample size calculation was performed using the G*power 188.8.131.52 software. The calculated sample size was eighteen, which was based on two studies (4, 24) that found an average Pmax of 288 Watt in a regular cycling test with a similar population and testing protocol, and a ten percent higher Pmax in CCALP than in regular cycling. The significance level was set at 0.05 and the power at 0.90.
All statistical analyses for significance were performed in SPSS 20 (IBM, Chicago, United States). Normality of the obtained data was determined by both visual analysis of the Q-Q plots and histograms in SPSS, and Shapiro-Wilk tests. This revealed normal distribution of the differences in Pmax, VO2max, Hfmax and TTE between both tests, as well as sub-maximal VO2 and Hf, and non-normal distribution of sub-maximal CE and EMG muscle activity. Significance of the differences in Pmax, TTE, as well as maximal and sub-maximal VO2 and Hf were determined through a Paired Samples T-test, supplemented with a Repeated Measures ANOVA for sub-maximal VO2 and Hf. The significance of the differences in sub-maximal EMG muscle activity and CE were determined through a Wilcoxon Signed-Rank Test. Alpha was set at 0.05 for all maximal parameters, and at 0.006 for all sub-maximal analyses, based on Bonferroni-correction. Hence, a type I error could be prevented.â€ƒ
There was no significant difference for mean Pmax between regular cycling, 350.2 Watt (Â± 58.9), and CCALP, 353.8 Watt (Â± 57.2) with p=0.506. Accordingly, mean TTE did not differ between both tests, with 610.9 seconds (Â± 145.0) in regular cycling and 612.6 seconds (Â± 140.1) in CCALP (p=0.826). Also mean VO2max did not significantly differ between both tests with 47.8 ml/kg/min. (Â± 7.0) for regular cycling and 49.0 ml/kg/min. (Â± 7.2) for CCALP (p=0.076), whereas mean Hfmax was significantly higher in CCALP, 189.9 BPM (Â± 11.3), than in regular cycling, 187.5 BPM (Â± 11.3) with p=0.020. Mean Pmax, TTE, VO2max and Hfmax during regular cycling and CCALP, and the differences between these variables are summarized in Table 1.
Mean VO2 showed an incremental trend over both tests; the progress of the test (in time steps of ten percent) had a statistically significant effect on VO2 (F (8,136) = 162.848, p < 0.001). The type of cycling test (CCALP or regular cycling) also had a significant effect on VO2 (F(1,17) = 56.054, p < 0.001), VO2 was higher in CCALP from 20 to 90 percent of TTE than in regular cycling (p < 0.006), as depicted in Figure 2A. Mean Hf also showed an incremental trend over both tests (F (8,136) = 216.80, p < 0.001), and the type of cycling test had a significant effect on Hf (F(1,17) = 21.695, p < 0.001). Moreover, the progress of the test (in time steps of ten percent) and type of cycling test were significantly interacted on Hf (F(8,136) = 3,366, p = 0,049). Mean Hf was significantly higher in CCALP from 20 to 100 percent of TTE (p < 0.006), which is depicted in Figure 2B. Mean CE during CCALP was significantly lower for all steps of ten percent, except 100 percent of TTE (p < 0.006), as depicted in Figure 2C. Mean data of both cycling tests, the differences between the cycling tests, and the level of significance (based on Bonferroni-corrected alpha = 0.006) are summarized per step of ten percent in Table 1 in the supplemental materials.
The difference in mean EMG muscle activity of the triceps brachialis between CCALP and regular cycling increased during the progress of the tests, this difference was statistically significant for all steps of ten percent of TTE (p < 0.006). The difference in mean EMG muscle activity of the deltoideus pars clavicularis did not increase during progress of the tests, although the difference between both tests was statistically significant as from 40 percent of TTE (p < 0.006). Mean EMG muscle activity of the soleus, gastrocnemius lateralis, gluteus maximus, vastus medialis, rectus femoris and vastus lateralis did not significantly differ between CCALP and regular cycling, with the exception of the rectus femoris at 100 percent of TTE. The differences in mean EMG muscle activity of all muscles is depicted in Figure 3, and the mean EMG data of both cycling tests, the differences between the cycling tests, and the significance (based on Bonferroni-corrected alpha = 0.006) are summarized in Table 2 in the supplemental materials. Although lacking statistically significance, Figure 3 shows differences between both tests of up to ten percent. Figure 3 illustrates a decreasing positive difference in EMG muscle activity of the soleus and gastrocnemius lateralis, as from 50 percent of TTE, a decreasing positive difference of the gluteus maximus, and a decreasing negative difference in the vastus medialis, rectus femoris and vastus lateralis as from 50 percent of TTE.â€ƒ
The current study investigated the effects of CCALP with the Powerbike compared to regular cycling. Opposite to the expectations, Pmax did not differ between both tests. This finding is contradictory to three previous studies (9, 18, 21) that observed a 12.3 to 18.0 percent higher Pmax in CCALP with a rotating movement for arm propulsion than in regular cycling. Associated with the higher Pmax in these studies (9, 18, 21), they also found an accompanying 5.3 to 14.1 percent higher VO2max, whereas one study (4) found a similar VO2max in CCALP and regular cycling. Although the mean difference of 1.2 ml/kg/min. in VO2max between CCALP with the Powerbike and regular cycling in current study was not statistically significant, it exposed a tendency of a higher VO2max in CCALP, which was confirmed by the low p-value of 0.076. The increase in VO2max in current and previous studies (9, 18, 21) can be attributed to the larger active muscle mass in CCALP, as this increases the oxidative capacity of cyclists by the larger active mitochondrial mass (2). The ostensibly increase in VO2max indicates additional capacity of the cardiorespiratory system in CCALP, and hence peripheral restrictions as limitation in regular cycling, which are being avoided in CCALP by the additional muscle mass. Although an increased cardiorespiratory capacity does not directly improve the cycling performance, it is an important precondition of the Pmax (8, 19). Even though the ostensibly increased VO2max and increased Hfmax in CCALP in current study indicate additional cardiorespiratory capacity on top of that during regular cycling (20), the lack of an increase in Pmax questions whether the arm propulsion system of the Powerbike permits cyclists to employ this additional capacity to improve their Pmax, and hence their cycling performance.
The discrepancy between previous studies (4, 9, 18, 21) and current study about differences in VO2max, and indirectly Pmax, between CCALP and regular cycling could be the consequence of different subject populations. The study of Billat et al. (4) included well-trained triathletes. Trained triathletes, just like trained cyclists, are able to attain a higher percentage of their VO2max in regular cycling than non-trained cyclists (6). In this case, VO2max can increase only marginally by adding arm propulsion to regular cycling. In contrast, non-trained cyclists, who have a more equal division of strength and power over the upper and lower extremity, can increase VO2max substantially by adding arm propulsion (6, 21). This finding was supported by Secher et al. (21), who included well-trained swimmers, kajakers, rowers, cyclists, runners and non-trained students in their study. For all subjects, the VO2max was higher in CCALP than in regular cycling, except for the well-trained cyclists (21). Current study included a general population of healthy, young subjects, although two of them were recreational road cyclists, and one was a professional road cyclist for over ten years. One of the recreational road cyclists as well as the former professional road cyclist had a lower VO2max in CCALP than in regular cycling, the other recreational road cyclist had a similar VO2max in both tests. Excluding these three cyclists from analysis resulted in a significant mean difference of 1.69 ml/kg/min. in VO2max between CCALP and regular cycling (p = 0.023). Previous finding illustrates that inclusion of the three cyclists has negated the effects of CCALP on the VO2max.
Besides inter-individual differences in the effects of CCALP, the experience of subjects with CCALP with the Powerbike could also influence the effects on Pmax. Multiple subjects of the current study reported that despite of the familiarization trial, they were still not completely accustomed to the combined arm and leg movements on the Powerbike, which might have confined or even negated the effects of CCALP on Pmax.
Another possible reason for the absence of a higher Pmax in CCALP is the reduced increase in power generation of the leg muscles compared to regular cycling, as from the moment that the power generation of the arms increases. The current study cannot statistically support this assumption, although visual analysis of Figure 3 shows substantial differences in EMG muscle activity between CCALP and regular cycling. The higher EMG muscle activity of the triceps brachialis and deltoideus pars clavicularis in CCALP can be explained by the activity of the arms in CCALP, whereas the arms are rather inactive in regular cycling. The trend of a reduced increase in EMG muscle activity of the legs in CCALP could be related to the considerable increase in activity of the arms as from 50 percent of TTE (1). The ostensibly reduced increase in EMG muscle activity of the leg muscles might be the reason for the absence of a higher Pmax in CCALP compared to regular cycling, whereas VO2max was ostensibly higher, since the arms utilize more oxygen than the legs to generate a similar power output (5). Although the reduced increase in EMG muscle activity of the legs might impeded an increase in Pmax, it could have delayed or even prevented peripheral exhaustion of the leg muscles, despite of increased central exhaustion as shown by the increased VO2max and Hfmax. The decreased peripheral exhaustion can be adopted by employing the legs more intensively at sub-maximal intensities or by spreading out de total power generating capacity of the leg muscles over a longer period.
To render results comparable at all intensities, the use of arm propulsion was obliged during the entire test for CCALP in current study, although the Powerbike permits subjects to use arm propulsion optionally on top of leg propulsion. However, the lower CE during all steps of CCALP indicates that subjects are performing the same workload at a higher VO2, hence they are utilizing more energy for an equal Pmax. (5, 8, 16). This finding indicates that additional arm propulsion does not yield beneficial effects at sub-maximal intensities, it can even be disadvantageous in prolonged cycling stages, in which energy depletion is suggested to be a crucial performance indicator (12). Furthermore, most of the subjects reported that it is hard to use additional arm propulsion effectively at a low resistance, as the legs are very dominant in this situation and do not need any support of the arms. Combining this remark with the lower CE at all sub-maximal intensities suggests that it could be more effective to use the additional arm propulsion only at higher resistances. In this manner, premature exhaustion of the arm muscles and inefficient arm propulsion can be prevented in early stages of the cycling test for CCALP.
Although the current study does not indicate an increased cycling performance in CCALP with the Powerbike compared to regular cycling, the increased Hfmax and near-significant increase in VO2max imply a higher cardiorespiratory capacity in CCALP than in regular cycling, which is an important physiological precondition for an increase in Pmax. More research is required to investigate whether the increased cardiorespiratory capacity can be converted into a higher Pmax, and hence an improved cycling performance. As current study did not identify any increase in Pmax, future research can be directed to the effectiveness of the arm propulsion mechanism of the Powerbike in converting additional cardiorespiratory capacity in CCALP into external power output, which could be done by comparing CCALP with the Powerbike, with CCALP with a rotating movement for arm propulsion, as previous studies found (9, 18, 21) increases in Pmax as wel as VO2max with these devices. As the lower CE in CCALP in current study clearly showed disadvantageous effects of CCALP with the Powerbike at sub-maximal intensities in non-specific trained subjects, it is suggested for future research to leave the additional arm propulsion optionally during the entire test for CCALP; subjects should decide themselves when they need the additional arm propulsion. Furthermore, it is suggested to investigate the effects of CCALP with the Powerbike in a larger sample size, which also enables the comparison between different populations. Future research can focus on the effects of CCALP in untrained subjects, subjects who are trained only in the lower extremity (e.g. cyclists) and subjects who are trained in the upper as well as lower extremity (e.g. rowers or swimmers), as previous studies found dissimilar effects of CCALP in these groups (6, 21). Moreover, the influence of training programs specific to CCALP with the Powerbike, on the differences between CCALP and regular cycling can be investigated, as multiple subjects in current study reported that they were not completely accustomed to the combined arm and leg movements on the Powerbike. Lastly, it would be interesting to measure EMG muscle activity of the trunk muscles as these appear to be active in ensuring the upright position of the cyclist as well as in the movement of the upper body from the left to the right lever, and vice versa.
As Pmax was the primary outcome of current study, the sample size calculation was based on this parameter. This, however, does not guarantee an adequate power of the secondary parameters, which can argue the lack of statistically significant differences in VO2max and EMG muscle activity, although Table 1 and Figure 3 respectively, demonstrate differences in these variables between CCALP and regular cycling. Furthermore, it should be noticed that some EMG recordings were excluded from analysis as these contained incorrect signals. This measure could also have affected the power of the EMG analyses.
...(download the rest of the essay above)