Recumbent cycling is a popular and well-established mode of exercise that is used in fitness clubs and clinical rehabilitation centers to treat a diverse range of disorders (25,38). In addition, anecdotal evidence suggests that the number of outdoor recumbent cyclists has increased in recent years, and more and more competitive recumbent cyclist races (usually classified under the “human-powered races”) are apparent.
Outdoor recumbent bicycles offer important aerodynamic benefits compared with the more traditional upright bikes leading to improvements in cycling velocity and subsequent performance. However, cycling endurance and fatigue are also sensitive to the vertical distance between the heart and the active muscles involved in cycling because of the gravitational effect acting across these muscles. Gravity enhances the perfusion pressure on active muscles, which helps explain the faster rate of increase in blood flow (blood flow kinetics) (13,35), the faster oxygen uptake (V˙O2) kinetics (9,23,24,30), and the lower lactate production (15,33) during exercise when active muscles are below the level of the heart compared with when they are above heart level. These effects appear to contribute to a reduction in the rate of muscle fatigue (17) and endurance improvements in graded (12,15,18,24,27,30) as well as high-intensity constant-load (15–18) cycling during upright compared with supine postures under standard laboratory conditions (i.e., when the effect of air resistance is removed).
In contrast to the substantial amount of published studies comparing performance outcomes between supine and upright cycling postures, and despite the increasing popularity of recumbent bicycles, fewer studies have explored these physiological effects at recumbent body angles between supine and upright. For instance, to our knowledge, it is not known if graded cycling performance or muscle fatigue and EMG activities during submaximal cycling are different at body angles between upright and supine. Recently, Egaña et al. (16) reported that exercise times to failure and V˙O2 kinetics were not altered when high-intensity submaximal cycling was performed at progressively lowering body angles from the upright (90°) to 65° recumbent (R) and 30° R postures, but that exercise times to failure were significantly shorter in the supine (0°) when compared with the three inclined postures. However, in the same study, cardiac output (CO) responses (recorded only during upright, 65° R and 30° R postures) were significantly larger during the two recumbent postures when compared with upright at 30 s into the exercise, raising the possibility that when the heart lies above the level of the active exercising muscles (as is the case when cycling recumbent angles are at or above 30°), the gravity-induced reductions in perfusion pressure in these two recumbent compared with the upright posture may be counteracted by an increased central and/or peripheral circulation (16). However, most competitive as well as many recreational recumbent cyclists adopt recumbent angles lower than 30° R, often as low as 15° (where the heart lies below the level of most active muscles) to maximize their aerodynamic benefits, but little is known if these low recumbent angles can be detrimental for cycling performance when compared with upright when the effect of air resistance is removed.
Accordingly, to further explore the postural effect on cycling performance, the main aim of the present study was to compare the endurance, V˙O2 kinetics, muscle fatigue, and EMG (from three leg muscles) responses during high-intensity constant-load cycling between a previously unexplored low recumbent angle of 15° upright, supine, and 30° R postures. In addition, we also compared the exercise time sustained during a graded cycling test among these four postures. It was hypothesized that exercise performance would be reduced, fatigue increased, and V˙O2 kinetics impaired during supine and 15° R when compared with 30° R and upright postures.
Two experiments were performed. Experiment 1 tested the effect of lower recumbent angle on performance during graded cycling and on muscle fatigue, whereas experiment 2 tested the effect of lower recumbent angle on V˙O2 kinetics, CO, and performance during constant-load cycling exercise. Ten young male subjects (mean ± SD; age = 24 ± 4 yr, height = 183 ± 4 cm, body mass = 74.4 ± 6.9 kg) who were recreationally active once or twice per week took part in experiment 1, which required them to visit the Human Performance Laboratory in the Department of Physiology of Trinity College Dublin, on 6 d separated by at least 72 h. Nine different healthy males (mean ± SD; age = 21 ± 1 yr, height = 181 ± 4 cm, body mass = 77.5 ± 7.6 kg) of similar activity levels took part in experiment 2, and they visited the same laboratory on five separate occasions (at least 72 h apart). Informed consent was obtained from all subjects before their participation in this study. All experimental procedures were carried out in accordance with the Declaration of Helsinki and were approved by the Trinity College Dublin Faculty of Health Science Research Ethics Committee.
Before each testing day, subjects were asked to refrain from consuming caffeine and alcohol and to avoid any strenuous exercise in the 24 h before testing. For testing days 1 to 4 of experiment 1 (i.e., “graded tests,” see following sections) and all five testing days of experiment 2, a standard cycle ergometer (Monark 874E; Monark Exercise AB, Vansbro, Sweden) was used. For testing days 5 and 6 of experiment 1 (i.e., “fatigue tests,” see following sections), exercise was performed on an electrically braked cycle ergometer (Lode Excalibur Sport, Groningen, the Netherlands) controlled via a connected PC running Lode Wingate software (v1.0.12, Groningen, the Netherlands). These two cycle ergometers have been shown to reach comparable power outputs (37). Both ergometers were modified to allow for recumbent cycling by the attachment of an inclining bench to the rear of the ergometer (16). The recumbent seat was approximately 10 cm below the center of the crank, which is similar to the distance between the crank and the seat on most commercially available recumbent bikes. The cycling cadence required for each test was 60 rpm. Failure in any exercise test was defined as an inability to maintain a minimum cadence of 50 rpm for 3 s. During exercise, HR was continuously monitored and recorded every 5 s using an HR monitor (Polar Electro, S725, Kempele, Finland). In addition, apart from visits 5 and 6 of experiment 1, subjects wore a face mask to continuously collect expired air using an online metabolic system (experiment 1: Metalyser, Cortex Biophysik, Leipzig, Germany; experiment 2: Innocor, Innovision A/S, Odense, Denmark). The Metalyser system measured the expiratory airflow with a volume transducer (Triple V® turbine, digital) connected to the metabolic analyzer. Expired gases were analyzed for oxygen (O2) with an electrochemical cell and for carbon dioxide (CO2) output with an infrared analyzer. Before each test, the CO2 and the O2 analyzers of the Metalyser were calibrated against room air as well as a reference gas of known composition (5% CO2, 15% O2, and 80% N2), and the volume was calibrated with a 3-L gas syringe. The Innocor system measured airflow using a pressure difference pneumotach. Carbon dioxide analysis was performed by using a photoacoustic gas analyzer, and oxygen was analyzed using an oxygen sensor (Oxigraf Inc., Mountain View, CA) based on the principle of laser diode absorption spectroscopy. The volume was calibrated with a 3-L syringe, and only the oxygen sensor was regularly calibrated (against room air) before each tests by the researcher, as both the oxygen sensor and the photoacoustic gas analyzer require multipoint calibration performed by the manufacturer periodically (6–12 months). The analysis of expired air allowed the determination of O2 uptake (V˙O2), CO2 output (V˙CO2), minute ventilation (V˙E), and the RER every 10 s (experiment 1) or breath by breath (experiment 2).
During the upright tests, the position of the head and body were maintained in a vertical plane with the arms held loosely by the sides to minimize any involvement from the upper body associated with gripping of the handlebars. In all other postures (30° R, 15° R, and supine), subjects lay comfortably behind the ergometer with the recumbent seat angle set as required for each posture, again with the arms lying loosely at the side (16). A harness was worn in the supine and 15° R posture to secure the subject to the ergometer and to prevent any rearward movement while cycling. Knee angles were kept constant at the start of the crank cycle between postures by adjusting the distance from the seat to the pedals. The magnitude of the hydrostatic pressure acting on the active muscles during cycling was estimated for each posture using the equation, ρgh, where ρ is the density of the fluid (i.e., blood), g is the gravitational constant, and h is the vertical distance (or height) between the right atrium of the heart (estimated to lie at the level of the third costal space; approximately 5 cm below the sternal angle) and the arteries feeding the proximal muscles engaged in cycling (estimated to lie at the midpoint of the two lateral iliac crests). Other measures specific to each Experiment are described in the next section.
Experiment 1 (graded cycling performance and fatigue and EMG activities during constant-load cycling).
In experiment 1, the maximal performance during graded cycling and muscle fatigue and EMG activities of the lower limbs during constant-load cycling were assessed in the upright, 30° R, 15° R, and supine postures. After a familiarizing session, subjects attended the laboratory on six subsequent separate days. On testing days 1 to 4, a graded test to failure was performed each day at one of the four randomly assigned cycling postures. After a resting period of 3 min in the exercise position, the graded test began with 3 min cycling at 60 W and increased incrementally by 30 W every 3 min until the subject reached 180 W (all subjects reached at least this work rate). Thereafter, the work rate was increased by 15 W every minute until failure. The peak work rate achieved was defined as the highest work rate that a subject could sustain for a minimum period of 50 s. Peak V˙O2 was defined as the highest 30-s mean value recorded before the subject’s volitional termination of the test. The work rate at which the ventilatory threshold (VT) occurred was determined using the V-slope method by identifying the power output at which a clear steeper increase of V˙CO2 as compared with V˙O2 occurs (2,5).
On testing days 5 and 6, subjects performed on each day two high-intensity constant-load tests to failure with 10-s efforts of all-out cycling interspersed every minute. The two tests were separated by 60 min of passive rest (i.e., sitting on a chair), so that in total, between the 2 d, four fatigue tests were performed. The fatigue tests were performed at the same four body postures (upright, 30° R, 15° R, and supine), and the order of the tests was selected at a counterbalanced fashion to remove any potential exercise order effects. The intensity for the fatigue tests was set at the work rate achieved at midpoint between the VT and the end exercise during the graded test performed in the upright posture (50% Δ upright) so that these tests were performed at the same absolute intensities. Initially, to determine the maximal power achievable in each posture, subjects completed a single peak effort test comprising a 45-s cycling at 100 W followed by a 10-s all-out cycling and then a 3-min rest. The fatigue test commenced with cycling at each subject’s specified work rate (50% Δ upright) for 45 s, after which subjects completed 10 s of all-out cycling followed by 5 s of unloaded cycling. This sequence was repeated until failure. The 60-min resting period was chosen because the effect of prior heavy exercise has been shown not to persist beyond approximately 45–60 min (7), and we were careful to ensure that HR responses before the second fatigue test were returned back to the levels seen before the initial fatigue test. The peak power achieved before the test and during the subsequent all-out cycling efforts was recorded to enable an estimation of the rate of fatigue while cycling in each of the four postures. The decline in peak power during each all-out effort was described using a linear function (y = a + bx), where y is power, x is time, parameter a provides power at t = 0 (i.e., predicted peak power), and parameter b represents the rate of fatigue. Although fatigue is often a nonlinear function of time, most individuals exhibited a linear response (Fig. 2). On this basis, we chose to describe fatigue as a linear function of time. The goodness of fit of the linear function to the fatigue responses in this study (Fig. 2) is reflected in the SE in estimating the y-intercept of this function. Across the four conditions, the average SE of estimates for the y-intercept was 3.9% ± 0.9%. For all the 10-s efforts of all-out cycling, the chosen torque factor was 0.65. A previous pilot study established that this was the most comfortable torque factor to perform the 10-s all-out efforts in the recumbent postures. Individual braking torque was calculated by multiplying the body mass of each subject by the torque factor.
In addition, during the fatigue tests, surface EMG recordings were taken from three muscles of the right leg: vastus lateralis (VL), biceps femoris (BF), and gastrocnemius medialis (GM). The skin recording sites were selected from the belly of the muscle and prepared by shaving, abrading, and cleaning with alcohol (70%). Two bipolar Ag/AgCl recording electrodes were placed on the skin at the recording sites 25 mm apart (center to center) and in a plane estimated to be parallel to the direction of the muscle shortening during contraction. A reference electrode was attached to the anterior superior iliac crest. EMG signals were band-pass filtered (10–500 Hz) and sampled at 1000 Hz using a Power Lab connected to a PC running chart recording software (version 6.0; ADInstruments, Sydney, Australia). On completion of the first test, the electrode locations were carefully marked with a permanent marker. The root mean square (RMS) values were calculated burst by burst throughout the test. The criteria for the onset and offset activation were based on a voltage threshold (3 SDs above baseline). The average RMS value during the initial 30 s of unloaded exercise was subtracted from all RMS measures during subsequent exercise, and this latter value was initially normalized to the maximum RMS (NEMG) achieved within each posture. This normalization was followed because the rates of fatigue (see previous section) were also calculated from the peak power achieved within each posture. However, given that EMG responses are often normalized to a set of “gold standard” and assuming that the upright posture was the gold standard in the present study, EMG data were also normalized to the upright maximum RMS. The maximum RMS was calculated by averaging three consecutive bursts when the maximum power output was achieved during the initial 10-s effort of all-out cycling (i.e., prefatigue test). NEMG measurements were then calculated during the fatigue test and were based on five consecutive bursts (i.e., crank cycles) recorded between the 20th and the 25th second and between the 40th and the 45th second during the first minute and between the 40th and the 45th second (before the 10-s all-out efforts) every minute thereafter. ΔNEMG was calculated as the difference between end-exercise NEMG and NEMG at the onset of exercise. Because of technical difficulties, data from two subjects were excluded from the final analysis.
Experiment 2 (V˙O2 kinetics, CO, and performance during constant-load cycling).
In experiment 2, the performance, CO, and V˙O2 kinetics during high-intensity constant-load cycling were assessed during upright, 30° R, 15° R, and supine postures. After a familiarization session, subjects attended the Human Performance Laboratory on five subsequent days. Initially (testing day 1), subjects performed a maximal graded test to failure in the upright posture as described earlier (see experiment 1, days 1–4) to determine peak work rate and ventilatory threshold (2,5). Subsequently (testing days 2–5), two bouts of constant-load cycling were performed each day at an intensity equivalent to the work rate achieved at midpoint between the VT and the end exercise during the upright graded test (50% Δ, mean ± SD, 229 ± 37 W). On each of these days, subjects completed the exercise in one of the four randomly assigned postures (upright, 30° R, 15° R, or supine). Bout 1 continued until failure, followed by a minimum of 60 min of passive rest (i.e., sitting on a chair), whereas the duration of bout 2 was set at 6 min. As during the fatigue tests, a minimum resting period of 60 min was chosen because it has been shown that approximately 45 min is sufficient to allow the restoration of “normal” V˙O2 response after prior high-intensity exercise in young participants (7), and we were careful to allow sufficient recovery time for restoration of HR and V˙O2 values. In addition, on day 5, before performing the two constant-load exercise tests, CO (Innocor; Innovision A/S, Odense, Denmark) was measured using an inert gas rebreathing technique based on the Fick principle as previously described (1,22,26,34), at rest and after an additional short 60-s bout of the constant-load exercise (50% Δ upright) given that stroke volume (SV) has been shown to peak at approximately 30 s during high-intensity cycling exercise (20,33). This was performed randomly in each of the four postures (upright, 30° R, 15° R, and supine). These short bouts were separated by a 15-min rest period and were then followed by the two constant-load bouts described earlier. Stroke volume was calculated as CO/HR.
Breath-by-breath values for V˙O2 collected during the first 6 min of each bout were linearly interpolated to provide values at 1-s intervals. For each subject, all data sets for the two bouts performed in the same posture were then time aligned and averaged. Data were then smoothed using a 5-s moving average filter and analyzed by fitting a three-component exponential curve to the results according to the following equation:
where the three exponential terms represent the cardiodynamic, primary, and slow components of V˙O2. Baseline V˙O2 (base) represents the mean oxygen uptake during the last 90 s before the exercise bout. For each exponential term, A c, A p, and A s are the asymptotic amplitudes; τ c , τ p, and τ s are the time constants; and TDc, TDp, and TDs are the time delays. The parameters U c, U p, and U s are conditional expressions that limit the fitting of a particular phase to the period at and beyond the time delay associated with that phase. Fitting the cardiodynamic phase allowed us to visually determine the transition between the cardiodynamic and the primary phase given that when this transition is determined using a set value (i.e., 20 s), the time constant of the primary phase can be overestimated (36). However, the cardiodynamic phase cannot be always described by an exponential term (29), and thus only the amplitude and duration (TDp–TDc) of this phase (given that TDc was not fixed at t = 0) are presented. End-exercise V˙O2 was defined as the mean oxygen uptake during the last 30 s of exercise. The physiologically relevant amplitude of the primary component (A′p) was defined as the sum of A c + A p. Because the asymptotic value (A s) of the exponential term describing the V˙O2 slow component may represent a higher value than that actually reached at the end of the exercise, the actual amplitude of the V˙O2 slow component at the end of exercise (A′s) was estimated as the difference between the end-exercise V˙O2 and the A′p. The functional gain of the entire response (i.e., end-exercise gain) was calculated as (end-exercise V˙O2 − V˙O2 base / Δ work rate). The relative contributions of the slow component and relevant amplitude of the primary component to the overall increase in V˙O2 at end exercise were also calculated. In addition, the mean response time (MRT) of the “overall” V˙O2 kinetics response was calculated by fitting a single-exponential curve from the onset to the end of heavy-intensity exercise. Data that exceeded the 95% prediction intervals during an initial fit of a model were excluded, and no more than 13 data points were removed from the original time series of data. The models were fitted to the data using a weighted least-squares nonlinear regression procedure (TableCurve 2D; Systat San Jose, CA).
Kinetics parameters of oxygen uptake and “peak” responses among postures were analyzed using a one-way repeated-measures ANOVA. Effects of body posture and time on peak power and EMG activities were identified using a two-way repeated-measures ANOVA. Differences were located using Tukey’s post hoc test. The relationship between the rate of fatigue and the ΔNEMG activities were examined using Pearson’s product moment correlation coefficients. Data are presented as mean ± SD. Significance was set at P < 0.05.
Exercise times and physiological responses during the graded test at the four postures are shown in Figure 1A and Table 1A, respectively. The time sustained and the peak work rate achieved during the graded test were significantly lower in the supine compared with the other three postures, and they were also lower in the 15° R compared with the upright posture. The absolute work rate (W) at VT was significantly lower in the supine compared with the upright posture but was similar among postures when expressed in relative (% peak) terms. Peak HR was significantly higher in the upright compared with the other three postures. However, peak V˙O2 (mL·kg−1·min−1), peak RER, and V˙O2 (mL·kg−1·min−1 and % peak) at VT were not different among the four postures.
Mean cycling time (min) during the fatigue tests was significantly shorter in the supine (2.6 ± 0.6) compared with the other three postures (upright = 5.0 ± 1.3, 30° R = 4.7 ± 1.8, and 15° R = 4.4 ± 1.3). Figure 2 shows the individual fatigue responses to the four postures (Figs. 2A–2D) and the mean normalized fatigue responses (Fig. 2E). The rate of fatigue (W·min−1) during the supine posture was significantly greater (−113 ± 77) than that in the upright posture (−75 ± 40) but not different compared with 15° R and 30° R postures (−95 ± 52 and −95 ± 58, respectively). A comparison of the mean normalized fatigue responses (% peak) showed that at minute 2, fatigue was significantly larger in the supine (67 ± 14) compared with upright (82 ± 14) but not different than 30° R (78 ± 13) and 15° R (76 ± 9) postures. Power outputs at failure were similar among the four postures and the equivalent of 61% ± 15% (upright), 56% ± 13% (30° R), 56% ± 11% (15° R), and 60% ± 17% (supine) of the maximum power outputs (i.e., at time = 0). These maximum power outputs (W) (i.e., at time = 0) were not different among the four postures (upright = 742 ± 152, 30° R = 761 ± 99, 15° R = 792 ± 131, supine = 742 ± 170).
The maximum RMS responses for all three muscles obtained during the initial peak effort test (i.e., time = 0) were not different among the four postures. NEMG responses (% maximum) are shown in Figure 3. Each mean value in the graphs shown in Figure 3 is based on responses of all subjects (i.e., n = 8) so that the maximum exercise time for each NEMG response shown (before “failure” time point) is limited by the subject who failed first. NEMG responses for the GM showed a significant posture by time interaction so that these responses during the supine and 15° R postures at minutes 1 and 2 and at failure were greater than during the upright posture. These NEMG data for GM were also significantly larger at failure than at exercise onset, minutes 1 and 2 within the supine, and 15° R postures, but no time effect was observed for either 30° R or upright postures. NEMG responses for VL also showed a significant posture by time interaction so that they were significantly higher in the supine compared with the upright posture at failure, and in addition, responses for VL were significantly larger at failure compared with at the onset of exercise for all four postures. BF NEMG responses also tended to be larger during the supine and 15° R compared with upright (main effect, posture; P = 0.09) and showed a significant main effect of time within each of the four postures with no posture by time interaction. The ΔNEMG activities were not affected by body postures for any of the three muscles investigated and were not correlated to rates of fatigue in any posture.
Cycling times to failure (min) during the constant-load tests were significantly longer in the upright and 30° R postures compared with 15° R and supine postures (Fig. 1B and Table 1B). Peak HR were also higher during upright and 30° R compared with 15° R and supine postures, but peak V˙O2 and peak V˙ E were not affected by posture (Table 1B).
V˙O2 kinetics responses during high-intensity exercise for a representative individual at the four postures are presented in Figure 4. Two subjects failed to complete 6 min of exercise during the two exercise bouts in the supine posture (times to failure for these two subjects were 4 and 5 min; Figure 1B); thus, the V˙O2 responses for all four postures for these two subjects were limited to the shortest exercise bout. The absolute (L·min−1) and the relative (percentage of V˙O2 increase at end exercise) amplitudes of the slow component of the V˙O2 response were significantly larger during the supine compared with the 30° R and upright postures, and the absolute amplitude was also larger during the 15° R compared with the 30° R posture (Table 2A). Conversely, the relative contribution of the primary component to the overall increase in V˙O2 at end exercise was significantly larger in the upright and 30° R compared with supine, and it was also larger in the upright than 15° R posture. The MRT of the entire response was significantly longer in the supine compared with the upright and the 30° R postures. The rest of the parameters were not different among the four postures.
CO, SV, and HR responses (0–60 s of exercise).
At rest, CO and SV were significantly lower in the upright compared with the other three postures (Table 2B). However, by 60 s into the constant-load bout, these variables were similar among all conditions. HR responses did not significantly differ among postures at rest or at 60 s during exercise (Table 2B).
The perfusion pressure (mm Hg) added by gravity acting on both sides of the vasculature in the quadriceps muscles (at their upper site of origin, near the iliac crest) was significantly reduced (on average between experiments 1 and 2) from the upright (25.0 ± 0.6) to the 30° R (18.4 ± 0.4) and 15° R (14.2 ± 0.3). Perfusion pressure was also lower during 15° R than 30° R, as well as in the supine (∼0) when compared with the other three postures.
The main findings of the present study were that the performance of graded exercise was progressively diminished when the body was reclined from the upright to supine postures with significant reductions apparent between upright and 15° R and supine postures, fatigue was progressively diminished at body angles above zero with significantly larger rates of fatigue in the supine compared with upright posture, and the time to failure during high-intensity constant-load cycling was significantly longer during upright and 30° R compared with supine and 15° R postures. The shorter times sustained during the constant-load efforts at 15° R and supine postures were accompanied with larger amplitudes of the slow component of the V˙O2 response and lower relative contribution of the primary component to the overall increase of the V˙O2 response in these two postures.
Graded cycling performance.
Peak work rate during graded cycling exercise has been consistently shown to be significantly (∼10%) greater during upright compared with supine posture (12,14,15,24,27,30). However, before this study, very limited data were available on the effect of recumbent angles between supine and upright postures (i.e., degree of inclination between 0° and 90°) on graded cycling performance. Kato et al. (2011) recently reported that peak oxygen uptake during a graded cycling test was significantly higher during high recumbent (∼75° R) compared with supine posture, although no data regarding peak work rate or time sustained during the tests were provided (28). To our knowledge, only one previous study investigated the effect of the degree of body inclination from supine to upright postures on graded muscle performance and showed that the time sustained during isometric calf plantarflexion exercise was similar when the body tilt angle was lowered from 90° to 47°, but that performance was significantly higher in these two postures compared with supine (0°) posture (13). The present findings are consistent with these observations and extend them to demonstrate that the time sustained and the peak work rate achieved during a graded cycling exercise are not significantly reduced, reclining the body from upright to 30° R, but that they are significantly reduced during low recumbent (15° R) and supine compared with upright postures.
However, despite these systematic reductions in peak power and exercise time, peak oxygen uptake responses in the present study were not significantly different among the four postures, suggesting an “excess” V˙O2 in the 15° R and supine postures. It is possible that this extra V˙O2 requirement is related to a larger contribution of fast twitch fibers in the low recumbent postures because of the reduced perfusion pressure. Fast twitch fibers are predominantly activated above VT and are believed to show a less effective V˙O2 response (greater O2 cost of contraction) than slow twitch fibers, which are predominantly recruited at intensities below VT (19,31,32). Consistent with this notion, DiMenna et al. (2010) have recently reported that the ΔV˙O2/ΔW slope (calculated by linear regression) above VT during a ramp graded exercise was significantly greater during supine compared with upright cycling, whereas this slope was similar for both postures at intensities below VT (11). In this context, the significantly lower power outputs at VT observed in the present study in the supine compared with the upright posture in the absence of any differences in V˙O2 were unexpected. It could be speculated that the apparent larger oxygen cost during the supine compared with upright posture at the point of VT may be due to a larger activation of inspiratory muscles and/or postural accessory muscles to overcome gravity in the supine position, as some participants might have not laid their backs completely on the bench due to lack of familiarity with this form of exercise. And if these extra postural and respiratory challenges contributed to the excess V˙O2 at intensities below VT in the supine posture, then it is reasonable to think that they also contributed, at least in part, to the excess V˙O2 at end exercise in the low recumbent postures. The design of the graded test (i.e., combination of step and ramp increments) used in the present study was not appropriate to explore the ΔV˙O2/ΔW relationship given that the slope of this relationship appears to be sensitive to the exercise protocol (step versus ramp increments) (6); thus, further studies using either ramp or step graded tests are needed to better understand how cycling efficiency is affected at body angles between upright and supine postures.
Fatigue and EMG activities.
As for graded exercise performance, before this study, there were no data on the effect of recumbent cycling on muscle fatigue responses. The present findings showed that the rates of fatigue (and the times sustained during the fatigue tests) were significantly different between supine and upright postures. The larger fatigue responses during supine cycling were evident by the second minute of exercise (Fig. 2E), confirming that the postural effect of fatigue is relatively rapid and manifest within the first minute or two of exercise (13,14,17,21,41,42). It is possible that the considerable variations in the postural effect on muscle fatigue and times to failure among participants (Fig. 2A–2D) may have precluded these effects to be significant between low recumbent and upright postures. In the present study, the postural effect on fatigue occurred in the absence of any significant postural effect on maximum force or power output before the onset of exercise (i.e., time = 0), and this is consistent with previous studies (13,14,17,21,40,41). Similarly, the maximum EMG responses for the three muscles assessed during the 10-s all-out efforts before the fatigue tests (i.e., time = 0) were not different among the four postures, and the normalized EMG responses were also similar at the onset of exercise for all postures. However, there was a progressive divergence of the NEMG responses mainly for GM and at a lesser extend for VL such that NEMG values for the upright and 30° R remained relatively constant (or increased slightly), whereas they increased more dramatically for supine and 15° R postures. Even if NEMG values for BF were not significantly affected by posture, they also tended to be larger in the supine and 15° R postures compared with upright posture (main effect, posture; P = 0.09).
During intense exercise, such NEMG behavior is normally attributed to a compensatory increase in motor unit recruitment and/or rate coding in the presence of fatigue. This notion is supported by the present findings as differences in fatigue and NEMG responses between upright and supine postures were significant at the same exercise time points, and differences, although not significant, between the 15° R and the upright postures were also apparent at similar exercise time points. The present findings are in agreement with Egaña et al. (17), who showed significantly larger rates of fatigue in the supine compared with upright posture together with larger NEMG responses in five lower-body muscles by minutes 1–4 and at the point of failure of a high-intensity (80% peak power) constant-load cycling exercise, and with Tachi et al. (2004), who reported significantly shorter times during a moderate intensity exhaustive dorsiflexion exercise and larger integrated EMG responses of the tibialis anterior at the end of the exercise in the supine compare with upright posture (41). In contrast, Denis and Perry (10) reported similar EMG activities in VL, rectus femoris, and BF muscles during upright and supine high-intensity cycling; however, these observations are likely to be related to the fact that they used same relative intensities (posture-specific VT plus 25 W), as research has previously shown that relative high-intensity constant-load exercise (posture specific 80% peak work rate) is similar between upright and supine cycling (15). However, caution needs to be taken interpreting the findings of the present study given that (a) fatigue indices (i.e., rate of fatigue) and ΔNEMG data were not significantly correlated and (b) when EMG (RMS) responses were normalized relative to the maximum upright condition (results not presented here), NEMG responses for GM were no longer significantly different across postures. Nevertheless, NEMG responses for VL and BF were not affected by the two different normalization protocols, and normalizing EMG responses to the upright condition when some of the remaining conditions are performed on different days is likely to be less appropriate than normalizing EMG data to the posture-specific peak RMS recorded on the same day.
Performance, CO, and V˙O2 kinetics during constant-load cycling.
Given the possibility that the dynamic response of V˙O2 may be linked to the postural effect on muscle fatigue and EMG, in experiment 2, we explored V˙O2 kinetics responses at the same four cycling postures using the same exercise intensity (work rate relative to 50% Δ of the upright graded test). In addition, to assess cycling endurance at this high-intensity constant-load effort, the first of the two bouts was brought to failure. Cycling endurance during a high-intensity constant-load effort is significantly larger in upright compared with supine cycling (15–18), and this effect is of a larger magnitude (≥100%) than the effect observed during graded cycling (∼10%) (12,15,18,24,27,30). In addition, when the body angle is lowered from the upright to 65° R and 30° R postures, the time sustained during a high-intensity cycling constant-load exercise and the V˙O2 kinetics responses are unaffected, but the time sustained is larger at these three postures (upright, 65° R, and 30° R) compared with supine (16). Consistent with Egaña et al. (16) the present findings showed that exercising times (and the parameters describing the V˙O2 kinetics response) were similar between upright and 30° R postures, but we observed that times to failure were significantly diminished when the body angle was further reduced to 15° R and 0° (supine) postures compared with both upright and 30° R postures.
In addition, these shorter exercising times at 15° R and supine postures were accompanied by larger amplitudes of the slow component and smaller relative amplitudes of the primary component of the V˙O2 kinetics response measured during the initial 6 min of exercise. In addition, the MRT was significantly longer during the supine posture compared with both upright and 30° R postures. This notion is consistent with observations of lower amplitudes of the primary component but larger amplitudes of the slow component of V˙O2 (30), as well as an overall slower kinetic response of the V˙O2 response (12,24,33) in the supine compared with upright high-intensity cycling. It has been suggested that the V˙O2 slow component is, at least in part, caused by increased motor unit activation as high-intensity exercise proceeds with increased recruitment of fast-twitch fibers that show longer time constant and greater oxygen cost of contraction than slow-twitch fibers (3,4,19,31,32,39). Therefore, the present EMG findings from the fatigue tests are consistent with a greater slow component amplitude during the supine and the 15° R conditions.
In the present study, the perfusion pressure acting on the active musculature was significantly reduced lowering the body from the upright to the 30° R, 15° R, and supine postures by causing a shift of blood volume toward the heart, which, in turn, augmented resting CO and SV, as shown elsewhere (16,33). However, during dynamic exercise, perfusion pressure is removed from the venous side upon muscle relaxation when the venous volume has been expelled during the preceding muscle contraction. This results in a greater increase in perfusion pressure in the upright compared with the recumbent and supine postures where the gravity-induced hydrostatic pressure component is lower or absent. As a consequence, by minute 1 of the heavy constant-load exercise, the upright CO values were similar to the recumbent and supine values, suggesting that central circulatory adaptation may not be responsible for the differences observed in performance. This is consistent with Leyk et al. (33), who reported similar CO responses between supine and upright postures at the onset of heavy intensity exercise. Further studies exploring the dynamic response of blood flow among these postures are needed to confirm the likely role of the peripheral circulation.
In the present study, participants in experiment 1 (days 5 and 6) completed two fatigue tests within the same session (separated by 60 min), and participants in experiment 2 completed the constant-load submaximal test to failure and the 6-min submaximal bout also within the same session (separated by 60 min). Although there is evidence that a period of 60 min after high-intensity exercise is sufficient for complete recovery to occur (i.e., to have no effect on V˙O2 kinetics during subsequent exercise ), we cannot exclude the possibility that the completion of the initial fatigue test had an effect on the physiological responses to the second fatigue test. To minimize this limitation, the order of the fatigue tests was selected at a counterbalanced fashion to remove any potential exercise order effects. For practical reasons, we did not normalize the work rate relative to each participant’s critical power (CP) during the constant-load tests to failure during experiment 2, and we cannot exclude the possibility that some participants could have exercised below their CP and others above. However, when intensities approximately 5% below CP were used during upright cycling, Burnley et al. (8) reported that 9 of their 11 participants were able to complete at least 30 min of cycling, whereas approximately 5% above CP exhaustion occurred within 13 min. Thus, given that the mean time to exhaustion in the upright posture in the present study was <13 min and that eight of the nine subjects reached exhaustion in less than 20 min, it is likely that most (if not all) subjects in the present study exercised above their CP.
In conclusion, this study showed that the time sustained during both graded cycling exercise as well as high-intensity constant-load cycling exercise is significantly reduced by supine and low recumbent (15° R) postures compared with 30° R and upright postures. The reductions in performance during the constant-load exercise during supine and 15° R postures were, in turn, accompanied with larger amplitudes of the V˙O2 slow component and larger EMG activities of some of the active muscles involved in cycling. Future studies should explore the potential ergogenic effects achieved due to reduced air resistance during low-recumbent compared with upright cycling at different degrees of air resistance to better understand under which environmental wind conditions the aerodynamic benefits associated with the 15° R posture would overcome the gravity-induced negative physiological repercussions.
No external funding was received for this research.
The authors report no conflict of interest.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Agostoni P, Cattadori G, Apostolo A, et al.. Noninvasive measurement of cardiac output during exercise by inert gas rebreathing technique: a new tool for heart failure evaluation. J Am Coll Cardiol
. 2005; 46 (9): 1779–81.
2. Amann M, Subudhi A, Foster C. Influence of testing protocol on ventilatory thresholds and cycling performance. Med Sci Sports Exerc
. 2004; 36 (4): 613–22.
3. Barstow TJ, Jones AM, Nguyen PH, Casaburi R. Influence of muscle fiber type and pedal frequency on oxygen uptake kinetics of heavy exercise. J Appl Physiol
. 1996; 81 (4): 1642–50.
4. Barstow TJ, Mole PA. Linear and nonlinear characteristics of oxygen uptake kinetics during heavy exercise. J Appl Physiol
. 1991; 71 (6): 2099–106.
5. Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol
. 1986; 60 (6): 2020–7.
6. Boone J, Koppo K, Barstow TJ, Bouckaert J. Effect of exercise protocol on deoxy[Hb + Mb]: incremental step versus ramp exercise. Med Sci Sports Exerc
. 2010; 42 (5): 935–42.
7. Burnley M, Doust JH, Jones AM. Time required for the restoration of normal heavy exercise VO2
kinetics following prior heavy exercise. J Appl Physiol
. 2006; 101 (5): 1320–7.
8. Burnley M, Doust JH, Vanhatalo A. A 3-min all-out test to determine peak oxygen uptake and the maximal steady state. Med Sci Sports Exerc
. 2006; 38 (11): 1995–2003.
9. Convertino VA, Goldwater DJ, Sandler H. Oxygen uptake kinetics of constant-load work: upright vs. supine exercise. Aviat Space Environ Med
. 1984; 55 (6): 501–6.
10. Denis R, Perry S. Influence of posture on pulmonary O2
uptake kinetics, muscle deoxygenation and myolectrical activity during heavy-intensity exercise. J Sports Sci Med
. 2006; 5: 254–65.
11. DiMenna FJ, Bailey SJ, Jones AM. Influence of body position on muscle deoxy[Hb+Mb] during ramp cycle exercise. Respir Physiol Neurobiol
. 2010; 173 (2): 138–45.
12. DiMenna FJ, Wilkerson DP, Burnley M, Bailey SJ, Jones AM. Priming exercise speeds pulmonary O2
uptake kinetics during supine “work-to-work” high-intensity cycle exercise. J Appl Physiol
. 2010; 108 (2): 283–92.
13. Egaña M, Green S. Effect of body tilt on calf muscle performance and blood flow in humans. J Appl Physiol
. 2005; 98 (6): 2249–58.
14. Egaña M, Green S. Intensity-dependent effect of body tilt angle on calf muscle fatigue in humans. Eur J Appl Physiol
. 2007; 99 (1): 1–9.
15. Egaña M, Green S, Garrigan EJ, Warmington S. Effect of posture on high-intensity constant-load cycling performance in men and women. Eur J Appl Physiol
. 2006; 96 (1): 1–9.
16. Egaña M, O’Riordan D, Warmington SA. Exercise performance and VO2
kinetics during upright and recumbent high-intensity cycling exercise. Eur J Appl Physiol
. 2010; 110 (1): 39–47.
17. Egaña M, Ryan K, Warmington SA, Green S. Effect of body tilt angle on fatigue and EMG activities in lower limbs during cycling. Eur J Appl Physiol
. 2010; 108 (4): 649–56.
18. Egaña M, Smith S, Green S. Revisiting the effect of posture on high-intensity constant-load cycling performance in men and women. Eur J Appl Physiol
. 2007; 99 (5): 495–501.
19. Endo MY, Kobayakawa M, Kinugasa R, et al.. Thigh muscle activation distribution and pulmonary VO2
kinetics during moderate, heavy, and very heavy intensity cycling exercise in humans. Am J Physiol Regul Integr Comp Physiol
. 2007; 293 (2): R812–20.
20. Faisal A, Beavers KR, Robertson AD, Hughson RL. Prior moderate and heavy exercise accelerate oxygen uptake and cardiac output kinetics in endurance athletes. J Appl Physiol
. 2009; 106 (5): 1553–63.
21. Fitzpatrick R, Taylor JL, McCloskey DI. Effects of arterial perfusion pressure on force production in working human hand muscles. J Physiol
. 1996; 495: (Pt 3): 885–91.
22. Fontana P, Boutellier U, Toigo M. Non-invasive haemodynamic assessments using Innocor during standard graded exercise tests. Eur J Appl Physiol
. 2010; 108 (3): 573–80.
23. Hughson RL, Shoemaker JK, Tschakovsky ME, Kowalchuk JM. Dependence of muscle VO2
on blood flow dynamics at onset of forearm exercise. J Appl Physiol
. 1996; 81 (4): 1619–26.
24. Hughson RL, Xing HC, Borkhoff C, Butler GC. Kinetics of ventilation and gas exchange during supine and upright cycle exercise. Eur J Appl Physiol Occup Physiol
. 1991; 63 (3-4): 300–7.
25. Jacobs PL, Nash MS. Exercise recommendations for individuals with spinal cord injury. Sports Med
. 2004; 34 (11): 727–51.
26. Jakovljevic DG, Nunan D, Donovan G, Hodges LD, Sandercock GR, Brodie DA. Comparison of cardiac output determined by different rebreathing methods at rest and at peak exercise. Eur J Appl Physiol
. 2008; 102 (5): 593–9.
27. Jones AM, Berger NJ, Wilkerson DP, Roberts CL. Effects of “priming” exercise on pulmonary O2
uptake and muscle deoxygenation kinetics during heavy-intensity cycle exercise in the supine and upright positions. J Appl Physiol
. 2006; 101 (5): 1432–41.
28. Kato M, Tsutsumi T, Yamaguchi T, Kurakane S, Chang H. Characteristics of maximum performance of pedaling exercise in recumbent and supine positions. J Sports Sci Med
. 2011; 10 (3): 491–7.
29. Koga S, Shiojiri T, Kondo N. Measuring VO2
kinetics. The practicalities. In: Jones AM, Poole DC, editors. Oxygen Uptake Kinetics in Sport, Exercise and Medicine
. London: Routledge; 2005. pp. 39–61.
30. Koga S, Shiojiri T, Shibasaki M, Kondo N, Fukuba Y, Barstow TJ. Kinetics of oxygen uptake during supine and upright heavy exercise. J Appl Physiol
. 1999; 87 (1): 253–60.
31. Krustrup P, Secher NH, Relu MU, Hellsten Y, Soderlund K, Bangsbo J. Neuromuscular blockade of slow twitch muscle fibres elevates muscle oxygen uptake and energy turnover during submaximal exercise in humans. J Physiol
. 2008; 586 (Pt 24): 6037–48.
32. Krustrup P, Soderlund K, Mohr M, Bangsbo J. Slow-twitch fiber glycogen depletion elevates moderate-exercise fast-twitch fiber activity and O2
uptake. Med Sci Sports Exerc
. 2004; 36 (6): 973–82.
33. Leyk D, Essfeld D, Hoffmann U, Wunderlich HG, Baum K, Stegemann J. Postural effect on cardiac output, oxygen uptake and lactate during cycle exercise of varying intensity. Eur J Appl Physiol Occup Physiol
. 1994; 68 (1): 30–5.
34. Mac Ananey O, Malone J, Warmington S, O’Shea D, Green S, Egaña M. Cardiac output is not related to the slowed O2
uptake kinetics in type 2 diabetes. Med Sci Sports Exerc
. 2011; 43 (6): 935–42.
35. MacDonald MJ, Shoemaker JK, Tschakovsky ME, Hughson RL. Alveolar oxygen uptake and femoral artery blood flow dynamics in upright and supine leg exercise in humans. J Appl Physiol
. 1998; 85 (5): 1622–8.
36. Mezzani A, Grassi B, Giordano A, Corra U, Colombo S, Giannuzzi P. Age-related prolongation of phase I of VO2
on-kinetics in healthy humans. Am J Physiol Regul Integr Comp Physiol
. 2010; 299 (3): R968–76.
37. Micklewright D, Alkhatib A, Beneke R. Mechanically versus electro-magnetically braked cycle ergometer: performance and energy cost of the Wingate Anaerobic Test. Eur J Appl Physiol
. 2006; 96 (6): 748–51.
38. Perell KL, Gregor S, Kim G, et al.. Comparison of cycling kinetics during recumbent bicycling in subjects with and without diabetes. J Rehabil Res Dev
. 2002; 39 (1): 13–20.
39. Pringle JS, Doust JH, Carter H, et al.. Oxygen uptake kinetics during moderate, heavy and severe intensity “submaximal” exercise in humans: the influence of muscle fibre type and capillarisation. Eur J Appl Physiol
. 2003; 89 (3–4): 289–300.
40. Reiser RF 2nd, Maines JM, Eisenmann JC, Wilkinson JG. Standing and seated Wingate protocols in human cycling. A comparison of standard parameters. Eur J Appl Physiol
. 2002; 88 (1–2): 152–7.
41. Tachi M, Kouzaki M, Kanehisa H, Fukunaga T. The influence of circulatory difference on muscle oxygenation and fatigue during intermittent static dorsiflexion. Eur J Appl Physiol
. 2004; 91 (5–6): 682–8.
42. Zange J, Beisteiner M, Muller K, Shushakov V, Maassen N. Energy metabolism in intensively exercising calf muscle under a simulated orthostasis. Pflugers Arch
. 2008; 455 (6): 1153–63.
Keywords:©2013The American College of Sports Medicine
RECUMBENT; CYCLING; V˙O2 KINETICS; FATIGUE; POSTURE