Effects Of Long Term

Journal of Sports Sciences, June 2007; 25(8): 927 – 935

Field and laboratory correlates of performance in competitive cross-country mountain bikers

LOUISE PRINS1,2, ELMARIE TERBLANCHE1, & KATHRYN H. MYBURGH2 1

Department of Sport Science and 2Department of Physiological Sciences, University of Stellenbosch, Stellenbosch, South Africa (Accepted 4 July 2006)

Abstract We designed a laboratory test with variable fixed intensities to simulate cross-country mountain biking and compared this to more commonly used laboratory tests and mountain bike performance. Eight competitive male mountain bikers participated in a cross-country race and subsequently did six performance tests: an individual outdoor time trial on the same course as the race and five laboratory tests. The laboratory tests were as follows: an incremental cycle test to fatigue to determine peak power output; a 26-min variable fixed-intensity protocol using an electronically braked ergometer followed immediately by a 1-km time trial using the cyclist’s own bike on an electronically braked roller ergometer; two 52-min variable fixed-intensity protocols each followed by a 1-km time trial; and a 1-km time trial done on its own. Outdoor competition time and outdoor time trial time correlated significantly (r ¼ 0.79, P 5 0.05). Both outdoor tests correlated better with peak power output relative to body mass (both r ¼ 70.83, P 5 0.05) than absolute peak power output (outdoor competition: r ¼ 70.65; outdoor time trial: r ¼ 70.66; non-significant). Outdoor performance times did not correlate with the laboratory tests. We conclude that cross-country mountain biking is similar to uphill or hilly road cycling. Further research is required to design sport-specific tests to determine the remaining unexplained variance in performance.

Keywords: Mountain bike, performance, onset of blood lactate accumulation, relative power output, time trial

Introduction Mountain biking is a comparatively new sport, with cross-country racing being officially recognized by the International Cycling Union in 1990. The first World Cup series took place in 1991 and mountain bike cross-country racing made its debut at the Olympics in 1996 in Atlanta. Despite the growing popularity of the sport, mountain bikers have not been investigated extensively by exercise physiologists. Four studies have investigated the physiological profiles of mountain bikers (Baron, 2001; Impellizzeri, Sassi, Rodriguez-Alonso, Mognoni, & Marcora, 2002; Mastroianni, Zupan, Chuba, Berger, & Wile, 2000; Wilber, Zawadzki, Kearney, Shannon, & Disalvo, 1997), while another two examined the effects of bicycle technology on physiological responses to cycling, without addressing performance (MacRae, Hise, & Allen, 2000; Seifert, Luetkemeier, Spencer, Miller, & Burke, 1997). In many ways, mountain biking is different from road cycling. Cross-country mountain bike racing

involves exercise intensities similar to those in short (540 km) road cycling time trials (Impellizzeri et al., 2002), but higher than those in longer (440 km) road cycling stages as predicted by Padilla et al. (2001). Specific physiological requirements for mountain biking may also differ from road cycling because of different riding techniques, terrain conditions, and strategies incorporated in the sport. Therefore, laboratory performance prediction tests used for road cyclists are probably not applicable to mountain bikers. A common laboratory performance test for road cyclists is the progressive incremental test to exhaustion. Maximal performance variables covered by this test include maximal oxygen consumption (V_ O2max) (Coyle, Coggan, Hopper, & Walters, 1988) and absolute sustained peak power output (Coyle et al., 1991; Hawley & Noakes, 1992), both typically achieved in the final minute of the test. Sub-maximal variables that could be associated with performance include those related to plasma lactate concentration or the dynamics of plasma lactate accumulation.

Correspondence: K. H. Myburgh, Department of Physiological Sciences, University of Stellenbosch, Private Bag XI, Matieland 7602, South Africa. E-mail: [email protected] ISSN 0264-0414 print/ISSN 1466-447X online Ó 2007 Taylor & Francis DOI: 10.1080/02640410600907938

928

L. Prins et al.

These include the lactate threshold, defined as the %V_ O2max when plasma lactate has increased by 1 mmol  l71 above the baseline value (Coyle et al., 1991), and the onset of blood lactate accumulation (Padilla, Mujika, Cuesta, & Goiriena, 1999), both of which have been shown to predict different aspects of road cycling. For example, peak power output was shown to be correlated with time trial performance (time to complete a set distance) in a simulated indoor race (Hawley & Noakes, 1992). Similarly, mean power output during a simulated indoor race correlated well with performance in an outdoor race event (Coyle et al., 1991). In the latter study, outdoor race performance was correlated with V_ O2 at the lactate threshold determined from an incremental test to exhaustion, and mean absolute V_ O2 during a simulated indoor 1-h performance test, but not mean V_ O2 expressed relative to body mass. However, cross-country mountain biking differs substantially from road cycling and the best laboratory predictor of uphill cycling over either 1 km or 6 km (on a treadmill with a gradient of 12% and 6% respectively) was shown to be the mean power produced during a Wingate test expressed relative to body mass (Davison, Swan, Coleman, & Bird, 2000). These contrasting results suggest that the requirements for success in flat and hilly cycling are completely different. In contrast to road cycling (Coyle et al., 1988, 1991; Mujika & Padilla, 2001; Padilla et al., 1999; Schabort, Hawley, Hopkins, Mujika, & Noakes, 1998), no cycling studies that have used laboratorybased tests to predict performance for mountain biking, or to design a sport-specific laboratory test for mountain bikers. One option would be to take account of the variability in terrain and thus the variability in intensity in a cross-country mountain bike race, and then design a test with variable intensities to reflect these characteristics. For such a test to be reproducible between participants but also variable, the intensity and duration at the various intensities should be fixed rather than freely selected such as during free-range exercise (Schabort et al., 1998; Terblanche, Wessels, Stewart, & Koeslag, 1999). Exercise tests with fixed but variable intensities are rare in the literature (Palmer, Borghouts, Noakes, & Hawley, 1999; Palmer, Noakes, & Hawley, 1997). This could be because they need to be performed in combination with a time trial to assess performance. The latter could be the time taken to complete a specific distance (the approach used in the present study), or a specific amount of external mechanical work that can be completed in a set time (as used by Coyle et al., 1991). The aims of this study were: (1) to design and evaluate a laboratory test to simulate cross-country

mountain bike performance; (2) to compare the predictive power of variables from commonly used laboratory tests as well as non-traditional tests; and (3) to determine the most relevant mountain bike performance prediction test(s).

Methods Participants Eight competitive male cross-country mountain bikers participated in the study. All participants had had to have competed in mountain bike races for at least two consecutive seasons. After being fully informed of the risks associated with the study, the participants provided their written consent. The study was approved by the ethics committee of the authors’ institution. The eight participants completed all of the tests. Test procedures: Overview All participants took part in an outdoor competition and then completed four laboratory tests and an outdoor time trial within 2 months of the competition. One of the laboratory tests was done on two occasions to determine reproducibility. All tests were completed in the winter season in temperate conditions with low humidity. Laboratory testing was performed during five separate visits, with a maximum of 7 days of recovery between tests. All participants completed the following laboratory tests in random order: 1.

2.

3.

4.

A progressive incremental exercise test to exhaustion to determine V_ O2max and peak power output. A 1-km time trial performed in a fresh condition (no previous laboratory test on the same day). A variable fixed-intensity bout lasting 26 min (approximately equivalent to one lap of the original race) and followed immediately by a 1-km time trial. A variable fixed-intensity test consisting of two consecutive bouts each lasting 26 min and followed immediately by a 1-km time trial.

The latter was repeated on a separate day to determine the reproducibility of the variable fixedintensity protocol. Field testing took place on two occasions. On the first occasion, all riders participated in the same outdoor competition. On the second occasion, they performed an individual outdoor time trial over the same course.

Mountain bike performance Progressive incremental test to exhaustion All participants completed a progressive incremental test to fatigue on a calibrated cycle ergometer (Technogym Bikerace, Gambettola, Italy) for the determination of V_ O2max, peak power output, and the onset of blood lactate accumulation. After a 10-min warm-up, the test began at an intensity of 3.33 W  kg71 body mass. Every 2½ min, the intensity was increased by 30 W until the participant reached exhaustion (modified from Kuipers, Verstappen, Keizer, & Guerten, 1985). The criteria for ending the test included one or a combination of the following: a heart rate greater than 90% of the age-predicted maximum heart rate, a respiratory exchange ratio (RER) greater than 1.1, or a plateau in oxygen consumption (5150 ml  min71 difference in oxygen consumption for the final two stages). During the test, the participants wore a mask that covered the nose and mouth and expired air passed through an on-line computer system attached to an automated gas analyser (Oxycon Version 4.5, Jaeger, Hoechberg, Germany). Before each test, the gas analyser was calibrated with room air and a carbon dioxide – oxygen – nitrogen gas mixture of known composition. Analyser outputs were processed by the computer, which calculated oxygen uptake, carbon dioxide production, minute ventilation, and the RER for every 10 s of the test. Each participant’s V_ O2max was taken as the highest oxygen uptake measured during any 10-s period of the test. Peak power output was calculated as follows (Kuipers et al., 1985): peak power output ¼ Wf þ ðt=150  30Þ where Wf is the final completed intensity and t is the time in seconds of the final uncompleted workload. Heart rate was recorded every 5 s using a downloadable heart rate monitor (Accurex Plus, Polar Electro, Kempele, Finland). Each participant’s maximal heart rate was taken as the highest heart rate measured during any 10-s period of the test. Blood samples were obtained from an indwelling cannula (JelcoTM IV Catheter, Brussels, Belgium) in the participant’s left forearm vein. Blood samples (5 ml) were taken at rest, 15 s before the end of each intensity, and at 2 and 4 min into recovery. Samples were immediately centrifuged at 48C at 3000 rev  min71 and plasma was frozen for later analysis. Plasma lactate concentration was later determined using an electroenzymatic technique with an automatic analyser (YSI1 1500 Sport, Yellow Springs Instruments, Yellow Springs, OH). Following the recommendations of the manufacturer, the analyser was calibrated before each batch with standard solutions of known lactate concentrations

929

(0 and 5 mmol  l71). The exercise intensity corresponding to the onset of blood lactate accumulation was identified on the lactate – power output curve by straight line interpolation between the two closest points eliciting a blood lactate concentration of 4 mmol  l71 (Sjo¨din & Jacobs, 1981). Percent maximal heart rate (%HRmax) was plotted against peak power output (PPO) for each participant. Each participant’s individual graph was used to calculate the %PPO for a specific %HRmax value to determine the variable fixed-intensity protocol, which aimed to simulate one lap of the outdoor time trial (see below for more details). Field tests Outdoor competition. All participants competed in the same regional cross-country championship race. Approximately 100 mountain bikers competed in the event. The course was approximately 8 km long and, depending on fitness, experience, and age, the riders had to complete either four or six laps of the course during competition. For the purpose of this study, the race time for the first four laps was used as a performance measure. Outdoor time trial. Each participant had individually to perform an outdoor time trial of four laps on the same course as used for the outdoor competition. They were requested to go ‘‘all out’’ as if they were competing in a race. Each participant wore a downloadable heart rate monitor (Accurex Plus, Polar Electro, Kempele, Finland) to record their heart rate and to determine when they finished a lap and the complete time trial. Total outdoor time trial time was also used as a performance measure. Variable fixed-intensity bouts After the first four participants completed the individual outdoor time trial, their heart rate data were used to design a variable fixed-intensity laboratory test. We decided to only use the heart rate data of the second lap to avoid the initial stress caused by the start of the outdoor time trial and the possible effects of cardiac drift (Jeukendrup & Van Diemen, 1998) during the last two laps. Heart rate (expressed as a percentage of each individual’s maximal heart rate obtained in the progressive incremental test to exhaustion) was then plotted against time standardized to 100% (every 5-s interval was expressed as a percentage of that person’s total time to complete the second lap of the outdoor time trial). The data for the four participants were then superimposed onto one graph and through inspection 15 stages with different durations and %HRmax values were identified as being typical for all four

930

L. Prins et al.

participants. The %HRmax – PPO graphs that were constructed after the peak power output test were then used to determine each individual’s corresponding %PPO for each of the 15 stages. The mean %PPO for the four participants was then calculated for each of the 15 stages. As one participant completed the time trial with great difficulty, we decided to reduce the intensities of the recovery stages to create ‘‘resting periods’’ more similar to those a mountain biker would experience when going downhill, but which were not reflected immediately in the participant’s heart rate. The average time it took for the participants to complete the second lap of the outdoor time trial was 26 min and that was taken as the duration of the one-lap variable fixed-intensity protocol (Figure 1). All participants performed one variable fixedintensity bout simulating one lap and two variable fixed-intensity bouts simulating two laps of the outdoor time trial on a calibrated cycle ergometer (Technogym Bikerace, Gambettola, Italy), relative to each individual’s peak power output. Heart rate was recorded using a downloadable heart rate monitor (Accurex Plus, Polar Electro, Kempele, Finland). The average %HRmax for each stage was then plotted against time. Each variable fixed-intensity bout was followed by a sprint time trial performance test (see below for details). Laboratory time trials Each participant completed a total of four 1-km time trials on separate occasions on a calibrated SpinTrainer (Techonogym, Gambettola, Italy) on which the participant’s own mountain bike was mounted.

Figure 1. One lap of the simulated cross-country mountain bike test consisting of a variable fixed-intensity protocol standardized for each individual. PPO ¼ peak power output.

As previously described (Myburgh, Viljoen, & Tereblanche, 2001), the Spin-Trainer was set to difficulty level 8 (a standardized electromagnetic resistance), calibrated for each person’s body weight, and the wheel pressure was standardized at 70 psi for all participants. The test – retest reproducibility for a 20-km time trial for 12 individuals who were previously accustomed to the equipment was 0.9% (coefficient of variation) in our laboratory (unpublished data). For 13 other individuals who were just accustomed to the equipment, the coefficient of variation for a 5-km time trial was 0.6% (unpublished data). For the latter participants, the correlation between peak power output and 5-km time trial time was r ¼ 70.66 (unpublished data). One time trial was undertaken on its own (TT0) after a 5-min warm-up. The other three time trials were done within 30 s of finishing either the one- (TT1) or the two-lap (TT2) variable fixed-intensity protocols. The participants were requested to ride ‘‘as fast as possible’’ and the times they achieved were used as a performance measure. A 1-km time trial was chosen because mountain biking seldom involves long stretches of sprints during cross-country competition. The TT0 time trial would be representative of a sprint at the start of the race to obtain a favourable racing position, while TT1 and TT2 would be representative of either sprints to get past other riders on the single track or sprints to the finish. Data analysis For descriptive between-participant comparisons and within-participant comparisons, data are presented as the mean and standard deviation (s). Paired t-tests were performed to compare the performance times of the outdoor competition and outdoor time trials. One-way analysis of variance (ANOVA) with repeated measures was used to compare the mean heart rates for the different laps of the outdoor time trials and to compare the mean times for the different laps of the outdoor competition. Paired t-tests were performed to compare the mean heart rates of the first laps of the different variable fixed-intensity bouts. Thereafter, we compared the mean heart rates of the second laps of the outdoor time trials with the second lap of the second two-lap variable fixed-intensity protocol, also using a paired t-test. Significant differences revealed by the ANOVA were further analysed using Tukey (HSD) post-hoc analysis. Pearson’s correlation coefficients were calculated to investigate the association between the different time trials, race times, and metabolic variables for the total group. Then, 95% confidence limits for the correlation coefficients (r) were calculated using the methods of Hopkins (2000). The coefficient of

Mountain bike performance variation (CV) was used to estimate the reproducibility of the variable fixed-intensity bouts. The coefficient of variation was calculated using a twoway ANOVA on the natural logarithm of the variable, then transforming the within-participant standard deviation using the following formula: CV ¼ 100(es 7 1), where e is the base of natural logarithms (Schabort et al., 1998). Statistical significance was set at P 5 0.05.

Results Characteristics of the participants and outdoor performance Physiological responses at maximal exercise and outdoor performance times for the eight participants are reported in Table I. There were no significant differences in lap times for laps 1 to 4 (26:52+1:38, 26:19+1:57, 26:59+1:29, and 26:53+1:39 min:s, respectively; P 4 0.05) of the outdoor competition. Both maximal RER and maximal blood lactate concentration indicate that maximal efforts were achieved. No difference was found between outdoor competition and outdoor time trial performance Table I. Participant (n ¼ 8) characteristics and outdoor performances.

Age (years) Body mass (kg) Maximal heart rate (beats  min71) V_ O2max (l  min71) V_ O2max (ml  kg71  min71) Peak power output (W) Peak power output (W  kg71) [La]max (mmol  l71) Maximal RER Outdoor competition time (min:s) Outdoor time trial time (min:s)

Mean + s

Range

28 + 5 72.9 + 5.6 189 + 5

20 – 35 66 – 81 183 – 196

4.65 + 0.64 63.6 + 5.7 372 + 37 5.1 + 0.4 11.9 + 4.5 1.22 + 0.05 106:47 + 6:43

3.70 – 5.68 56.1 – 72.8 314 – 446 4.5 – 5.7 5.2 – 17.0 1.15 – 1.31 99:06 – 120:49

109:00 + 4:41

99:46 – 116:00

Note: [La]max, maximal blood lactate concentration.

931

(P 4 0.05) and they also correlated (r ¼ 0.79; P 5 0.05). The onset of blood lactate accumulation occurred at a power output of 289+60 W, at a power output relative to body mass of 4.0+0.7 W  kg71, at 78.7+9.0% V_ O2max, and at 85.9+6.6% maximal heart rate. Only peak power output relative to body mass correlated with both outdoor competition and outdoor time trial time (Table II). Power output relative to body mass at OBLA correlated with outdoor time trial time only (Table II). Other variables that did not correlate with outdoor competition were age (r ¼ 0.12), body mass (r ¼ 0.07), absolute V_ O2max (r ¼ 70.35), relative V_ O2max (r ¼ 70.59), absolute power output at the onset of lactate accumulation (r ¼ 70.59), and peak power output (r ¼ 70.65). Similarly, none of the above variables correlated with the outdoor time trial time: age (r ¼ 0.24), body mass (r ¼ 0.07), absolute V_ O2max (r ¼ 70.34), relative V_ O2max (r ¼ 70.55), absolute power output at the onset of lactate accumulation (r ¼ 70.66), and peak power output (r ¼ 70.67). 1-km time trials The reproducibility of both 1-km time trial tests after the two-lap variable fixed-intensity protocol was good (CV ¼ 3.3%). For further analysis, the best of the two attempts was used. The group means for 1-km time trial performances for TT0, TT1, and the best of the two TT2 trials (hereafter called TT2B) are shown in Figure 2. There were significant differences between the group means for performance time for the TT0 (83.4+5.2 s) and both TT1 (88.3+5.7 s) and TT2 B (87.4+5.3 s) (P 5 0.05). There was no statistically significant difference between TT1 and TT2B. There was no correlation between any 1-km time trial performance time and either outdoor competition or outdoor times times, although better r-values were seen for TT2 and the best of the two TT2 trials (i.e. TT2B) (Table III). In addition, we determined whether there was a relationship between either outdoor competition or outdoor time trial time and the decrement in 1-km

Table II. The relationship between outdoor performances and different laboratory power output variables (95% confidence limits in parentheses). Outdoor competition time (min:s) PPO (W) PPO (W  kg71) PO at OBLA (W) PO at OBLA (W  kg71)

r ¼ 70.65 r ¼ 70.83 r ¼ 70.56 r ¼ 70.64

(70.93 (70.97 (70.91 (70.93

to to to to

0.10) 70.30) 0.24) 0.12)

P-value

Outdoor time trial time (min:s)

P-value

N.S. 50.05 N.S. N.S.

r ¼ 70.66 r ¼ 70.83 r ¼ 70.67 r ¼ 70.74

N.S. 50.05 N.S. 50.05

(70.93 (70.97 (70.93 (70.95

to to to to

0.08) 70.30) 0.07) 70.07)

Note: PPO ¼ peak power output; OBLA ¼ onset of blood lactate accumulation; PO at OBLA ¼ power output at a blood lactate concentration of 4 mmol  l71; N.S. ¼ non-significant.

932

L. Prins et al. (87.9+6.1% vs. 89.3+4.3%; P 4 0.05). The clearly demarcated differences in exercise intensity for the stages of the variable fixed-intensity protocol (Figure 1) are not reflected by the group mean %HRmax profiles for the variable fixed-intensity protocol or the outdoor time trial (Figure 3). However, the reproducibility of %HRmax between the first and second two-lap variable fixed-intensity protocol was estimated by comparing the coefficient of variation for heart rate at each stage. The coefficient of variation for all stages ranged from 1.0 to 4.1 (mean CV for all stages was 2.0).

time trial performance (DTT) when comparing either TT1 or TT2B with TT0 (e.g. TT1 7 TT0 ¼ DTT1). No significant relationships were observed between outdoor competition performance and DTT1 (r ¼ 0.43, non-significant) or DTT2B (r ¼ 0.38, non-significant). Even poorer r-values were seen for the relationship between outdoor time trial time and DTT1 (r ¼ 0.04, non-significant) or DTT2B (r ¼ 0.29, non-significant). Heart rate monitoring The mean exercise intensity (expressed as a percentage of HRmax) for the outdoor time trial was 88.4+0.6%. There were no significant differences in mean heart rate (expressed as a percentage of HRmax) for laps 1 to 4 of the outdoor time trial (88.3+6.0%, 89.3+4.3%, 88.3+3.8%, and 87.8+5.1% respectively; P 4 0.05). Figure 3 shows that there was no significant difference in mean heart rate (expressed as a percentage of HRmax) between the second lap of the second two-lap variable fixed-intensity protocol and the second lap of the outdoor time trial

Discussion In this study, mountain bikers took part in a competitive race and subsequently also completed an outdoor time trial over the same course, as well as other laboratory tests. The main finding of the study was that both outdoor competition and outdoor time trial performances were better related to peak power output in relation to body mass than to absolute peak power output or any other physiological variable, or performance in laboratory tests. Participants The mountain bikers in our study had lower absolute (4.65 vs. 4.86 l  min71) and relative (63.6 vs. 75.2 ml  kg71  min71) V_ O2max values than international high-standard mountain bikers (Impellizzeri et al., 2002), but better relative V_ O2max values than recreational off-road cyclists (52.7 ml  kg71  min71) (Mastroianni et al., 2000). Their maximal power (372 W and 5.1 W  kg71) was also somewhat lower compared with cyclists representing the US National Off-Road Bicycle Association’s (NORBA) cross-country team (420 W and 5.9 W  kg71) (Wilber et al., 1997). Several of the mountain bikers in our study only competed in regional mountain bike competitions, while the NORBA and high-standard mountain bikers were internationally competitive cyclists and therefore we assume, better trained. Nevertheless, the above data indicate that our participants were competitive athletes rather than recreationally active cyclists.

Figure 2. Performance times for 1-km time trials presented as group means (+s) for time trials done on their own (TT0), after one (TT1) or after two laps of the variable fixed-intensity protocol. Since the 1-km time trial after two laps of the protocol was performed on two occasions, the best performance is presented here (TT2B). Asterisks denote difference between time trials (P 5 0.05).

Table III. The relationships between outdoor performances and performance times for the novel laboratory tests (95% confidence limits in parentheses).

TT0 TT1 TT2B

Outdoor competition time (min:s)

P-value

Outdoor time trial time (min:s)

P-value

r ¼ 0.29 (70.52 to 0.83) r ¼ 0.53 (70.28 to 0.90) r ¼ 0.59 (70.20 to 0.91)

N. S .

r ¼ 0.24 (70.56 to 0.81) r ¼ 0.25 (70.55 to 0.81) r ¼ 0.46 (70.36 to 0.88)

N.S.

N. S . N. S .

N.S. N.S.

Note: TT0 ¼ 1-km time trial from rested; TT1 ¼ 1-km time trial after one lap of the variable fixed-intensity protocol; TT2B ¼ 1-km time trial after two laps of the variable fixed-intensity protocol (best of two tests); N.S. ¼ non-significant.

Mountain bike performance

Figure 3. Mean %HRmax for the second lap of the second two-lap variable fixed-intensity protocol (~) and the second lap of the outdoor time trial (¤).

Outdoor competition vs. outdoor time trial performance The outdoor competition and outdoor time trial were both completed over four laps of the same outdoor course. The only difference was that one was completed in a race with a mass start and the other as an individual time trial. Despite the lack of competition, there were no differences in the performances of mountain bikers during the outdoor competition and outdoor time trial (P 4 0.05). Three participants performed better during the competition than the outdoor time trial, while the other five performed better in the outdoor time trial. In mountain biking it is very important to obtain a good position at the start of the race, since there are usually few chances of passing other riders, especially if the course consists of mainly ‘‘single-track’’. Other factors that could prevent passing are the steepness of the hills and hazards such as rocks embedded in the track. Performing well in a race is therefore dependent on fitness, technique, and clever strategies to overcome these difficulties. These factors could explain why the relatively good relationship (r ¼ 0.79; P 5 0.05) between the performance times of the two outdoor tests in our current study was not even better. Maximal exercise and other participant characteristics For prediction of performance, physiological variables can be expressed both in absolute terms and relative to anthropometric variables. For example, body mass and frontal area influence gravity-dependent or aerodynamic resistance (Padilla et al., 1999). Maximal oxygen uptake is commonly expressed relative to body mass, but many authors have chosen to report absolute values (Febbraio & Koukoulas, 2000; Palmer et al., 1999; Van Loon, Jeukendrup,

933

Saris, & Wagenmakers, 1999). Not all of our participants fitted the criteria for national elite cyclists and it was therefore not unexpected that varying abilities were observed. This was clearly evident in the wide ranges of V_ O2max, age, and body mass (Table I). Although V_ O2max is traditionally regarded as an important determinant of endurance performance (Coyle et al., 1988), neither absolute nor relative V_ O2max (r ¼ 70.35 and r ¼ 70.59 respectively) correlated with outdoor performance times. It has been shown that the absolute peak power output obtained during a maximal incremental cycle test can be used as a predictor of performance in endurance road cyclists (Coyle et al., 1991; Schabort, Killian, St. Clair Gibson, Hawley, & Noakes, 2000) and that it is in fact a better predictor than V_ O2max (Hawley & Noakes, 1992). Schabort et al. (2000) reported a high correlation (r ¼ 70.91; P 5 0.01) between maximum power output during a progressive incremental cycling test to exhaustion and performance time for 40-km cycling (n ¼ 10) during National Triathlon Championships. In the present study, the relationship between peak power output and both outdoor competition and outdoor time trial performance (r ¼ 70.65 and r ¼ 70.66 respectively) was not significant (P 4 0.05). On the other hand, Davison et al. (2000) found that the best predictor of hill climbing performance in road cycling was the mean power per unit of body mass as tested during the Wingate test (r ¼ 70.90). Padilla et al. (1999) also found that uphill road cycling specialists competing in the Tour de France had the highest power output relative to body mass (tested during an incremental test) when compared with cyclists who specialized in other areas, such as flat terrain or time trials. Our results support these previous findings, as peak power output relative to body mass (obtained during an incremental test) correlated better with both outdoor performances (r ¼ 70.83; P 5 0.05) than absolute peak power, despite a relatively small sample size in our study. This could indicate that mountain biking can be compared with uphill or hilly road cycling, but to a lesser extent with time trialling. Indeed, even outdoor time trialling on the mountain bike course did not correlate better with outdoor competition than did peak power output relative to body mass. Higher correlations were obtained when power output at the onset of blood lactate accumulation was expressed relative to body mass (r ¼ 70.74 for the outdoor time trial; P 5 0.05). This again stresses the fact that mountain biking can be classified in the same way as road cycling in hilly terrain in terms of the physiological requirements for good performance. This finding could be important for cyclists are yet to choose between road and off-road racing as

934

L. Prins et al.

their specialty. If a cyclist has a particularly good power to weight ratio, this advantage might be more clearly seen in mountain bike races. A finding that is more difficult to explain is the non-significant correlation between power output relative to body mass at the onset of blood lactate accumulation and outdoor competition performance (r ¼ 70.64). This could have arisen because competition performance (as opposed to individual performance over the same course) is influenced by the other competitors and tactical decisions related to the other competitors, which might be different from the tactical decisions made when cycling alone. Less traditional testing Laboratory-based testing in this study included a variety of 1-km time trials (in the rested condition and after variable fixed intensities equivalent to one or two laps of the outdoor time trial course). A 1-km time trial takes 80 to 95 s to complete. Therefore, it is longer than a Wingate test, but a lot shorter than other tests such as the incremental test to fatigue. No correlation or trend was observed between the sprint done on its own and outdoor competition performance. A difference was observed between TT0 and both TT1 and TT2B (P 5 0.05). Sprinting in a race is never done on its own, but always in conjunction with various other exercise intensities. Therefore, a sprint following varying-intensity bouts lasting about an hour could be a better representation of a mountain biker’s racing ability. The correlation between the sprint done after the twolap variable fixed-intensity protocol and outdoor competition performance, although not significant (r ¼ 0.59), was higher. It is possible that it could be a Type II error (see confidence limits in Table III), so with more participants it could be a meaningful performance test. Sprint cycling has a high anaerobic component to ATP provision and promotes rapid muscular fatigue, as evidenced by, for example, typical power decrements seen in the 30-s Wingate test. We also used the decrement in 1-km time trial performance between those tested in the rested condition and those tests done after the variable fixed-intensity tests to assess fatigability in our cyclists. Individual fatigability was, however, also not related to outdoor competition or outdoor time trial performances. In contrast to our study, 1-km as well as 4-km sprints were used by Schabort et al. (1998). They designed a laboratory test of cycling performance incorporating a series of sprints in a 100-km time trial to evaluate its reproducibility. One of their conclusions was that laboratory protocols in which participants are allowed freely to choose their effort might be more reliable than constant-load exercise tests.

Evaluating the variable fixed-intensity protocol Individual heart rates (expressed as a percentage of HRmax) between the two simulated two-lap variable fixed-intensity protocols were reproducible when the coefficient of variation was calculated for each stage. The mean coefficient of variation of all the stages was 2.0%. This indicates that individual variation from day to day was minimal. The heart rate response during the outdoor time trial was related to course profile, which is in line with Ferna´ndez-Garcı´a and colleagues (Ferna´ndezGarcı´a, Terrados, Pe´rez-Landaluce, & Rodrı´guezAlonso, 2000), who recorded heart rate during stages of the Tour de France and the Vuelta e Espan˜a (albeit both road stage races). The mean heart rate (88.4% HRmax) we observed for the outdoor time trial was comparable to the mean heart rate (90.0% HRmax) of four international cross-country mountain bike competitions (Impellizzeri et al., 2002). No difference was found between the %HRmax values for laps one to four of the outdoor time trial. Mean %HRmax of lap four was slightly lower than for the other laps, which could be due to slower completion time of that lap and the effect of fatigue setting in. Within each lap, %HRmax typically varied between 80 and 95% in both the outdoor time trial and the indoor variable fixed-intensity protocol (see Figure 3 for a representative example of one lap), indicating that the intensity did vary quite substantially from the mean of 88%. The ideal way of designing a simulated variable fixed-intensity laboratory test would be to base the intensities on power output variations measured in the field (and the duration at each specific power output). Although this technology is not yet widely used, it will benefit future research. Nevertheless, heart rate during the variable fixed-intensity protocol followed closely the same pattern as during the outdoor time trial (Figure 3), despite more clearly demarcated and abrupt changes in exercise intensity for the variable fixed-intensity protocol (Figure 1). This could be explained by the relatively slow rate of recovery for heart rate when exercise intensity was reduced. Therefore, our conclusions would correspond with those of Gilman (1996), that heart rate does not always reflect energy utilization. Factors other than exercise intensity that could influence heart rate include cardiac drift, dehydration, environmental factors, and competition stress. Also, although power output can drop dramatically in a short time (e.g. when a steep downhill follows quickly after a steep uphill), the heart rate response to the easier workload of downhill cycling may take some time to resolve due to the oxidative requirements of recovery from anaerobic work.

Mountain bike performance Conclusions The main finding of this study was that both outdoor competition and outdoor time trial performances were better related to peak power output in relation to body mass than to absolute peak power output (Table II). This indicates that mountain biking can be placed in the same category as uphill or hilly road cycling. Also, peak power output in relation to body mass correlated as well with outdoor competition performance as with individual outdoor time trial performance. This may indicate that the typical incremental test to fatigue used for many years by sport scientists is sufficient to predict mountain bike performance and there is no need for a more sportspecific laboratory test. On the one hand, the results of our study did not support our hypothesis that parameters from our specific variable fixed-intensity protocol would correlate better with outdoor performances than parameters from an incremental test to exhaustion. On the other hand, peak power output per kilogram only explained about 70% of the variance in mountain biking race performance in our participants. Therefore, future studies should design alternative performance-related tests that will benefit our ability to predict mountain bike race performance. We suggest a shift in the focus of researchers studying the physiology of mountain biking towards the design of less traditional, and more sport-specific, laboratory tests.

References Baron, R. (2001). Aerobic and anaerobic power characteristics of off-road cyclists. Medicine and Science in Sports and Exercise, 33, 1387 – 1393. Coyle, E. F., Coggan, A. R., Hopper, M. K., & Walters, T. J. (1988). Determinants of endurance in well-trained cyclists. Journal of Applied Physiology, 64, 2622 – 2630. Coyle, E. F., Feltner, M. E., Kautz, S. A., Hamilton, M. T., Montain, S. J., Baylor, A. M. et al. (1991). Physiological and biomechanical factors associated with elite endurance cycling performance. Medicine and Science in Sports and Exercise, 23, 93 – 107. Davison, R. C., Swan, D., Coleman, D., & Bird, S. (2000). Correlates of simulated hill climb performance. Journal of Sports Sciences, 18, 105 – 110. Febbraio, M. A., & Koukoulas, I. (2000). HSP72 gene expression progressively increases in human skeletal muscle during prolonged, exhaustive exercise. Journal of Applied Physiology, 89, 1055 – 1060. Ferna´ndez-Garcı´a, B., Terrados, N., Pe´rez-Landaluce, J., & Rodrı´guez-Alonso, M. (2000). Intensity of exercise during road race pro-cycling competition. Medicine and Science in Sports and Exercise, 32, 1002 – 1006. Gilman, M. B. (1996). The use of heart rate to monitor the intensity of endurance training. Sports Medicine, 21, 73 – 79. Hawley, J. A., & Noakes, T. D. (1992). Peak sustained power output predicts V_ O2max and performance in trained cyclists. European Journal of Applied Physiology, 65, 79 – 83.

935

Hopkins, W. G. (2000). A new view of statistics. Internet Society for Sport Science (available at: http://www.sportsci.org/resource/stats/). Impellizzeri, F., Sassi, A., Rodriguez-Alonso, M., Mognoni, P., & Marcora, S. (2002). Exercise intensity during off-road cycling competitions. Medicine and Science in Sports and Exercise, 34, 1808 – 1813. Jeukendrup, A., & Van Diemen, A. (1998). Heart rate monitoring during training and competition in cyclists. Journal of Sports Sciences, 16, S91 – S99. Kuipers, H., Verstappen, F. T. J., Keizer, H. A., & Guerten, P. (1985). Variability of aerobic performance in the laboratory and its physiological correlates. International Journal of Sports Medicine, 6, 197 – 201. MacRae, H. S., Hise, K. J., & Allen, P. J. (2000). Effects of front and dual suspension mountain bike systems on uphill cycling performance. Medicine and Science in Sports and Exercise, 32, 1276 – 1280. Mastroianni, G. R., Zupan, M. F., Chuba, D. M., Berger, R. C., & Wile, A. L. (2000). Voluntary pacing and energy cost of offroad cycling and running. Applied Ergonomics, 31, 479 – 485. Mujika, I., & Padilla, S. (2001). Physiological and performance characteristics of male professional road cyclists. Sports Medicine, 31, 479 – 487. Myburgh, K. H., Viljoen, A., & Tereblanche, S. (2001). Plasma lactate concentrations for self-selected maximal effort lasting 1 h. Medicine and Science in Sports and Exercise, 33, 152 – 156. Padilla, S., Mujika, I., Cuesta, G., & Goiriena, J. J. (1999). Level ground and uphill cycling ability in professional road cycling. Medicine and Science in Sports and Exercise, 31, 878 – 885. Padilla, S., Mujika, I., Orbananos, J., Santisteban, J., Angulo, F., & Goiriena, J. J. (2001). Exercise intensity and load during mass-start stage races in professional road cycling. Medicine and Science in Sports and Exercise, 33, 796 – 802. Palmer, G. S., Borghouts, L. B., Noakes, T. D., & Hawley, J. A. (1999). Metabolic performance responses to constant-load vs. variable-intensity exercise in trained cyclists. Journal of Applied Physiology, 87, 1186 – 1196. Palmer, G. S., Noakes, T. D., & Hawley, J. A. (1997). Effects of steady-state versus stochastic exercise on subsequent cycling performance. Medicine and Science in Sports and Exercise, 29, 684 – 687. Schabort, E. J., Hawley, J. A., Hopkins, W. G., Mujika, I., & Noakes, T. D. (1998). A new reliable laboratory test of endurance performance for road cyclists. Medicine and Science in Sports and Exercise, 20, 1744 – 1750. Schabort, E. J., Killian, S. C., St. Clair Gibson, A., Hawley, J. A., & Noakes, T. D. (2000). Prediction of triathlon race time from laboratory testing in national triathletes. Medicine and Science in Sports and Exercise, 32, 844 – 849. Seifert, J. G., Luetkemeier, M. J., Spencer, M. K., Miller, D., & Burke, E. R. (1997). The effects of mountain bike suspension systems on energy expenditure, physical exertion, and time trial performance during mountain bicycling. International Journal of Sports Medicine, 18, 197 – 200. Sjo¨din, B., & Jacobs, I. (1981). Onset of blood lactate accumulation and marathon running performance. International Journal of Sports Medicine, 2, 23 – 26. Terblanche, E., Wessels, J. A., Stewart, R. I., & Koeslag, J. H. (1999). A computer simulation of free-range exercise in the laboratory. Journal of Applied Physiology, 87, 1386 – 1391. Van Loon, L. U. C., Jeukendrup, A. E., Saris, W. H. M., & Wagenmakers, A. J. M. (1999). Effect of training status on fuel selection during submaximal exercise with glucose ingestion. Journal of Applied Physiology, 87, 1413 – 1420. Wilber, R. L., Zawadzki, K. M., Kearney, J. T., Shannon, M. P., & Disalvo, D. (1997). Physiological profiles of elite off-road and road cyclists. Medicine and Science in Sports and Exercise, 29, 1090 – 1094.