- Email: [email protected]
The effects of exercise intensity or drafting during swimming on subsequent cycling performance in triathletes
Journal of Science and Medicine in Sport (2007) 10, 234—243
ORIGINAL PAPER
The effects of exercise intensity or drafting during swimming on subsequent cycling performance in triathletes David J. Bentley a,b,∗, Sebastien Libicz c, Aurelie Jougla c, Olivier Coste c, Jerome Manetta c, Karim Chamari e, Gregoire P. Millet c,d a
Health and Exercise, School of Medical Science, University of NSW, Sydney, Australia Department of Human and Health Science, University of Westminster, London, United Kingdom c Faculty of Sport Sciences, University of Montpellier, Montpellier, France d Aspire, Academy for Sports Excellence, Doha, Qatar e National Centre of Medicine and Science in Sport (CNMSS), Tunis, Tunisia b
Received 19 January 2006 ; received in revised form 1 May 2006; accepted 1 May 2006 KEYWORDS Ergometer; Triathlon; Athletes; Power; Metabolism; Time trial
∗
Summary The purpose of this study was to compare the affects of drafting or a reduction of exercise intensity during swimming on the power output sustained (Pmean ) during a subsequent cycle time trial (TT). In addition the relationship between peak power output (PPO) and Pmean generated during the cycle TT after swimming was examined. Nine well-trained triathletes performed an incremental cycling test to exhaustion for determination of PPO. In addition, each subject performed three swim-cycle (SC) trials consisting of 20 min cycle TT preceded by a 400 m swimming trial completed as (1) ‘‘all out’’ and in a non-drafting situation (SC100% ); (2) at 90% of SC100% in a non-drafting situation (SC90% ); (3) in a drafting position at the same controlled velocity as SC100% (SCdrafting ). Swimming velocity (m s−1 ) was significantly (p < 0.01) lower at each time point during the 400 m swimming trial in SC90% compared with SC100% and SCdrafting . There was no significant difference in velocity between SC100% and SCdrafting . Blood lactate (BLA) concentration was also significantly (p < 0.01) lower after swimming in SC90% compared to SC100% and SCdrafting (3.8 ± 0.9 versus 7.3 ± 2.4 and 7.9 ± 2.4 mM). The Pmean was also significantly (p < 0.05) lower in SC100% relative to the SC90% and SCdrafting (226 ± 15 versus 253 ± 33 and 249 ± 36 W). There was no significant correlation between PPO (W) and Pmean for SC100% (r = −0.32), SC90% (r = 0.65; p = 0.058) or SCdrafting (r = 0.54). This study indicates that drafting or swimming at a lower velocity did not induce any conflicting affects on power output during a subsequent cycle TT. However, this study confirms that Pmean during a cycle TT is reduced when prior swimming is performed. Furthermore the positive relationship typically observed between PPO and Pmean is
Corresponding author. E-mail address: [email protected] (D.J. Bentley).
1440-2440/$ — see front matter © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsams.2006.05.004
Swimming and subsequent cycling performance
235
disrupted by swimming activity performed before a cycling TT. This factor should be considered in terms of physiological analysis of triathletes. © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
Introduction The term ‘drafting’ is used to describe the tactic of performing a mode of activity in a sheltered position. Drafting has been researched in many sports such as cycling,1 kayaking,2 roller skating,3 triathlon4,5 and swimming.6—8 In swimming, it has been shown that the metabolic cost of exercise is lower by 5—10% when in a drafting position or that the maximal velocity is increased by 3.2—6.0% due to a reduction in frontal resistance or drag of 10—26%.2—8 Triathlon swimming is unique from competitive land based pool swimming since athletes are able to gain a benefit from drafting. During triathlon events, drafting is a tactic by which swimmers can stay with the leading competitors and conserve energy for the remaining disciplines.9,10 Drafting induces modification of stroke parameters such as stroke length and stroke index. Stroke length is particularly increased, due to a longer entry phase in combination with the decreased frontal drag, whereas the stroke rate is unchanged.7,8 The decrease in drag (10—26%)7,11 is important as it results in a lower blood lactate (BLA) concentration and improvement in exercise efficiency which are thought to be associated with an improvement in swimming performance within the range of 3.2—6.0%.7 Drafting successfully is a skilled and technical task and the distance to the feet of the lead swimmer has been shown to influence the change in metabolic responses; the optimal being between 0 and 50 cm11 but variable between different athletes.12 Hence the energy cost of drafting and indeed drafting during swimming could be important for improving swimming and subsequent cycling performance in a triathlon event. However this has not been clearly established in the scientific literature. The research concerning the affects of swimming on subsequent cycling performance in triathletes have shown that cycling performance is decreased compared to cycling without prior swimming.13—15 However other studies have not shown this reduction in performance.16 Such an affect could be a function of the duration or the intensity of the swimming or cycling trial.14,16 The swimming distances used in previous studies have been 750,14,16 80013 or 3000 m.16 Surprisingly the distance of the swim in an Olympic distance triathlon (1500 m) has
never been used in an experimental investigation concerning triathlon. Recently there has been an interest in the affects of drafting during swimming on subsequent physiological and biomechanical responses in submaximal cycling exercise.17,18 In these studies it has been shown that torque exerted on the crank system of a bike is lower during the down phase of pedalling when an athlete performs swimming in a non-drafting situation prior to the cycling bout.17 Furthermore, the metabolic cost is higher during submaximal cycling activity completed after a swimming trial compared to cycling alone.18 However, in cycling the use of metabolic efficiency as a variable to assess endurance performance has been questioned with some studies showing that this variable is not associated with performance in well-trained or elite athletes.19,20 Furthermore, the competitive demands are far different in elite triathlons as compared with recreational triathlon events.9,10 Also, whilst it is accepted that drafting in swimming results in a lower energy cost, no studies have directly compared the affects of drafting per se and swimming at a lower intensity (like which occurs during drafting) on subsequent cycling performance. The affects of drafting during swimming on short distance cycling time trial (TT) performance, as in an elite triathlon competition, have not yet been investigated. In contemporary elite Olympic distance triathlon the tactic of drafting during the cycle stage influences the demands of this stage as well as swimming and running tactics.9 It is known that in elite triathlon, the start and first third of the swim bout are a major determinant of the final race result.10 Vleck et al.10 reported that in an ITU World Cup, the top performers in the triathlon overall were significantly faster in the first 400—500 m of the swim stage. In addition, the first half of the cycling bout has also been shown to influence to a great extent the overall results9,10 essentially due to tactical reasons. For example Vleck et al.10 reported that during an ITU world Cup, there is a significant difference in cycling performance in the first 20 km between the slower and the faster swimmers, i.e., the slower swimmers cycled significantly faster in the first 20 km of the cycle stage than the faster ones. Therefore, one may argue that integrating these tactical data (analyses of the ‘critical moments’) rather than to simulate the overall
236
D.J. Bentley et al.
event might be more relevant to elite performers in order to assess their physiological responses in a very specific way. Therefore, the purpose of the present study was to investigate the physiological responses and performance during a cycling TT preceded by swimming bouts at different intensities including a trial where a triathlete was allowed to draft behind another swimmer. The peak power output (PPO) defined as the power output related to VO2,max during an incremental test to exhaustion is a variable that has been shown to be related to cycling TT performance in a triathlon.21 A secondary purpose of this study was to investigate the relationship between PPO obtained during an incremental exercise test to exhaustion and the average power output during a cycle TT performed after swimming in a drafting or non-drafting situation. We hypothesised that a 10% decrease in swimming velocity or in a drafting position would induce a significantly higher power output during a subsequent cycling time trial because of the energy saving that has occurred during the swimming activity. This in turn would affect the relationship between PPO and cycling performance after swimming exercise.
(Orion, S.T.E., Toulouse, France) to determine VO2,max and peak power output (PPO). On another three separate occasions a 20 min cycling time trial (TT) was performed in three different conditions: (1) preceded by 400 m freestyle at maximal velocity (SC100% ); (2) preceded by 400 m freestyle at 90% of SC100% (SC90% ); (3) preceded by 400 m freestyle in a drafting position at the same controlled velocity as SC100% (SCdrafting ). During SC100% , the subjects were instructed to swim as fast as possible over the 400 m and to be constant in order to achieve the best performance. SC100% was performed first then the order between the remaining trials was randomized, because the swim velocity during these two trials was calculated from the individualized 50 m pacing of each triathlete in SC100% . All swimming trials took place in an indoor 25 m swimming pool (temperature ∼20 ◦ C). In order to standardize the duration of the swim-cycle transition (2—3 min), the subsequent cycling tests were conducted adjacent to the swimming pool an environmentally controlled room (temperature ∼20 ◦ C; relative humidity ∼50%). The testing was performed during a period of 15 days with at least 48 h between each test session. Each test combination commenced at the same time of the day to control for any possible time of day variation in performance.22
Methods Incremental exercise test Subjects Nine regional to national level triathletes participated in the study on a voluntary basis and provided written, informed consent prior to participation. The Montpellier University institutional ethics committee approved the study. The subjects were all familiar with testing procedures, having regularly been tested with the same equipment and personnel as part of their training evaluation. The physical and training characteristics of the subjects are shown in Table 1. On average, the subjects’ weekly training loads were 7 ± 2 h in swimming, 4 ± 1 h in cycling and 5 ± 1 h in running.
Experimental design Each subject performed a maximal test to exhaustion using an electromagnetic cycle ergometer Table 1
The initial warm-up power of 30 W was maintained for 3 min after which the required power output was increased by 30 W every minute until fatigue. Fatigue coincided with a decrease in pedalling frequency of <70 rpm and the inability to maintain the required power output. The PPO (W) was defined as the workload corresponding to the highest fully completed stage. If fatigue occurred during a stage, PPO was calculated as the power output corresponding to the preceding stage plus a fraction of the final workload.23 Throughout the incremental test respiratory exchange data were collected continuously using a breath-by-breath automated metabolic system (CPX, Medical Graphics, MN, USA). Following completion of the test oxygen uptake (VO2 , l min−1 ) was averaged every 60 s. Heart rate (HR, b min−1 ) was continuously monitored by a 12-lead electrocardio-
Physical and training characteristics of the participating subjects (mean ± S.D.)
Subjects (n = 9)
Age (year)
Height (cm)
Body mass (kg)
Training time (h week−1 )
VO2,max (ml min−1 kg−1 )
HRmax (b min−1 )
PPO (W)
Mean S.D.
25.1 5.8
175.8 6.5
69.5 7.2
15.6 4.6
69.3 3.6
184.7 2.7
321.1 28.5
Swimming and subsequent cycling performance gram (ECG) (Medical Graphics, MN, USA). VO2,max was determined as the highest 60 s average VO2 values during the test. HRmax was measured as the highest consecutive HR values during the test.
Swimming-cycling trials Before each swimming trial a standardized warmup consisting of 800 m of swimming at low intensity (<60% of their best time on 400 m). During the SC90% , the required velocity calculated with the split times corresponding to 90% of the velocity of each 50 m of SC100% was set by an observer walking along the side of the pool and adjusted every 8.33 m. The SCdrafting was performed at the same velocity as in SC100% , but behind a lead swimmer in drafting condition, the pace was also set by an observer walking along the side of the pool as described previously. The draftee was instructed to be as close as possible to the lead swimmer. The average swimming velocity (m s−1 ) and stroke rate (SR, s min−1 ) were measured for each 50 m using a chronometer (KingTech, Taiwan) with frequency meter base 3. The stroke length (SL, m) was calculated for each 50 m by dividing the velocity by SR. The time taken (2—3 min) for the transition from the swimming pool to commencement of the cycling time trial was standardised. In the transition period, the subjects changed into their cycle shoes, a portable gas analysis system (K4b2 , Cosmed, Rome, Italy) was attached to the subject and the BLA concentration (mM) together with the RPE measured (see physiological measurements). For each cycling TT, the athletes exercised on the same bicycle used in training which was set on a home-trainer (Travel, Elite, Italy) and fitted with an SRM® ‘professional’ crankset (SRM® , J¨ ulich, Welldorf, Germany). During each cycling trial, power output (P) and pedalling cadence (PC) where measured continuously at a 1 s acquisition frequency from the SRM crankset via an SRM interface and stored in SRM software operating in Windows XP. Data were then averaged every 60 s. The average power (Pmean ) and cadence (PCmean ) of the 20-min trial were also calculated. The SRM crankset was calibrated by the same person according to the manufacturers’ recommended procedure before each test. The reliability and validity of the SRM ‘professional’ crank set for maximal exercise testing and TT has been previously examined.24 The BLA concentration was measured prior to each trial at rest (BLApre ), at the end of swimming (BLAswim ) and at the 10th minute in cycling (BLAcycling ). In addition BLA concentration was determined immediately at the end of cycling, after 1 min and then every 2 min until the concentration was found to
237 decrease. The blood samples were obtained from the finger using previously described procedures.25 The within-assay coefficient of variation (CV) for BLA as determined by repetitive measurements of the same sample was 5.2% and the between-assay CV was 6.1%. The highest value was recorded as the end-exercise concentration (BLApost ). The subject indicated RPE after swimming (RPEswim ), at the 10th minute of cycling trials (RPEcycle ) and immediately after exercise (RPEpost ) by pointing to the 6—20 Borg’s scale.26
Physiological measurements Oxygen uptake and HR where measured continuously by the K4b2 device throughout the cycle TT. The portable gas analysis system has previously been experimentally validated by comparison with the criterion Douglas bags.27 During the experiment, the respiratory data were stored breath by breath. Data were then averaged every 60 s. The average VO2 and HR for the total duration (20-min) of each TT were also calculated. Prior to each test, the K4b2 system was calibrated using ambient air (which partial O2 composition was assumed to be 20.9%) and a gas of known O2 (16%) and CO2 (5%) concentration. The calibration of the turbine flowmeter of the K4b2 was performed with a 3-L syringe (Quinton Instruments, Seattle, USA).
Statistical analysis The results are presented as mean ± S.D. The normality of the distribution of the variables and the homogeneity of variance were tested and accepted (SigmaStat, Jandel scientific, San Rafael, CA). A series of two-way (trial × sample time) analysis of variance (ANOVA) complemented by a Tuckey post hoc test was used to identify differences in all parameters between the three cycling time trials. Pearson correlation coefficients were used to test the relationships between PPO and Pmean in the three conditions. For all statistical analyses, a p value of 0.05 was accepted as the level of statistical significance.
Results Swimming performance The overall velocity (m s−1 ) in SC90% was significantly (p < 0.01) lower compared to SC100% and SCdrafting (1.13 ± 0.10 versus 1.26 ± 0.11 and 1.27 ± 0.10). There was no significant difference in velocity between SC100% and SCdrafting . The overall
238
D.J. Bentley et al.
Figure 1 The mean (±S.D.) peak power output (PPO) obtained in the incremental test and mean power output (Pmean ) during the cycling time trial in each condition. * p < 0.05 Significant difference between SC90% and SCdraft . $ Significantly (p < 0.05) different from PPO.
SR was significantly (p < 0.01) lower in SC90% (31 ± 4) compared to SC100% (37 ± 4) and SCdrafting (38 ± 4). There was no significant difference in SR between SC100% and SCdrafting . There were no significant differences for SL between each trial (2.08 ± 0.14 versus 2.19 ± 0.14 versus 2.05 ± 0.15 m for SC100% , SC90% and SCdrafting , respectively).
Figure 2 The mean power output (W) (A) and pedalling cadence (PC) (rpm) (B) during the cycling time trial (TT) in each swimming-cycling (SC) condition. No significant differences between conditions.
Cycling performance The Pmean in each condition corresponded to 72 ± 8, 78 ± 7 and 80 ± 13% of PPO for SC100% , SC90% and SCdrafting , respectively. The Pmean was significantly lower relative to PPO in each SC condition (p < 0.01) (Fig. 1). The Pmean was also significantly lower in SC100% relative to the SC90% and SCdrafting (p < 0.05) (Fig. 1). The power output in SC100% decreased continuously during the first 10 min of the trial, and then remained stable during the second half. In contrast in SCdraft , the power output remained stable throughout the whole 20-min. Despite this, the power output (W) was not significantly different at each time point in each trial (Fig. 2A). There was no significant correlation between PPO (W) and Pmean for SC100% (r = −0.32), SC90% (r = 0.65; p = 0.058) or SCdrafting (r = 0.54). As shown in Fig. 2B, the variability in PC was larger in SC100% than in the other conditions. Despite this the PC was not significantly different at each time point in each of the three trials (Fig. 2B). There was also no significant difference in PCmean between the three conditions (93 ± 7, 96 ± 8, and 90 ± 7 in SC100% , SC90% and SCdrafting , respectively).
Physiological parameters during cycling The average VO2 during SC100% was not significantly different compared to SC90% and SCdrafting
(57 ± 5 versus 56 ± 6 and 54 ± 6 ml kg−1 min−1 , respectively, corresponding to 81 ± 7, 81 ± 9 and 79 ± 8 of VO2,max for SC100% , SC90% and SCdrafting , respectively. No significant difference for VO2 was observed at each time point during the cycling TT in the three conditions (Fig. 3A). The mean HR recorded during SC100% (177 ± 5) was significantly higher than in SCdrafting (171 ± 6) (p < 0.05). There was no significant difference in the mean HR between SC100% and SC90% . The HR in SCdrafting was significantly lower compared with SC90% and SC100% between the 6th and 13th minutes of the trial (Fig. 3B).
Blood lactate and rating of perceived exertion The BLA concentration was significantly higher at each time point (swim, cycle and post) relative to rest regardless of the SC condition (p < 0.01). The BLAswim was significantly higher in SC100% and SCdrafting compared with SC90% (p < 0.01). Aside from this result, BLAcycle and BLApost were not significantly different between each SC condition. RPEswim was significantly higher in SC100% compared in SC90% whereas no further differences in RPE were observed between the different conditions at each time point (Fig. 4B).
Swimming and subsequent cycling performance
Figure 3 The mean oxygen uptake (VO2 ) (ml kg−1 min−1 ) (A) and heart rate (HR) (b min−1 ) (B) during the cycle time trial (TT) in each swimming-cycle (SC) condition. $ Significantly (p < 0.05) different between SCdraft and SC90% ; # significantly (p < 0.05) different between SCdraft and SC100% .
Discussion This study was conducted to compare the affects of drafting or exercise at a lower relative exercise intensity in swimming on subsequent cycle TT performance in trained triathletes. The most important finding was that the power output during a 20 min TT was significantly lower after a 400 m all out freestyle swimming trial performed at maximal effort as compared to swimming at either 90% of this velocity or in a drafting situation. Importantly there was no significant difference in the power output during cycling performance after swimming at 90% or in a drafting position. The results of this study also demonstrate that all out swimming prior to cycling may affect the positive relationship between PPO and the average power output generated during a cycle TT which is typically observed when examining a cycle TT performed in isolation. Studies investigating the effects of drafting in triathlon have predominantly concerned only the cycle-run transition.4,5 It has been shown that drafting during the cycle stage of a simulated triathlon results in a significant reduction in metabolic load, leading to a higher running
239
Figure 4 A The mean (±S.D.) blood lactate (BLA) concentration (mM) (A) and rating of perceived exertion (RPE) (B) before, during and after the swimming and cycle section of each swimming-cycle condition. ** Significantly (p < 0.01) different between SC90% compared with SC100% and SCdraft .
efficiency.4,5 The reduction in cycling performance during the SC100% shown in this study is similar to that observed by previous investigators who found that drafting positively influenced the physiological responses during submaximal cycle exercise.17,18 Specifically these experiments have shown that the decrease in metabolic load when swimming in a drafting position could lead either to a higher cycling efficiency17 and induce significantly lower pedal rate or higher mean and peak resultant torques during submaximal set workload cycling.18 It is difficult to compare the results of the present study to those previous investigations17,18 as the cycle exercise performed in these experiments (10 or 15 min, set workload trial) was far different to the 20 min TT that was used in our study. In the present study, we showed that drafting during swimming could lead to an increase in performance. However the results of the present study and those of other researchers17,18 clearly show that drafting has a positive affect on cycling performance. The power output generated by elite triathletes is typically higher in the early stages
240 of the 40 km distance in order to catch up with or to break away from the lead group of athletes.10 Therefore, limiting the negative effects of swimming activity on cycling power output may be particularly important in a draft-legal event, since it would provide a means of establishing a good position and maintaining contact with the leading cyclists. Previous studies concerning the affects of swimming on cycling in triathletes have only compared the affects on cycling performance of swimming at a typical competition intensity compared to a cycling trial in isolation.13,15,16 Furthermore, the comparative affects of drafting during swimming or at a lower intensity have not been examined. Kreider et al.13 found that 800 m of sub-maximal swimming before 75 min of cycling at 75% of maximal oxygen uptake (VO2,max ) resulted in a significant reduction in power output (159 W versus 191 W) when compared to a control 75 min cycle without prior swimming. Peeling et al.14 were able to demonstrate that 20-km cycling performance was significantly lower when 750 m of swimming was performed at 90% compared to an all out trial. The results of present study confirm these findings, but also extends this data by showing that drafting or swimming at 90% of the velocity during an all out 400 m trial induces similar performance affects during a cycling TT. This data should be considered by coaches and scientists when modeling performance trials or training strategies. The reduction in Pmean during the SC100% observed in the present study could be associated with the higher BLA concentration after swimming in SC100% . However it is interesting that similar lactate levels were observed following swimming in the SCdrafting . These results are conflicting with previous studies where a lower BLA concentration was reported in the drafting condition.6,7 However in these studies the swimming velocity in the drafting and non-drafting conditions was the same. Therefore, BLA concentration was bound to be lower in a drafting situation indicative of lower exercise stress. Furthermore the ability to draft is a skill that requires practice and may result in very different responses between different athletes. It is likely that the present subjects were not skilled in this technique. In addition, since the swim trials were performed in a 25 m swimming pool, a large part of the exercise time was when each athlete would turn after each lap. Therefore it is not surprising that the decrease in BLA concentration or change in stroke pattern are lower than in studies performed in 50-m pool. Hence it is not likely that BLA concentration was not associated with the reduction in power output in the SC100% .
D.J. Bentley et al. It has been reported that drafting in swimming induces a change in stroking parameters (i.e., increased stroke length and stroke index and an unchanged stroke frequency). In addition, the kicking pattern is also modified and drafting is associated to a decrease in leg frequency.8 Although we did not control the leg kicking frequency of the subjects, one may speculate that the change in kicking pattern could induce a lower leg fatigue and explain the improved performance in the subsequent cycling performance. Indeed whilst we did not find any significant differences between conditions in the first 5—10 min, the power output appeared to be lower in this part of the cycle TT in SC100% . Hence we interpret this result as a sign of fatigue as some subjects took longer to adjust the PC and power output in SC100% compared with the other conditions. That aside the results of this study indicate that swimming at a maximal effort does indeed affect cycling performance relative to swimming at lower intensity or in a drafting situation. Furthermore short distance cycle TT performance remains unchanged after swimming at a lower intensity or in a drafting position. However as demonstrated in Fig. 2B, the variability in PC is larger in SC100% than in the other conditions. We interpret this result as a sign of fatigue since the subjects took longer to adjust their PC to the right value in SC100% than in the other conditions. The distances chosen for this study (400 m swim, 20 min cycle) were shorter than in an ‘Olympic distance’ triathlon. However research indicates that the first 400 m of the swimming stage and the first 15 km of the cycling stages are important for deciding overall race positions in an elite triathlon.10 Therefore, in this study we are confident that the trials administered were specific to elite triathletes and easier to administer in the context of the design of this study. This aside it could be that different duration and intensity swimming bouts (as in ‘sprint’, ‘Olympic’ or ‘Long Distance’ triathlon) may elicit different physiological and performance changes during a subsequent cycling task. Laursen et al.16 demonstrated that a 3000 m swimming at a self-selected intensity induced no significant differences in the average power output sustained over a 3 h cycle time trial as compared to a control conditions without prior swimming (212 W versus 222 W). Whereas other recent studies have shown higher intensity swimming bouts to affect cycling performance.14 In a longer distance triathlon, reduction in power output would reflect a residual fatigue whereas during a short-distance triathlon, decrease in power output would be linked to the physiological changes
Swimming and subsequent cycling performance directly due to the swim-cycle transition. Whyte et al.28 showed that factors such as a long duration swim prior to the cycling component, may impact upon cycle performance in ultra-endurance triathlon, but suggest that the more conditioned athletes may be better able to maintain a high cardiac output during prolonged cycling and therefore reduce the decrease in cycling performance. During a longer distance triathlon race the effects of swimming on cycling may not be as significant because the power output required in the early stages of the cycling trial might be lower as compared to a short distance triathlon. All studies conducted so far examining the swim to cycle transition have used a model where one velocity is selected in the swimming trial. Indeed the velocity during the different 400 m trials of the present study was relatively constant and not different between the conditions. However the swimming velocity (and conceivably exercise intensity) in an elite triathlon has been shown to be far greater in the early stages of the swimming stage as compared to the latter stages.10 This is a limitation of most studies examining the swimming to cycling transition in triathlon. Hence future investigations concerning swimming in triathlon should consider the specific physiological demands of this stage and replicate during experimental trials. It is of interest that the increase in cycling performance subsequent to a 400 m swim bout at submaximal intensity or in drafting condition was of ∼7% in the present study. This is in line with previous experiments where the increase in performance or efficiency was of 5—7%.14,16,17 The improvement in performance is therefore not significantly higher due to the shorter swimming bout. The PPO is a physiological variable that has been shown to be related to cycling performance in the field21—23 or in a laboratory based setting.29,30 Typically correlation coefficients of r ≥ 0.85 have been reported in the literature between PPO and the average power output during a cycling TT.19,29,30 In this study we did not find any significant correlation between PPO and Pmean during the cycle TT in each SC condition. These results suggest that whatever the swimming condition prior to a cycling performance trial, the relationship between PPO and performance is affected. Sport scientists often quantify PPO in conjunction with other physiological measures to predict performance or to determine the power output generated during a cycling trial.21,23,30 Hence the validity of using this variable for these purposes in triathletes is questioned.
241 In the present study we did not include a control cycle TT. In addition no running bout was performed after the cycling. Based on previous studies, it is likely that the power output generated in an isolated trial would have been far greater than that in the three trials that were performed in this study.13,15 Previous investigations have also included a running ‘stage’ during an experimental trial.13,14 In one study, the subjects swam 800 m, cycled for 75 min and ran 10 km, meaning that they may have adjusted their pacing strategy in cycling due to the following run.13 In a more recent investigation the affects of exercise intensity in swimming were examined in running and cycling in a sprint distance (750 m, 20 km and 5 km).14 These authors found no significant change in running performance regardless of the intensity of exercise in swimming. These findings more than likely demonstrate that running performance is a function of the residual fatigue occurring in the lower limb muscles (i.e., vastus lateralis and medialis, gastrocnemius) following cycling. Other researchers have also recommended that at least in a sprint triathlon, athletes should perform the cycle stage as maximally as possible as manipulation of the intensity in the cycle trial has no marked affects on running performance (Suriano, unpublished). Hence it is likely that the cycle TT results would have been the same if the running performance trial had also been performed after the completion of the cycle stage. That aside the primary purpose of this study was to examine any performance or physiological differences in swimcycle trials with varying swimming intensity and characteristics. We have demonstrated that whilst swimming may affect cycling performance, drafting results in a similar performance response during cycling. This has obvious implications for training of triathletes or the tactical approach to swimming in an elite triathlon event. However, the present results cannot be transposed directly to the ‘Olympic distance triathlon’ since the effects of a 400-m swim are certainly different than those of a 750 or 1500 m swim. However it is suggested that for tactical reasons, the present experimental design was more relevant to elite than to sub-elite triathletes. In summary swimming over 400 m at the maximal velocity prior cycling led to a decreased cycling performance in trained triathletes. However this result was not observed when the swim intensity was 10% lower or performed in drafting position. Regardless of what type of swimming trial that was performed prior to cycling, the relationship between PPO and cycling performance was affected. Further studies should examine the
242 interaction between swimming duration and intensity on cycling performance. Furthermore the specific physiological demands of the swimming stage of a competition should be replicated in studies examining the swimming to cycling transition in triathlon.
Practical implications • Training for triathletes should incorporate consist of specific swimming-cycling sessions incorporating swimming bouts simulating a competition intensity at 90% of a maximal velocity for a given race distance. • Specific swimming-cycling training sessions need to be completed regularly in drafting conditions. • Athletes may have to control and stabilise their pedalling cadence when performing cycling after swimming in training.
Acknowledgements The authors thank the subjects for their enthusiastic participation and their availability and Mr. Roudil, director of the Inter-universities swimming pool for his help.
References 1. Olds TS, Norton KI, Lowe EL, Olive S, Reay F, Ly S. Modeling road-cycling performance. J Appl Physiol 1995;78:1596—611. 2. Gray GL, Matheson GO, McKenzie DC. The metabolic cost of two kayaking techniques. Int J Sports Med 1995;16: 250—4. 3. Rundell KW. Effects of drafting during short-track speed skating. Med Sci Sports Exerc 1996;28:765—71. 4. Hausswirth C, Lehenaff D, Dreano P, Savonen K. Effects of cycling alone or in a sheltered position on subsequent running performance during a triathlon. Med Sci Sports Exerc 1999;31:599—604. 5. Hausswirth C, Vallier JM, Lehenaff D, Brisswalter J, Smith D, Millet G, et al. Effect of two drafting modalities in cycling on running performance. Med Sci Sports Exerc 2001;33:485—92. 6. Bassett Jr DR, Flohr J, Duey WJ, Howley ET, Pein RL. Metabolic responses to drafting during front crawl swimming. Med Sci Sports Exerc 1991;23:744—7. 7. Chatard JC, Chollet D, Millet G. Performance and drag during drafting swimming in highly trained triathletes. Med Sci Sports Exerc 1998;30:1276—80. 8. Chollet D, Hue O, Auclair F, Millet G, Chatard JC. The effects of drafting on stroking variations during swimming in elite male triathletes. Eur J Appl Physiol 2000;82: 413—7.
D.J. Bentley et al. 9. Bentley DJ, Millet GP, Vleck VE, McNaughton LR. Specific aspects of contemporary triathlon: implications for physiological analysis and performance. Sports Med 2002;32:345—59. 10. Vleck VE, Burgi A, Bentley DJ. The consequences of swim, cycle, and run performance on overall result in elite olympic distance triathlon. Int J Sports Med 2006;27: 43—8. 11. Chatard JC, Wilson B. Drafting distance in swimming. Med Sci Sports Exerc 2003;35:1176—81. 12. Millet G, Chollet D, Chatard JC. Effects of drafting behind a two- or a six-beat kick swimmer in elite female triathletes. Eur J Appl Physiol 2000;82:465—71. 13. Kreider R, Boone T, Thompson W, Burkes S, Cortes C. Cardiovascular and thermal responses of triathlon performance. Med Sci Sports Exer 1988;20:385—90. 14. Peeling PD, Bishop DJ, Landers GJ. Effect of swimming intensity on subsequent cycling and overall triathlon performance. Br J Sports Med 2005;39: 960—4. 15. Delextrat A, Brisswalter J, Hausswirth C, Bernard T, Vallier JM. Does prior 1500-m swimming affect cycling energy expenditure in well-trained triathletes? Can J Appl Physiol 2005;30:392—403. 16. Laursen PB, Rhodes EC, Langill RH. The effects of 3000-m swimming on subsequent 3-h cycling performance: implications for ultraendurance triathletes. Eur J Appl Physiol 2000;83:28—33. 17. Delextrat A, Tricot V, Bernard T, Vercruyssen F, Hausswirth C, Brisswalter J. Drafting during swimming improves efficiency during subsequent cycling. Med Sci Sports Exerc 2003;35:1612—9. 18. Delextrat A, Tricot V, Bernard T, Vercruyssen F, Hausswirth C, Brisswalter J. Modification of cycling biomechanics during a swim-to-cycle trial. J Appl Biomech 2005;21: 297—308. 19. Bentley DJ, Vleck VE, Millet GP. The isocapnic buffering phase and mechanical efficiency: relationship to cycle time trial performance of short and long duration. Can J Appl Physiol 2005;30:46—60. 20. Moseley L, Achten J, Martin JC, Jeukendrup AE. No differences in cycling efficiency between world-class and recreational cyclists. Int J Sports Med 2004;25: 374—9. 21. Bentley DJ, Wilson GJ, Davie AJ, Zhou S. Correlations between peak power output, muscular strength and cycle time trial performance in triathletes. J Sports Med Phys Fitness 1998;38:201—7. 22. Atkinson G, Reilly T. Circadian variation in sports performance. Sports Med 1996;21:292—312. 23. Hawley JA, Noakes TD. Peak power output predicts maximal oxygen uptake and performance time in trained cyclists. Eur J Appl Physiol 1992;65:79—83. 24. Balmer J, Davison RC, Coleman DA, Bird SR. The validity of power output recorded during exercise performance tests using a Kingcycle air-braked cycle ergometer when compared with an SRM powermeter. Int J Sports Med 2000;21:195—9. 25. Davison RC, Coleman D, Balmer J, Nunn M, Theakston S, Burrows M, et al. Assessment of blood lactate: practical evaluation of the Biosen 5030 lactate analyzer. Med Sci Sports Exerc 2000;32:243—7. 26. Borg G. Perceived exertion as an indicator of somatic stress. Scand J Rehabil Med 1970;2:92—8. 27. McLaughlin JE, King GA, Howley Jr DR, Ainsworth BE. Validation of the COSMED K4 b2 portable metabolic system. Int J Sports Med 2001;22:280—4.
Swimming and subsequent cycling performance 28. Whyte G, Lumley S, George K, Gates P, Sharma S, Prasad K, et al. Physiological profile and predictors of cycling performance in ultra-endurance triathletes. J Sports Med Phys Fit 2000;40:103—9. 29. Bentley DJ, McNaughton LR, Thompson D, Vleck VE, Batterham AM. Peak power output, the lactate threshold, and
243 time trial performance in cyclists. Med Sci Sports Exerc 2001;33:2077—81. 30. Balmer J, Davison RC, Bird SR. Peak power predicts performance power during an outdoor 16, 1-km cycling time trial. Med Sci Sports Exerc 2000;32:1485—90.
ORIGINAL PAPER
The effects of exercise intensity or drafting during swimming on subsequent cycling performance in triathletes David J. Bentley a,b,∗, Sebastien Libicz c, Aurelie Jougla c, Olivier Coste c, Jerome Manetta c, Karim Chamari e, Gregoire P. Millet c,d a
Health and Exercise, School of Medical Science, University of NSW, Sydney, Australia Department of Human and Health Science, University of Westminster, London, United Kingdom c Faculty of Sport Sciences, University of Montpellier, Montpellier, France d Aspire, Academy for Sports Excellence, Doha, Qatar e National Centre of Medicine and Science in Sport (CNMSS), Tunis, Tunisia b
Received 19 January 2006 ; received in revised form 1 May 2006; accepted 1 May 2006 KEYWORDS Ergometer; Triathlon; Athletes; Power; Metabolism; Time trial
∗
Summary The purpose of this study was to compare the affects of drafting or a reduction of exercise intensity during swimming on the power output sustained (Pmean ) during a subsequent cycle time trial (TT). In addition the relationship between peak power output (PPO) and Pmean generated during the cycle TT after swimming was examined. Nine well-trained triathletes performed an incremental cycling test to exhaustion for determination of PPO. In addition, each subject performed three swim-cycle (SC) trials consisting of 20 min cycle TT preceded by a 400 m swimming trial completed as (1) ‘‘all out’’ and in a non-drafting situation (SC100% ); (2) at 90% of SC100% in a non-drafting situation (SC90% ); (3) in a drafting position at the same controlled velocity as SC100% (SCdrafting ). Swimming velocity (m s−1 ) was significantly (p < 0.01) lower at each time point during the 400 m swimming trial in SC90% compared with SC100% and SCdrafting . There was no significant difference in velocity between SC100% and SCdrafting . Blood lactate (BLA) concentration was also significantly (p < 0.01) lower after swimming in SC90% compared to SC100% and SCdrafting (3.8 ± 0.9 versus 7.3 ± 2.4 and 7.9 ± 2.4 mM). The Pmean was also significantly (p < 0.05) lower in SC100% relative to the SC90% and SCdrafting (226 ± 15 versus 253 ± 33 and 249 ± 36 W). There was no significant correlation between PPO (W) and Pmean for SC100% (r = −0.32), SC90% (r = 0.65; p = 0.058) or SCdrafting (r = 0.54). This study indicates that drafting or swimming at a lower velocity did not induce any conflicting affects on power output during a subsequent cycle TT. However, this study confirms that Pmean during a cycle TT is reduced when prior swimming is performed. Furthermore the positive relationship typically observed between PPO and Pmean is
Corresponding author. E-mail address: [email protected] (D.J. Bentley).
1440-2440/$ — see front matter © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jsams.2006.05.004
Swimming and subsequent cycling performance
235
disrupted by swimming activity performed before a cycling TT. This factor should be considered in terms of physiological analysis of triathletes. © 2006 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
Introduction The term ‘drafting’ is used to describe the tactic of performing a mode of activity in a sheltered position. Drafting has been researched in many sports such as cycling,1 kayaking,2 roller skating,3 triathlon4,5 and swimming.6—8 In swimming, it has been shown that the metabolic cost of exercise is lower by 5—10% when in a drafting position or that the maximal velocity is increased by 3.2—6.0% due to a reduction in frontal resistance or drag of 10—26%.2—8 Triathlon swimming is unique from competitive land based pool swimming since athletes are able to gain a benefit from drafting. During triathlon events, drafting is a tactic by which swimmers can stay with the leading competitors and conserve energy for the remaining disciplines.9,10 Drafting induces modification of stroke parameters such as stroke length and stroke index. Stroke length is particularly increased, due to a longer entry phase in combination with the decreased frontal drag, whereas the stroke rate is unchanged.7,8 The decrease in drag (10—26%)7,11 is important as it results in a lower blood lactate (BLA) concentration and improvement in exercise efficiency which are thought to be associated with an improvement in swimming performance within the range of 3.2—6.0%.7 Drafting successfully is a skilled and technical task and the distance to the feet of the lead swimmer has been shown to influence the change in metabolic responses; the optimal being between 0 and 50 cm11 but variable between different athletes.12 Hence the energy cost of drafting and indeed drafting during swimming could be important for improving swimming and subsequent cycling performance in a triathlon event. However this has not been clearly established in the scientific literature. The research concerning the affects of swimming on subsequent cycling performance in triathletes have shown that cycling performance is decreased compared to cycling without prior swimming.13—15 However other studies have not shown this reduction in performance.16 Such an affect could be a function of the duration or the intensity of the swimming or cycling trial.14,16 The swimming distances used in previous studies have been 750,14,16 80013 or 3000 m.16 Surprisingly the distance of the swim in an Olympic distance triathlon (1500 m) has
never been used in an experimental investigation concerning triathlon. Recently there has been an interest in the affects of drafting during swimming on subsequent physiological and biomechanical responses in submaximal cycling exercise.17,18 In these studies it has been shown that torque exerted on the crank system of a bike is lower during the down phase of pedalling when an athlete performs swimming in a non-drafting situation prior to the cycling bout.17 Furthermore, the metabolic cost is higher during submaximal cycling activity completed after a swimming trial compared to cycling alone.18 However, in cycling the use of metabolic efficiency as a variable to assess endurance performance has been questioned with some studies showing that this variable is not associated with performance in well-trained or elite athletes.19,20 Furthermore, the competitive demands are far different in elite triathlons as compared with recreational triathlon events.9,10 Also, whilst it is accepted that drafting in swimming results in a lower energy cost, no studies have directly compared the affects of drafting per se and swimming at a lower intensity (like which occurs during drafting) on subsequent cycling performance. The affects of drafting during swimming on short distance cycling time trial (TT) performance, as in an elite triathlon competition, have not yet been investigated. In contemporary elite Olympic distance triathlon the tactic of drafting during the cycle stage influences the demands of this stage as well as swimming and running tactics.9 It is known that in elite triathlon, the start and first third of the swim bout are a major determinant of the final race result.10 Vleck et al.10 reported that in an ITU World Cup, the top performers in the triathlon overall were significantly faster in the first 400—500 m of the swim stage. In addition, the first half of the cycling bout has also been shown to influence to a great extent the overall results9,10 essentially due to tactical reasons. For example Vleck et al.10 reported that during an ITU world Cup, there is a significant difference in cycling performance in the first 20 km between the slower and the faster swimmers, i.e., the slower swimmers cycled significantly faster in the first 20 km of the cycle stage than the faster ones. Therefore, one may argue that integrating these tactical data (analyses of the ‘critical moments’) rather than to simulate the overall
236
D.J. Bentley et al.
event might be more relevant to elite performers in order to assess their physiological responses in a very specific way. Therefore, the purpose of the present study was to investigate the physiological responses and performance during a cycling TT preceded by swimming bouts at different intensities including a trial where a triathlete was allowed to draft behind another swimmer. The peak power output (PPO) defined as the power output related to VO2,max during an incremental test to exhaustion is a variable that has been shown to be related to cycling TT performance in a triathlon.21 A secondary purpose of this study was to investigate the relationship between PPO obtained during an incremental exercise test to exhaustion and the average power output during a cycle TT performed after swimming in a drafting or non-drafting situation. We hypothesised that a 10% decrease in swimming velocity or in a drafting position would induce a significantly higher power output during a subsequent cycling time trial because of the energy saving that has occurred during the swimming activity. This in turn would affect the relationship between PPO and cycling performance after swimming exercise.
(Orion, S.T.E., Toulouse, France) to determine VO2,max and peak power output (PPO). On another three separate occasions a 20 min cycling time trial (TT) was performed in three different conditions: (1) preceded by 400 m freestyle at maximal velocity (SC100% ); (2) preceded by 400 m freestyle at 90% of SC100% (SC90% ); (3) preceded by 400 m freestyle in a drafting position at the same controlled velocity as SC100% (SCdrafting ). During SC100% , the subjects were instructed to swim as fast as possible over the 400 m and to be constant in order to achieve the best performance. SC100% was performed first then the order between the remaining trials was randomized, because the swim velocity during these two trials was calculated from the individualized 50 m pacing of each triathlete in SC100% . All swimming trials took place in an indoor 25 m swimming pool (temperature ∼20 ◦ C). In order to standardize the duration of the swim-cycle transition (2—3 min), the subsequent cycling tests were conducted adjacent to the swimming pool an environmentally controlled room (temperature ∼20 ◦ C; relative humidity ∼50%). The testing was performed during a period of 15 days with at least 48 h between each test session. Each test combination commenced at the same time of the day to control for any possible time of day variation in performance.22
Methods Incremental exercise test Subjects Nine regional to national level triathletes participated in the study on a voluntary basis and provided written, informed consent prior to participation. The Montpellier University institutional ethics committee approved the study. The subjects were all familiar with testing procedures, having regularly been tested with the same equipment and personnel as part of their training evaluation. The physical and training characteristics of the subjects are shown in Table 1. On average, the subjects’ weekly training loads were 7 ± 2 h in swimming, 4 ± 1 h in cycling and 5 ± 1 h in running.
Experimental design Each subject performed a maximal test to exhaustion using an electromagnetic cycle ergometer Table 1
The initial warm-up power of 30 W was maintained for 3 min after which the required power output was increased by 30 W every minute until fatigue. Fatigue coincided with a decrease in pedalling frequency of <70 rpm and the inability to maintain the required power output. The PPO (W) was defined as the workload corresponding to the highest fully completed stage. If fatigue occurred during a stage, PPO was calculated as the power output corresponding to the preceding stage plus a fraction of the final workload.23 Throughout the incremental test respiratory exchange data were collected continuously using a breath-by-breath automated metabolic system (CPX, Medical Graphics, MN, USA). Following completion of the test oxygen uptake (VO2 , l min−1 ) was averaged every 60 s. Heart rate (HR, b min−1 ) was continuously monitored by a 12-lead electrocardio-
Physical and training characteristics of the participating subjects (mean ± S.D.)
Subjects (n = 9)
Age (year)
Height (cm)
Body mass (kg)
Training time (h week−1 )
VO2,max (ml min−1 kg−1 )
HRmax (b min−1 )
PPO (W)
Mean S.D.
25.1 5.8
175.8 6.5
69.5 7.2
15.6 4.6
69.3 3.6
184.7 2.7
321.1 28.5
Swimming and subsequent cycling performance gram (ECG) (Medical Graphics, MN, USA). VO2,max was determined as the highest 60 s average VO2 values during the test. HRmax was measured as the highest consecutive HR values during the test.
Swimming-cycling trials Before each swimming trial a standardized warmup consisting of 800 m of swimming at low intensity (<60% of their best time on 400 m). During the SC90% , the required velocity calculated with the split times corresponding to 90% of the velocity of each 50 m of SC100% was set by an observer walking along the side of the pool and adjusted every 8.33 m. The SCdrafting was performed at the same velocity as in SC100% , but behind a lead swimmer in drafting condition, the pace was also set by an observer walking along the side of the pool as described previously. The draftee was instructed to be as close as possible to the lead swimmer. The average swimming velocity (m s−1 ) and stroke rate (SR, s min−1 ) were measured for each 50 m using a chronometer (KingTech, Taiwan) with frequency meter base 3. The stroke length (SL, m) was calculated for each 50 m by dividing the velocity by SR. The time taken (2—3 min) for the transition from the swimming pool to commencement of the cycling time trial was standardised. In the transition period, the subjects changed into their cycle shoes, a portable gas analysis system (K4b2 , Cosmed, Rome, Italy) was attached to the subject and the BLA concentration (mM) together with the RPE measured (see physiological measurements). For each cycling TT, the athletes exercised on the same bicycle used in training which was set on a home-trainer (Travel, Elite, Italy) and fitted with an SRM® ‘professional’ crankset (SRM® , J¨ ulich, Welldorf, Germany). During each cycling trial, power output (P) and pedalling cadence (PC) where measured continuously at a 1 s acquisition frequency from the SRM crankset via an SRM interface and stored in SRM software operating in Windows XP. Data were then averaged every 60 s. The average power (Pmean ) and cadence (PCmean ) of the 20-min trial were also calculated. The SRM crankset was calibrated by the same person according to the manufacturers’ recommended procedure before each test. The reliability and validity of the SRM ‘professional’ crank set for maximal exercise testing and TT has been previously examined.24 The BLA concentration was measured prior to each trial at rest (BLApre ), at the end of swimming (BLAswim ) and at the 10th minute in cycling (BLAcycling ). In addition BLA concentration was determined immediately at the end of cycling, after 1 min and then every 2 min until the concentration was found to
237 decrease. The blood samples were obtained from the finger using previously described procedures.25 The within-assay coefficient of variation (CV) for BLA as determined by repetitive measurements of the same sample was 5.2% and the between-assay CV was 6.1%. The highest value was recorded as the end-exercise concentration (BLApost ). The subject indicated RPE after swimming (RPEswim ), at the 10th minute of cycling trials (RPEcycle ) and immediately after exercise (RPEpost ) by pointing to the 6—20 Borg’s scale.26
Physiological measurements Oxygen uptake and HR where measured continuously by the K4b2 device throughout the cycle TT. The portable gas analysis system has previously been experimentally validated by comparison with the criterion Douglas bags.27 During the experiment, the respiratory data were stored breath by breath. Data were then averaged every 60 s. The average VO2 and HR for the total duration (20-min) of each TT were also calculated. Prior to each test, the K4b2 system was calibrated using ambient air (which partial O2 composition was assumed to be 20.9%) and a gas of known O2 (16%) and CO2 (5%) concentration. The calibration of the turbine flowmeter of the K4b2 was performed with a 3-L syringe (Quinton Instruments, Seattle, USA).
Statistical analysis The results are presented as mean ± S.D. The normality of the distribution of the variables and the homogeneity of variance were tested and accepted (SigmaStat, Jandel scientific, San Rafael, CA). A series of two-way (trial × sample time) analysis of variance (ANOVA) complemented by a Tuckey post hoc test was used to identify differences in all parameters between the three cycling time trials. Pearson correlation coefficients were used to test the relationships between PPO and Pmean in the three conditions. For all statistical analyses, a p value of 0.05 was accepted as the level of statistical significance.
Results Swimming performance The overall velocity (m s−1 ) in SC90% was significantly (p < 0.01) lower compared to SC100% and SCdrafting (1.13 ± 0.10 versus 1.26 ± 0.11 and 1.27 ± 0.10). There was no significant difference in velocity between SC100% and SCdrafting . The overall
238
D.J. Bentley et al.
Figure 1 The mean (±S.D.) peak power output (PPO) obtained in the incremental test and mean power output (Pmean ) during the cycling time trial in each condition. * p < 0.05 Significant difference between SC90% and SCdraft . $ Significantly (p < 0.05) different from PPO.
SR was significantly (p < 0.01) lower in SC90% (31 ± 4) compared to SC100% (37 ± 4) and SCdrafting (38 ± 4). There was no significant difference in SR between SC100% and SCdrafting . There were no significant differences for SL between each trial (2.08 ± 0.14 versus 2.19 ± 0.14 versus 2.05 ± 0.15 m for SC100% , SC90% and SCdrafting , respectively).
Figure 2 The mean power output (W) (A) and pedalling cadence (PC) (rpm) (B) during the cycling time trial (TT) in each swimming-cycling (SC) condition. No significant differences between conditions.
Cycling performance The Pmean in each condition corresponded to 72 ± 8, 78 ± 7 and 80 ± 13% of PPO for SC100% , SC90% and SCdrafting , respectively. The Pmean was significantly lower relative to PPO in each SC condition (p < 0.01) (Fig. 1). The Pmean was also significantly lower in SC100% relative to the SC90% and SCdrafting (p < 0.05) (Fig. 1). The power output in SC100% decreased continuously during the first 10 min of the trial, and then remained stable during the second half. In contrast in SCdraft , the power output remained stable throughout the whole 20-min. Despite this, the power output (W) was not significantly different at each time point in each trial (Fig. 2A). There was no significant correlation between PPO (W) and Pmean for SC100% (r = −0.32), SC90% (r = 0.65; p = 0.058) or SCdrafting (r = 0.54). As shown in Fig. 2B, the variability in PC was larger in SC100% than in the other conditions. Despite this the PC was not significantly different at each time point in each of the three trials (Fig. 2B). There was also no significant difference in PCmean between the three conditions (93 ± 7, 96 ± 8, and 90 ± 7 in SC100% , SC90% and SCdrafting , respectively).
Physiological parameters during cycling The average VO2 during SC100% was not significantly different compared to SC90% and SCdrafting
(57 ± 5 versus 56 ± 6 and 54 ± 6 ml kg−1 min−1 , respectively, corresponding to 81 ± 7, 81 ± 9 and 79 ± 8 of VO2,max for SC100% , SC90% and SCdrafting , respectively. No significant difference for VO2 was observed at each time point during the cycling TT in the three conditions (Fig. 3A). The mean HR recorded during SC100% (177 ± 5) was significantly higher than in SCdrafting (171 ± 6) (p < 0.05). There was no significant difference in the mean HR between SC100% and SC90% . The HR in SCdrafting was significantly lower compared with SC90% and SC100% between the 6th and 13th minutes of the trial (Fig. 3B).
Blood lactate and rating of perceived exertion The BLA concentration was significantly higher at each time point (swim, cycle and post) relative to rest regardless of the SC condition (p < 0.01). The BLAswim was significantly higher in SC100% and SCdrafting compared with SC90% (p < 0.01). Aside from this result, BLAcycle and BLApost were not significantly different between each SC condition. RPEswim was significantly higher in SC100% compared in SC90% whereas no further differences in RPE were observed between the different conditions at each time point (Fig. 4B).
Swimming and subsequent cycling performance
Figure 3 The mean oxygen uptake (VO2 ) (ml kg−1 min−1 ) (A) and heart rate (HR) (b min−1 ) (B) during the cycle time trial (TT) in each swimming-cycle (SC) condition. $ Significantly (p < 0.05) different between SCdraft and SC90% ; # significantly (p < 0.05) different between SCdraft and SC100% .
Discussion This study was conducted to compare the affects of drafting or exercise at a lower relative exercise intensity in swimming on subsequent cycle TT performance in trained triathletes. The most important finding was that the power output during a 20 min TT was significantly lower after a 400 m all out freestyle swimming trial performed at maximal effort as compared to swimming at either 90% of this velocity or in a drafting situation. Importantly there was no significant difference in the power output during cycling performance after swimming at 90% or in a drafting position. The results of this study also demonstrate that all out swimming prior to cycling may affect the positive relationship between PPO and the average power output generated during a cycle TT which is typically observed when examining a cycle TT performed in isolation. Studies investigating the effects of drafting in triathlon have predominantly concerned only the cycle-run transition.4,5 It has been shown that drafting during the cycle stage of a simulated triathlon results in a significant reduction in metabolic load, leading to a higher running
239
Figure 4 A The mean (±S.D.) blood lactate (BLA) concentration (mM) (A) and rating of perceived exertion (RPE) (B) before, during and after the swimming and cycle section of each swimming-cycle condition. ** Significantly (p < 0.01) different between SC90% compared with SC100% and SCdraft .
efficiency.4,5 The reduction in cycling performance during the SC100% shown in this study is similar to that observed by previous investigators who found that drafting positively influenced the physiological responses during submaximal cycle exercise.17,18 Specifically these experiments have shown that the decrease in metabolic load when swimming in a drafting position could lead either to a higher cycling efficiency17 and induce significantly lower pedal rate or higher mean and peak resultant torques during submaximal set workload cycling.18 It is difficult to compare the results of the present study to those previous investigations17,18 as the cycle exercise performed in these experiments (10 or 15 min, set workload trial) was far different to the 20 min TT that was used in our study. In the present study, we showed that drafting during swimming could lead to an increase in performance. However the results of the present study and those of other researchers17,18 clearly show that drafting has a positive affect on cycling performance. The power output generated by elite triathletes is typically higher in the early stages
240 of the 40 km distance in order to catch up with or to break away from the lead group of athletes.10 Therefore, limiting the negative effects of swimming activity on cycling power output may be particularly important in a draft-legal event, since it would provide a means of establishing a good position and maintaining contact with the leading cyclists. Previous studies concerning the affects of swimming on cycling in triathletes have only compared the affects on cycling performance of swimming at a typical competition intensity compared to a cycling trial in isolation.13,15,16 Furthermore, the comparative affects of drafting during swimming or at a lower intensity have not been examined. Kreider et al.13 found that 800 m of sub-maximal swimming before 75 min of cycling at 75% of maximal oxygen uptake (VO2,max ) resulted in a significant reduction in power output (159 W versus 191 W) when compared to a control 75 min cycle without prior swimming. Peeling et al.14 were able to demonstrate that 20-km cycling performance was significantly lower when 750 m of swimming was performed at 90% compared to an all out trial. The results of present study confirm these findings, but also extends this data by showing that drafting or swimming at 90% of the velocity during an all out 400 m trial induces similar performance affects during a cycling TT. This data should be considered by coaches and scientists when modeling performance trials or training strategies. The reduction in Pmean during the SC100% observed in the present study could be associated with the higher BLA concentration after swimming in SC100% . However it is interesting that similar lactate levels were observed following swimming in the SCdrafting . These results are conflicting with previous studies where a lower BLA concentration was reported in the drafting condition.6,7 However in these studies the swimming velocity in the drafting and non-drafting conditions was the same. Therefore, BLA concentration was bound to be lower in a drafting situation indicative of lower exercise stress. Furthermore the ability to draft is a skill that requires practice and may result in very different responses between different athletes. It is likely that the present subjects were not skilled in this technique. In addition, since the swim trials were performed in a 25 m swimming pool, a large part of the exercise time was when each athlete would turn after each lap. Therefore it is not surprising that the decrease in BLA concentration or change in stroke pattern are lower than in studies performed in 50-m pool. Hence it is not likely that BLA concentration was not associated with the reduction in power output in the SC100% .
D.J. Bentley et al. It has been reported that drafting in swimming induces a change in stroking parameters (i.e., increased stroke length and stroke index and an unchanged stroke frequency). In addition, the kicking pattern is also modified and drafting is associated to a decrease in leg frequency.8 Although we did not control the leg kicking frequency of the subjects, one may speculate that the change in kicking pattern could induce a lower leg fatigue and explain the improved performance in the subsequent cycling performance. Indeed whilst we did not find any significant differences between conditions in the first 5—10 min, the power output appeared to be lower in this part of the cycle TT in SC100% . Hence we interpret this result as a sign of fatigue as some subjects took longer to adjust the PC and power output in SC100% compared with the other conditions. That aside the results of this study indicate that swimming at a maximal effort does indeed affect cycling performance relative to swimming at lower intensity or in a drafting situation. Furthermore short distance cycle TT performance remains unchanged after swimming at a lower intensity or in a drafting position. However as demonstrated in Fig. 2B, the variability in PC is larger in SC100% than in the other conditions. We interpret this result as a sign of fatigue since the subjects took longer to adjust their PC to the right value in SC100% than in the other conditions. The distances chosen for this study (400 m swim, 20 min cycle) were shorter than in an ‘Olympic distance’ triathlon. However research indicates that the first 400 m of the swimming stage and the first 15 km of the cycling stages are important for deciding overall race positions in an elite triathlon.10 Therefore, in this study we are confident that the trials administered were specific to elite triathletes and easier to administer in the context of the design of this study. This aside it could be that different duration and intensity swimming bouts (as in ‘sprint’, ‘Olympic’ or ‘Long Distance’ triathlon) may elicit different physiological and performance changes during a subsequent cycling task. Laursen et al.16 demonstrated that a 3000 m swimming at a self-selected intensity induced no significant differences in the average power output sustained over a 3 h cycle time trial as compared to a control conditions without prior swimming (212 W versus 222 W). Whereas other recent studies have shown higher intensity swimming bouts to affect cycling performance.14 In a longer distance triathlon, reduction in power output would reflect a residual fatigue whereas during a short-distance triathlon, decrease in power output would be linked to the physiological changes
Swimming and subsequent cycling performance directly due to the swim-cycle transition. Whyte et al.28 showed that factors such as a long duration swim prior to the cycling component, may impact upon cycle performance in ultra-endurance triathlon, but suggest that the more conditioned athletes may be better able to maintain a high cardiac output during prolonged cycling and therefore reduce the decrease in cycling performance. During a longer distance triathlon race the effects of swimming on cycling may not be as significant because the power output required in the early stages of the cycling trial might be lower as compared to a short distance triathlon. All studies conducted so far examining the swim to cycle transition have used a model where one velocity is selected in the swimming trial. Indeed the velocity during the different 400 m trials of the present study was relatively constant and not different between the conditions. However the swimming velocity (and conceivably exercise intensity) in an elite triathlon has been shown to be far greater in the early stages of the swimming stage as compared to the latter stages.10 This is a limitation of most studies examining the swimming to cycling transition in triathlon. Hence future investigations concerning swimming in triathlon should consider the specific physiological demands of this stage and replicate during experimental trials. It is of interest that the increase in cycling performance subsequent to a 400 m swim bout at submaximal intensity or in drafting condition was of ∼7% in the present study. This is in line with previous experiments where the increase in performance or efficiency was of 5—7%.14,16,17 The improvement in performance is therefore not significantly higher due to the shorter swimming bout. The PPO is a physiological variable that has been shown to be related to cycling performance in the field21—23 or in a laboratory based setting.29,30 Typically correlation coefficients of r ≥ 0.85 have been reported in the literature between PPO and the average power output during a cycling TT.19,29,30 In this study we did not find any significant correlation between PPO and Pmean during the cycle TT in each SC condition. These results suggest that whatever the swimming condition prior to a cycling performance trial, the relationship between PPO and performance is affected. Sport scientists often quantify PPO in conjunction with other physiological measures to predict performance or to determine the power output generated during a cycling trial.21,23,30 Hence the validity of using this variable for these purposes in triathletes is questioned.
241 In the present study we did not include a control cycle TT. In addition no running bout was performed after the cycling. Based on previous studies, it is likely that the power output generated in an isolated trial would have been far greater than that in the three trials that were performed in this study.13,15 Previous investigations have also included a running ‘stage’ during an experimental trial.13,14 In one study, the subjects swam 800 m, cycled for 75 min and ran 10 km, meaning that they may have adjusted their pacing strategy in cycling due to the following run.13 In a more recent investigation the affects of exercise intensity in swimming were examined in running and cycling in a sprint distance (750 m, 20 km and 5 km).14 These authors found no significant change in running performance regardless of the intensity of exercise in swimming. These findings more than likely demonstrate that running performance is a function of the residual fatigue occurring in the lower limb muscles (i.e., vastus lateralis and medialis, gastrocnemius) following cycling. Other researchers have also recommended that at least in a sprint triathlon, athletes should perform the cycle stage as maximally as possible as manipulation of the intensity in the cycle trial has no marked affects on running performance (Suriano, unpublished). Hence it is likely that the cycle TT results would have been the same if the running performance trial had also been performed after the completion of the cycle stage. That aside the primary purpose of this study was to examine any performance or physiological differences in swimcycle trials with varying swimming intensity and characteristics. We have demonstrated that whilst swimming may affect cycling performance, drafting results in a similar performance response during cycling. This has obvious implications for training of triathletes or the tactical approach to swimming in an elite triathlon event. However, the present results cannot be transposed directly to the ‘Olympic distance triathlon’ since the effects of a 400-m swim are certainly different than those of a 750 or 1500 m swim. However it is suggested that for tactical reasons, the present experimental design was more relevant to elite than to sub-elite triathletes. In summary swimming over 400 m at the maximal velocity prior cycling led to a decreased cycling performance in trained triathletes. However this result was not observed when the swim intensity was 10% lower or performed in drafting position. Regardless of what type of swimming trial that was performed prior to cycling, the relationship between PPO and cycling performance was affected. Further studies should examine the
242 interaction between swimming duration and intensity on cycling performance. Furthermore the specific physiological demands of the swimming stage of a competition should be replicated in studies examining the swimming to cycling transition in triathlon.
Practical implications • Training for triathletes should incorporate consist of specific swimming-cycling sessions incorporating swimming bouts simulating a competition intensity at 90% of a maximal velocity for a given race distance. • Specific swimming-cycling training sessions need to be completed regularly in drafting conditions. • Athletes may have to control and stabilise their pedalling cadence when performing cycling after swimming in training.
Acknowledgements The authors thank the subjects for their enthusiastic participation and their availability and Mr. Roudil, director of the Inter-universities swimming pool for his help.
References 1. Olds TS, Norton KI, Lowe EL, Olive S, Reay F, Ly S. Modeling road-cycling performance. J Appl Physiol 1995;78:1596—611. 2. Gray GL, Matheson GO, McKenzie DC. The metabolic cost of two kayaking techniques. Int J Sports Med 1995;16: 250—4. 3. Rundell KW. Effects of drafting during short-track speed skating. Med Sci Sports Exerc 1996;28:765—71. 4. Hausswirth C, Lehenaff D, Dreano P, Savonen K. Effects of cycling alone or in a sheltered position on subsequent running performance during a triathlon. Med Sci Sports Exerc 1999;31:599—604. 5. Hausswirth C, Vallier JM, Lehenaff D, Brisswalter J, Smith D, Millet G, et al. Effect of two drafting modalities in cycling on running performance. Med Sci Sports Exerc 2001;33:485—92. 6. Bassett Jr DR, Flohr J, Duey WJ, Howley ET, Pein RL. Metabolic responses to drafting during front crawl swimming. Med Sci Sports Exerc 1991;23:744—7. 7. Chatard JC, Chollet D, Millet G. Performance and drag during drafting swimming in highly trained triathletes. Med Sci Sports Exerc 1998;30:1276—80. 8. Chollet D, Hue O, Auclair F, Millet G, Chatard JC. The effects of drafting on stroking variations during swimming in elite male triathletes. Eur J Appl Physiol 2000;82: 413—7.
D.J. Bentley et al. 9. Bentley DJ, Millet GP, Vleck VE, McNaughton LR. Specific aspects of contemporary triathlon: implications for physiological analysis and performance. Sports Med 2002;32:345—59. 10. Vleck VE, Burgi A, Bentley DJ. The consequences of swim, cycle, and run performance on overall result in elite olympic distance triathlon. Int J Sports Med 2006;27: 43—8. 11. Chatard JC, Wilson B. Drafting distance in swimming. Med Sci Sports Exerc 2003;35:1176—81. 12. Millet G, Chollet D, Chatard JC. Effects of drafting behind a two- or a six-beat kick swimmer in elite female triathletes. Eur J Appl Physiol 2000;82:465—71. 13. Kreider R, Boone T, Thompson W, Burkes S, Cortes C. Cardiovascular and thermal responses of triathlon performance. Med Sci Sports Exer 1988;20:385—90. 14. Peeling PD, Bishop DJ, Landers GJ. Effect of swimming intensity on subsequent cycling and overall triathlon performance. Br J Sports Med 2005;39: 960—4. 15. Delextrat A, Brisswalter J, Hausswirth C, Bernard T, Vallier JM. Does prior 1500-m swimming affect cycling energy expenditure in well-trained triathletes? Can J Appl Physiol 2005;30:392—403. 16. Laursen PB, Rhodes EC, Langill RH. The effects of 3000-m swimming on subsequent 3-h cycling performance: implications for ultraendurance triathletes. Eur J Appl Physiol 2000;83:28—33. 17. Delextrat A, Tricot V, Bernard T, Vercruyssen F, Hausswirth C, Brisswalter J. Drafting during swimming improves efficiency during subsequent cycling. Med Sci Sports Exerc 2003;35:1612—9. 18. Delextrat A, Tricot V, Bernard T, Vercruyssen F, Hausswirth C, Brisswalter J. Modification of cycling biomechanics during a swim-to-cycle trial. J Appl Biomech 2005;21: 297—308. 19. Bentley DJ, Vleck VE, Millet GP. The isocapnic buffering phase and mechanical efficiency: relationship to cycle time trial performance of short and long duration. Can J Appl Physiol 2005;30:46—60. 20. Moseley L, Achten J, Martin JC, Jeukendrup AE. No differences in cycling efficiency between world-class and recreational cyclists. Int J Sports Med 2004;25: 374—9. 21. Bentley DJ, Wilson GJ, Davie AJ, Zhou S. Correlations between peak power output, muscular strength and cycle time trial performance in triathletes. J Sports Med Phys Fitness 1998;38:201—7. 22. Atkinson G, Reilly T. Circadian variation in sports performance. Sports Med 1996;21:292—312. 23. Hawley JA, Noakes TD. Peak power output predicts maximal oxygen uptake and performance time in trained cyclists. Eur J Appl Physiol 1992;65:79—83. 24. Balmer J, Davison RC, Coleman DA, Bird SR. The validity of power output recorded during exercise performance tests using a Kingcycle air-braked cycle ergometer when compared with an SRM powermeter. Int J Sports Med 2000;21:195—9. 25. Davison RC, Coleman D, Balmer J, Nunn M, Theakston S, Burrows M, et al. Assessment of blood lactate: practical evaluation of the Biosen 5030 lactate analyzer. Med Sci Sports Exerc 2000;32:243—7. 26. Borg G. Perceived exertion as an indicator of somatic stress. Scand J Rehabil Med 1970;2:92—8. 27. McLaughlin JE, King GA, Howley Jr DR, Ainsworth BE. Validation of the COSMED K4 b2 portable metabolic system. Int J Sports Med 2001;22:280—4.
Swimming and subsequent cycling performance 28. Whyte G, Lumley S, George K, Gates P, Sharma S, Prasad K, et al. Physiological profile and predictors of cycling performance in ultra-endurance triathletes. J Sports Med Phys Fit 2000;40:103—9. 29. Bentley DJ, McNaughton LR, Thompson D, Vleck VE, Batterham AM. Peak power output, the lactate threshold, and
243 time trial performance in cyclists. Med Sci Sports Exerc 2001;33:2077—81. 30. Balmer J, Davison RC, Bird SR. Peak power predicts performance power during an outdoor 16, 1-km cycling time trial. Med Sci Sports Exerc 2000;32:1485—90.