Confidence in the reliability and validity of measurement data gained from the laboratory testing of athletes is crucial if coaches are to develop sustained improvements in athletic performance. In sports such as elite-level cycling, it is well known that the difference between first and second place can be marginal (15), and therefore any objectively collected laboratory data need to accurately reflect the demands of the race environment.
In cycle sports, including bicycle motocross (BMX), the availability of reliable field-based testing equipment such as Schoberer Rad Messtechnik (SRM) power meters (25), timing gates (24), and mobile metabolic systems (8) has led some researchers to question the exclusive use of laboratory-based testing and have argued that field testing may be a more relevant method of analysis (11).
For the past 2 decades, cycling studies have demonstrated that a strong relationship exists between performance data obtained both in the laboratory and in the field (2,13,19). However, despite the ecological validity of laboratory-based testing acknowledged within cycling as a whole, there still remains some dispute between individual cycling disciplines (9,20,22).
For instance, Jobson et al. (17) investigated the ecological validity of laboratory testing in 23 competitive road cyclists. Velocity was analyzed in two 25-mile time trials, one completed on a road course the other on a king cycle ergometer. The results revealed a 4% lower velocity recording on the road when compared with the laboratory test, which the authors found to be statistically significant. Interestingly, this 4% difference only occurred within riders who had a greater frontal surface area and it was concluded that the environmental factor of drag coefficients had an influence on the final test data.
Moreover, research conducted to examine the influence of drag between sprints performed in a laboratory and those of riders in an indoor gymnasium reported a 4% decrease in peak power between the 2 testing environments (4). The research also compared the differences in seated and standing sprints of both environments. Bertucci et al. (4) established a 32% higher force during standing sprints in the field compared with the laboratory environment. The authors concluded that the large increases in force and peak power were because of actual cycling locomotion, which enabled natural medial and lateral oscillations of the bike in field testing.
In contrast, Gardner et al. (10) analyzed the relationship between laboratory and field results in 7 international track sprint riders. The riders performed 2 maximal 6-second cycle ergometer sprints and two 65-m standing starts on a standard track bike. No statistical differences were reported between maximum torque, maximum power, and optimal pedaling rate in the laboratory vs. field testing. These findings led the authors to conclude that as the laboratory data and field testing data presented no statistical differences, they could be used interchangeably.
More recently, Karsten et al. (18) conducted a laboratory vs. field study to ascertain the validity of field testing power output. Karsten et al. (18) analyzed 14 trained cyclists on a laboratory cycle ergometer and on an indoor velodrome. This study reported a high level of agreement (10.98–10.8 W) between the power outputs in the field (234 ± 24 W) when compared with findings from the laboratory (234 ± 25 W). These findings further support those of Gardner et al. (10) as the 2 environments did not show a statistically significant difference.
To the authors knowledge, Bertucci and Hourde (3) have conducted the only BMX study using laboratory and field testing environments. The research examined the percentage contribution of upper- and lower-body on power production. In their study, Bertucci and Hourde (3) also identified a number of physical factors that can influence the outcome of BMX performance: peak power, time of power production, and torque. However, the study did not investigate any possible variations between the performance factors obtained in the 2 environments. Consequently, any variation in the measurement of peak power, torque, and time of power production among BMX riders in the laboratory vs. field are unknown.
Therefore, because of the limited peer reviewed data, the aim of this study is to ascertain any variation in peak power, torque, and time of power production between the 2 environments in BMX cyclists. It is anticipated that the results will enable BMX coaches and researchers to make an informed decision about the validity of data collected in the laboratory and field environment.
Experimental Approach to the Problem
To establish the validity of analyzing a BMX rider's peak power, time of power production, and torque in a laboratory and in the field, 2 separate trials were conducted.
The first trials were conducted in a laboratory environment. Each subject performed 3 repeated sprint tests on a Laboratory SRM using a 32 strange gauge cycle ergometer. The ergometer was adapted to more accurately mimic a BMX bike used in competition. This was achieved through attaching a standard 70-cm straight bar to the ergometer, along with Shimano SPD (Shimano Pedal Dynamics; Shimano, Inc., Osaka, Japan) pedals, and altering the crank arm length to 175 cm. Before testing, each rider adjusted the bar height and stem length to their preferred position. The saddle was lowered so that it did not interfere with the riders when performing each sprint. An inertial load of 50.2 kg·m−2 was added to the cycle ergometer, which according to Debraux et al. (6) equates to a standard BMX gear ratio of 43 (front chain ring)/16 (rear chain ring). Before the testing protocol, each rider performed their standard precompetition warm-up consisting of seated cycling and a series of standing short sprints. Each rider then performed three 10-second sprints in a standing position and was encouraged to reach maximal power in the shortest possible time duration. The riders were instructed to perform the sprints from a stationary position with their self-selected leg in the lead position. A rest period of 10 minutes (23) was given between tests. Data from all 8 riders' sprints were recorded using SRMWin software version 7.
The second trials were undertaken at the British National Indoor BMX Centre, Manchester, United Kingdom, 3 weeks after the initial laboratory trials. The indoor track has a 5-meter high start ramp with a 28° decent. The track measures 400 meters in length, has 4 straights with a number of technical jumps on each straight section, and 3 berms (corners).
The riders performed a structured self-paced warm-up consisting of a series of standing short sprints. Riders' then performed three 10-second sprint tests from the 5-meter high start ramp using a standard electronic start gate (Pro-Gate, Rockford, IL, USA). The riders used their own cycles, which were all fitted with an (SRM) 8 strange gauge crank and a standard gear ratio of 43/16. Data from all 8 riders' sprints were recorded using SRMWin software version 7.
Seventeen riders in the United Kingdom currently hold an elite-level license, and from this population, 8 elite male BMX riders volunteered to take part in the study (mean age 21 ± 2 years). Stature of the riders was recorded using a Harpenden stadiometer (Cranlea, Bournville, Birmingham, UK) to the nearest 0.1 cm, whereas body mass and percentage body fat were recorded using air displacement plethysmography (Bod Pod; Life Systems International, Concord, Calif). The riders mean characteristics were stature 170 ± 6 cm, body mass 69 ± 3 kg, and body fat 10 ± 2%. All the riders had previous experience of using laboratory SRM cycle ergometers and had ridden the Manchester BMX track. Written consent was obtained from all participants, and a detailed description of the test protocol was issued to all participants before the study. The research protocol and experimental design received ethical approval from the University of Derby Ethics Human Studies Board and in accordance with the Declaration of Helsinki.
Power and cadence data from both the laboratory and SRM crank power meter was used to calculate the rider's peak torque. Torque (T) was calculated as follows:
where P is the power, T is the torque, and R is the cadence.
Descriptive statistics were used to analyze the percentage difference between the laboratory and field tests. A paired samples t-test was used to calculate variations between power, torque, time to peak power in the laboratory and field, and the alpha value was set at p ≤ 0.05.
Agreement was established using 95% limits of agreement (5,1). Differences between the 2 measures were plotted against the mean values and analyzed for heteroscedasticity (Figures 1–3). Where this was evident, data were logarithmically transformed to calculate the ratio limits of agreement.
The main purpose of this study was to examine the validity of testing elite BMX rider's peak power, torque, and time of power production in a laboratory and comparing these data with testing conducted in a field-based environment. The results below provide a detailed breakdown for each of the individual BMX riders. Using a paired samples t-test, the results revealed significant differences between the laboratory testing and field testing of peak power t (7) = −11.38 (p ≤ 0.01) and the relationship between the 2 environments was estimated as r = 0.78. The relationship between the laboratory and field testing time of power production was estimated as r = 0.86 and also reported significant differences, t (7) = 8.64 (p ≤ 0.01). However, no significant difference existed between torque in the laboratory vs. field test, t (7) = −1.48 (p = 0.18, r = 0.61; Table 1).
Figure 1 demonstrates the limits of agreement for peak power between the laboratory and field testing. The upper limits of agreement estimate at between 246 and 714 W showing that a greater peak power is produced in field testing than the laboratory.
Figure 2 below shows the limits of agreement for time to peak power between the laboratory and field testing. The riders were able to reduce the time of peak power production in the field compared with the laboratory with lower limits of agreement of −3.39 to −0.73 seconds.
Figure 3 demonstrates the limits of agreement for peak torque between the laboratory and field testing environments. The limits of agreement for torque were −32.99 to 19.07 for the lower and upper limits, respectively. These limits reveal slightly higher torque values in the laboratory, but they were not statistically significant (p = 0.18).
The purpose of this study was to determine any variations in data collected in a laboratory compared with that in a field testing environment among elite BMX riders. The results revealed that peak power was significantly higher (p ≤ 0.01) in the field tests when compared with the data recorded in the laboratory, with values of 34 ± 9%. Additionally, time of power production was significantly reduced (p ≤ 0.01) in the field environment by 105 ± 24%. These results therefore pose the following questions: (a) Is the field data overestimating the values recorded? and (b) Is the laboratory under estimating the values recorded?
In an attempt to answer these questions, it is worth considering the findings of Herman et al. (14) who analyzed time of peak power production and peak power in 5 elite BMX riders including 2 Olympic medalists. They analyzed the standing start sprint results from the 5 riders performed on a flat paved surface. The results showed a high peak power of 2,087 ± 157 W and a reduction in time of power production, 1.6 seconds, which have a closer comparison to the field data in this study than the laboratory data. In a separate, study Debraux et al. (6) also found high power outputs in 7 elite BMX riders who performed standing starts on a flat asphalt surface. The riders in his study were recorded to have similar peak powers (1,631 ± 368 W) as the field results from this study (1,671 ± 188 W). Although Herman et al. (14) and Debraux et al. (6) did not undertake their research on a BMX track, both studies established that peak power values were higher (2,087 ± 156.8 W and 1,631 ± 368 W, respectively) than the laboratory data recorded in the current study (1,191 ± 188 W).
One possible explanation for the high field values may be the contribution of upper body to force development in BMX cyclists. Several studies have examined the effect of upper-body muscle activation on power and performance in cycling (7,12,16,21). They all concluded that upper-body activation had a statistically significant influence on performance. To enable riders in this study to use their upper-body effectively in the laboratory testing, the SRM cycle ergometer was adjusted to accurately replicate the field riding position. The standard SRM cycle ergometer road bars were removed and replaced with a 70-cm flat bar. Each rider then adjusted the stem length and height of the ergometer to their preferred position. However, the riders produced 480 ± 119 W less on the ergometer compared with riding their own bikes on the BMX track. A possible explanation for this reduction in peak power was presented in a study by Bertucci et al. (3). The authors inferred that BMX riders rely on the upper body to produce power disproportionally compared with other cycling disciplines.
The BMX riders use their upper body by oscillation of the bike and associative leverage. As the ergometer is a static rigid piece of equipment, riders were unable to oscillate during sprints. This could be a major factor in explaining the difference in power and torque. However, this does not explain the reduction in time of power production. A possible explanation for this decrease in time of power production may be because of the geometry of the track. The gravitational force acting on the rider may be increased as the riders descend the 28° start ramp. However, this was unfortunately outside the remit of this particular study.
The data in this study suggest that significantly lower results are recorded in the laboratory testing environment, when compared with the field testing environment.
These findings have implications when comparing power and time of peak power production research data conducted in a laboratory to field data, and the utilization of laboratory data for competitive benchmarking. Crucially, this does not mean that data collected in a laboratory cannot be used. Instead the application of BMX riders' data used interchangeable between the laboratory and the field should be viewed with an amount of caution. Based on the discussion in this article, it may be useful to investigate the contribution of a BMX riders' upper body on performance variables such as velocity in a race.
The authors gratefully acknowledge the British Cycling Federation Coach Jeremy Hayes for his time and dedication to the research project. There are no conflicts of interest for this work.
1. Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Medicine 26: 217–238, 1998.
2. Balmer J, Davison RR, Bird SR. Peak power
predicts performance power during an outdoor 16.1-km cycling time trial. Med Sci Sports Exerc 32: 1485–1490, 2000.
3. Bertucci W, Hourde C. Laboratory testing and field performance in BMX riders. J Sports Sci Med 10: 417–419, 2011.
4. Bertucci W, Taiar R, Grappe F. Differences between sprint tests under laboratory and actual cycling conditions. J Sports Sci Med 45: 277–283, 2005.
5. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: 307–310, 1986.
6. Debraux P, Manolova AV, Soudain-Pineau M, Hourde C, Bertucci W. Maximal torque
and power pedaling rate relationships for high level BMX riders in field tests. J Sci Cycling 2: 51–57, 2013.
7. Doré E, Baker SJ, Jammes A, Graham M, New K, Van Praagh E. Upper body contribution during leg cycling peak power
in teenage boys and girls. Res Sports Med 14: 245–257, 2006.
8. Eriksson JS, Rosdahl H, Schantz P. Validity of the oxycon mobile metabolic system under field measuring conditions. Eur J Appl Physiol 112: 345–355, 2012.
9. Faria EW, Parker DL, Faria IE. The science of cycling: Physiology and training—Part 1. Sports Med 35: 285–312, 2005.
10. Gardner SA, Martin JC, Martin DT, Barras M, Jenkins DG. Maximal torque
- and power-pedaling rate relationships for elite sprint cyclists in laboratory and field tests. Eur J Appl Physiol 101: 287–292, 2007.
11. Goosey-Tolfrey VL, Leicht CA. Field-based physiological testing of wheelchair athletes. Sports Med 4: 77–91, 2013.
12. Grant MC, Roberts R, Baird MF, Baker JS. The effect of prior upper body exercise on subsequent wingate performance. Biomed Res Int 2014: 329328, 2014.
13. Hawley JA, Noakes TD. Peak power
output predicts maximal oxygen uptake and performance time in trained cyclists. Eur J Appl Physiol Occup Physiol 65: 79–83, 1992.
14. Herman CW, McGregor SJ, Allen H, Bolt EM. Power capabilities of elite bicycle motocross
racers during field testing in preparation for 2008 Olympics. Med Sci Sports Exerc 41: 306–307, 2009.
15. Hopker J, Jobson S. Performance Cycling: The Science of Success. London: A&C Black, 2012.
16. Hurst HT, Swarén M, Hébert-Losier K, Ericsson F, Sinclair J, Atkins S, Holmberg HC. Influence of course type on upper body muscle activity in elite cross-country and downhill mountain bikers during off road downhill cycling. J Sci Cycling 1: 2–9, 2012.
17. Jobson SA, Nevill AM, Palmer GS, Jeukendrup AE. The ecological validity of laboratory cycling: Does body size explain the difference between laboratory- and field-based cycling performance? J Sports Sci 25: 3–9, 2007.
18. Karsten B, Jobson SA, Hopker J, Jimenez A, Beedie C. High agreement between laboratory and field estimates of critical power in cycling. Int J Sports Med 35: 298–303, 2014.
19. Padilla S, Mujika I, Angulo F, Goiriena JJ. Scientific approach to the 1-h cycling world record: A case study. J Appl Physiol (1985) 89: 1522–1527, 2000.
20. Palmer GS, Dennis SC, Noakes TD, Hawley JA. Assessment of the reproducibility of performance testing on an air-braked cycle ergometer. Int J Sports Med 17: 293–298, 1996.
21. Segerström ÅB, Holmbäck AM, Elzyri T, Eriksson KF, Ringsberg K, Groop L, Ola Thorsson O, Wollmer P. Upper body muscle strength and endurance in relation to peak exercise capacity during cycling in healthy sedentary male subjects. J Strength Cond Res 25: 1413–1417, 2011.
22. Smith MF, Davison RCR, Balme RJ, Bird SR. Reliability of mean power recorded during indoor and outdoor self-paced 40 km cycling time-trials. Int J Sports Med 22: 270–274, 2001.
23. Thatcher R, Gifford R, Howatson G. The influence of recovery duration after heavy resistance exercise on sprint cycling performance. J Strength Cond Res 26: 3089–3094, 2012.
24. Waldron M, Worsfold P, Twist C, Lamb K. Concurrent validity and test–retest reliability of a global positioning system (GPS) and timing gates to assess sprint performance variables. J Sports Sci 29: 1613–1619, 2011.
25. Wooles AL, Robinson AJ, Keen PS. A static method for obtaining a calibration factor for SRM bicycle power cranks. Sports Eng 8: 137–144, 2005.