Abstract
Objectives
This study aimed to investigate predictors of cycling performance in U23 cyclists by comparing traditional approaches to a novel method – the compound score. Thirty male U23 cyclists (N = 30, age 20.1 ± 1.1 yrs, body mass 69.0 ± 6.9 kg, height 182.6 ± 6.2 cm, O2max 73.8 ± 2.5 mL·kg−1·min−1) participated in this study.
Design
Power output information was derived from laboratory and field-testing during pre-season and mean maximal power outputs (MMP) from racing season. Absolute and relative 5-min MMP, 5-min MMP after 2000 kJ (MMP2000 kJ), allometric scaling and the compound score were compared to the race score and podium (top 3) performance during a competitive season.
Methods
Positive and negative predictive values were calculated for all significant performance variables for the likelihood of a podium performance.
Results
The absolute 5-min MMP of the field test revealed the highest negative predictive capacity (82.4%, p = 0.012) for a podium performance. The compound score of the 5-min MMP2000 kJ demonstrated the highest positive and average predictive capacity (83.3%, 78.0%, p = 0.007 - respectively). The multi-linear regression analysis revealed a significant predictive capacity between performance variables and the race score (R2 = 0.55, p = 0.015).
Conclusions
Collectively the results of the present study reveal that the compound score, alongside absolute power, was able to predict the highest positive and average likelihood for a podium performance. These findings can help to better understand performance capacity from field data to predict future cycling success.
Keywords
Practical implications
The proposed concept of the compound score allows practitioners to use power output information from the laboratory and field scenarios to predict the likelihood of race success in single day races. This methodology enables performance comparisons and benchmarking within a team, or in the long-term athlete development process. The assessment does not require expensive equipment nor access to special infrastructure. It only requires a well calibrated and reliable mobile power meter and scale.
41
,42
1. Introduction
Professional road cycling is characterized by racing over varied topography, ranging from flat to extremely mountainous terrain.
1
,2
Researchers have frequently attempted to quantify the characteristics of performance in elite cycling.3
,4
Primarily the focus has been on examining the relationships between physiological variables and race performance5
,6
and in some cases, specific tests have been developed in order to predict performance in subsequent races.7
, 8
, 9
However, while this information is useful, it would be of great(er) benefit to practitioners to have ‘performance thresholds’ that identify the level of performance needed to achieve a certain level of performance.
10
,11
In professional cycling, the goal is to achieve a win or podium placing. However, the ‘performance threshold’ that equates to the minimum level required to achieve a podium or win in a professional cycling race remains elusive. Ideally, these performance thresholds would be derived from standard physiological and performance test measures, or even field data, to minimize the amount of standardized testing athletes need to undergo. To date, efforts to do this have primarily focused on climbing performance, however, the exact performance threshold is unclear. Van Erp and colleagues12
suggested that athletes need to produce 5.9 W·kg−1 for ~30 min on the final climb of a mountainous race to be in contention for the victory in a Grand Tour, however, Leo et al.13
reported that this may only be sufficient to achieve a place in the top 10 of a mountainous professional multi-stage race. This focus on climbing performance is perhaps due to power output normalized to body mass being a very good predictor of climbing performance.14
However, as aforementioned, professional cycling races take place on varied terrain. On flatter terrain, absolute power output and aerodynamic drag are the primary components of performance.14
,15
An important consideration is that absolute power output is proportionally related to the body mass of the rider, while relative power output is inversely related to the cyclist's mass. Thus, these two factors are diverging with changes in mass and might explain why neither is able to predict performance with a high degree of accuracy in different racing terrain (flat, semi-mountainous and mountainous).3
,14
For this reason, allometric scaling was introduced to account for the influence of the rider's mass on power output.16
, 17
, 18
However, the existing research surrounding allometric scaling has not assessed the value of the metric to predict race performance. Finally, it has been suggested that the power outputs that athletes can produce in key moments of the race, rather than maximal values, are an important predictor of race performance12
; these key moments may occur on either uphill or flat terrain.19
Recent research has focused on the decline in the power profile after certain quantities of accumulated work. For example, it has been shown that after 2000 kJ power output drops significantly in U23 cyclists, this decline is determined by physiological characteristics such as gross efficiency and substrate utilization.13
,20
,21
Thus, it follows that if we wish to predict performance on varied terrain, both absolute and relative power outputs in fresh and in a fatigued state are important components of race performance.- Spragg J.
- Leo P.
- Swart J.
The relationship between physiological characteristics and durability in male professional cyclists.
Med Sci Sports Exerc. 2022; (Published online)https://doi.org/10.1249/MSS.0000000000003024
The aim of this study is to derive performance threshold values that are required to achieve a podium performance in U23 one day cycling races. We hypothesized that beside existing methods, a novel adaptation of both absolute and relative power output values; ‘the compound’ score, may offer a better predictive capacity than either relative or absolute power outputs or the existing allometric scaling options.
2. Methods
Power data were recorded from power meters fitted to the participants' bicycle during laboratory and field testing, training as well as racing during a competitive racing season (February to October). Prior to the season, within a 10-day period, participants performed both a laboratory peak power output test (PPO) and a 5-min maximum effort field test. Body mass (SECA 878) and height (SECA 217, SECA GmbH & Co. KG, Hamburg, Germany) were collected in conjunction with laboratory measures. The performance metrics included, absolute power output, relative power output, allometric scaling of power output and the compound score from the laboratory PPO test, 5-min field test, 5-min mean maximal power output (MMP) and 5-min MMP after 2000 kJ of accumulated total work and intensity (MMP2000 kJ). The 5-min duration was selected because it allows indirect comparison with laboratory derived peak power and maximum oxygen uptake (O2max).
10
,22
In a first step differences in absolute and relative power output, allometric scaling of power output and the compound score of the laboratory PPO test, 5-min field test, 5-min MMP and 5-min MMP2000 kJ were assessed. In a second step the relationship between race score and all performance variables was calculated. Thirdly performance variables, which showed a significant relationship to the race score were then incorporated into a multi-regression analysis to determine predictive capacity for a podium vs. non-podium performance.
Thirty male U23 professional cyclists participated in the study (age, 20.1 ± 1.1 yrs, body mass 69 ± 6.9 kg, height 182.6 ± 6.2 cm, body mass index 20.7 ± 1.4 kg·m−2, body surface area 1.9 ± 0.1 m2, O2max 73.8 ± 2.5 mL·kg−1·min−1). All participants were members of an international cycling union (UCI) continental licensed team during the cycling season analyzed. In cases of any prolonged illness, injury (defined as having completed less than 20 total race days), or termination of their cycling careers, participants were excluded from all analysis. Written informed consent was obtained after each participant was given a verbal and written explanation of the experimental protocol and fully understood the possible risks involved in taking part in the study. The study protocol was approved by the Ethical Review Board at the University of Innsbruck and followed the principles as set out in the declaration of Helsinki.
Participants were asked to avoid any exhaustive exercise for 24 h prior to the test, to avoid caffeine in the 12 h prior to the test and arrive 3 h postprandial in a hydrated state. The following PPO protocol was applied: initial power output 150 W, increment 20 W·min−1. Open circuit spiro ergometry with a breath-by-breath technique (ZAN600, nSpire Health GmbH, Oberthulba, Germany) was used.
23
Volume and flow were calibrated with a 1 L syringe before each trial. Gas calibration was completed before each measurement according to the manufacturer's recommendations (4.9 vol% CO2, 15.9 vol% O2, 79.2 vol% N2, nSpire Health GmbH, Oberthulba, Germany). All participants continuously wore a facemask and breathed through a flow sensor (Flow Sensor Type II, nSpire Health GmbH, Oberthulba, Germany). O2max and PPO were defined as the highest 30 s rolling average achieved before task failure. Continuous recordings of heart rate (HR) were performed via short-range telemetry with a 1 Hz sampling rate (V800, Polar Electro Oy, Kempele, Finland). The PPO test was performed in a controlled environment (temperature, 19–22 °C; 40–50% humidity) on the participant's own road bike mounted on an electromagnetically braked stationary trainer with a 1 Hz sampling rate (Cyclus2, RBM Elektronik-automation GmbH, Leipzig, Germany) which was calibrated prior to the start of lab-testing using the standard manufacturer specified protocol.A field test was carried out within two weeks of the initial laboratory visit, with an ambient temperature of 15–20 °C utilizing a standardized climb with an average gradient of 5.5%. The field test consisted of a 5-min maximum effort trial. Participants were encouraged to produce the highest possible workload and asked to maintain their preferred cadence.
24
Power output values for the 5-min maximum effort trial were analyzed using an open-source software package (Golden Cheetah, version 3.6, Cranleigh, United Kingdom).Power output was recorded using a standardized crank system (SRAM Red, Quarq, Spearfish, South Dakota, USA) with a 1 Hz sampling rate and monitored on a portable head unit device (Garmin Edge 520, Schaffhausen, Switzerland). A static calibration of the power meter was undertaken prior to the laboratory visit according to Wooles et al.
25
Participants were instructed to perform a ‘zero-offset’ before each training or racing session using the calibration function in their portable head unit device according to the manufacturer specified protocol.Additionally, 5-min MMP; representing the highest average power output recorded during a racing season, and 5-min MMP2000 kJ – the highest average power output after 2000 kJ of work
26
, 27
, 28
– were analyzed using an online cycling software platform (EnDuRa, Blue Cat Technical Ltd., Midhurst, West Sussex, United Kingdom). The 2000 kJ accumulated work was chosen as previous research has shown that this magnitude of accumulated work is sufficient to induce a reduction in power output in U23 and elite cycling populations.26
, 27
, 28
All power output data were manually checked for data spikes applying a three standard deviation threshold.29
- Morris A.S.
- Langari R.
Morris A.S. Langari R.B.T.M. Third E. Statistical Analysis of Measurements Subject to Random Errors. Academic Press,
2020https://doi.org/10.1016/B978-0-12-817141-7.00004-9
2.1 The compound score and allometric scaling
The compound score was calculated for all performance metrics from absolute power output and power output normalized to body mass as follows:
(1)
Allometric scaling of body mass was applied to all performance metrics using a mass exponent
18
of 0.32 as follows:(2)
2.2 Race performance
Race performances during the season for each participant were screened to select the best single day result, which was then used for further analysis. These results were achieved across a range of race categories (see Table 1). Therefore, the UCI points' score table was used to assign a weighting factor to each race category (see Table 1). As UCI points in race categories 1.2 and 1.2U were only assigned to the top 10 places, the mean race points' score of the top 10 places in each category was used to generate a weighting for each result as follows:
(3)
Table 1Race categorization.
| Race categories | Weighting factor |
|---|---|
| 1.1 NC | 5.46 |
| 1.1 | 5.40 |
| NC | 2.73 |
| 1.2 | 1.54 |
| 1.2 U | 1.11 |
(4); UCI race categories
43.
with the corresponding weighting factor based on the mean top 10 UCI points scale; NC – nations cup; WF – weighting factor; n – position.Positive, negative, and average predictivity were calculated for each performance variable for the likelihood of a podium performance during the season. Threshold values for positive and negative predictivity have been determined using receiver operating characteristics (ROCs).
30
All values are expressed as mean ± standard deviation and Shapiro Wilk's significance test assessed normality. A repeated measure analysis of variance (ANOVA) assessed the difference within performance values for absolute and relative power, allometric scaling and the compound score. In case of significant differences, a post hoc Bonferroni procedure was applied for pairwise comparisons. Spearman's linear product moment correlation was performed between all performance variables and the race score. In a second step a multi linear regression analysis investigated the predictive capacity of the significant performance variables on the podium vs. non-podium performance. The linear regression was applied for the ordinary scaled (race score) variable.
31
Cohen's d effect sizes were used for the Spearman correlation coefficient (rho) for small (0.1 to 0.3), medium (0.3 to 0.5) and large (>0.5) effects. ROC curves assessed the threshold values for each performance variable for the likelihood of a podium vs. non-podium performance.30
Positive and negative predictive capacity was only determined if the performance value reached statistical significance in the ROC analysis. All statistical analyses were completed using a commercially available software package (JASP version 0.16.3 for Windows 11, University of Amsterdam, Amsterdam, The Netherlands). All graphs and figures were created using GraphPad Prism (version 8.0.0 for Windows 11, GraphPad Software, San Diego, USA).3. Results
3.1 Performance characteristics
In total five UCI 1.1, four UCI 1.NC, 19 UCI NC, 37 UCI 1.2 and 25 UCI 1.2U races were included in the analysis, and 315 ± 16 data files were analyzed per athlete. Group wise comparison revealed that the 5-min MMP2000 kJ was significantly lower compared to the PPO test, 5-min field test and 5-min MMP for absolute power (p ≤ 0.001 to 0.002), relative power (p ≤ 0.001 to 0.003), allometric scaling (p ≤ 0.001 to 0.006) and the compound score (p ≤ 0.001 to 0.002). In addition, the 5-min relative power output in the field test was lower compared to the PPO test (p = 0.027) – see Table 2.
Table 2Performance characteristics.
| Variable | Absolute value | Relative value | Allometric scaling | Compound score |
|---|---|---|---|---|
| Lab test | 458 + 38 W | 6.7 ± 0.4 W·kg−1 | 118 ± 7 W·kg−0.32 | 3054 ± 299 |
| 5-min field test | 449 ± 30 W | 6.5 ± 0.3 W·kg−1 | 115 ± 6 W·kg−0.32 | 2900 ± 299 |
| 5-min MMP | 454 ± 6.6 W | 6.6 ± 0.3 W·kg−1 | 117 ± 6 W·kg−0.32 | 2995 ± 264 |
| 5-min MMP2000 kJ | 431 ± 40 W | 62 ± 0.3 W·kg−1 | 111 ± 7 W·kg−0.32 | 2667 ± 283 |
Mean ± SD of all performance characteristics (n = 30); MMP – mean maximum power; MMP2000 kJ – mean maximum power after 2000 kJ of accumulated work and intensity.
# Significantly different from lab test.
3.2 Relationship and predictive capacity between performance characteristics and race score
A total of 598 UCI points were scored by the participants, which included 8 podium performances and 23 top ten places.
Table 3 lists the relationship between the performance characteristics and the race score.
Table 3Relationship between performance characteristics and race performance score.
| Performance variable | Correlation coefficient (rho) | p-Value | Effect size |
|---|---|---|---|
| PPO (W) | 0.056 | 0.766 | Small |
| PPO (W·kg−1) | −0.207 | 0.271 | Small |
| PPO (W·kg−0.32) | −0.071 | 0.708 | Small |
| PPO Comp (W2·kg−1) | −0.047 | 0.803 | Small |
| 5-min field test (W) | 0.396 | 0.030 | Medium |
| 5-min field test (W·kg−1) | 0.045 | 0.811 | Small |
| 5-min field test (W·kg−0.32) | 0.364 | 0.047 | Medium |
| 5-min field test Comp (W2·kg−1) | 0.217 | 0.248 | Small |
| 5-min MMP (W) | 0.343 | 0.063 | Medium |
| 5-min MMP (W·kg−1) | 0.171 | 0.366 | Small |
| 5-min MMP (W·kg−0.32) | 0.404 | 0.026 | Medium |
| 5-min MMP Comp (W2−kg−1) | 0.396 | 0.030 | Medium |
| 5-min MMP2000 kJ (W) | 0.372 | 0.042 | Medium |
| 5-min MMP2000 kJ (W·kg−1) | 0.399 | 0.028 | Medium |
| 5-min MMP2000 kJ (W·kg−0.32) | 0.442 | 0.014 | Medium |
| 5-min MMP2000 kJ Comp (W2·kg−1) | 0.476 | 0.007 | Medium |
PPO – peak power output; MMP – mean maximum power output, Comp – compound score; MMP2000 kJ – mean maximum power output after 2000 kJ accumulated work and intensity.
The multi linear regression analysis revealed a significant predictive capacity between performance variables and the race score including absolute MMP and allometric scaling value of the 5-min field test, the compound score and allometric scaling value of 5-min MMP, as well as absolute, relative, the compound score and allometric scaling value of 5-min MMP2000 kJ (p = 0.015, coefficient of determination (R2) = 0.55) – see Fig. 1.
Fig. 1Represents the predictive capacity between podium vs. non podium performances based on the following variables: absolute MMP and allometric scaling value of 5-min field test, Comp and allometric scaling value of 5-min MMP, as well as absolute MMP, relative MMP, Comp and allometric scaling value of 5-min MMP2000 kJ.
Show full caption
Regression equation: 0.5527x + 19.83.
MMP – mean maximum power; MMP2000 kJ - mean maximum power output after 2.000 kJ accumulated work and intensity; Comp – Compound score; red dotted horizontal line represents the cutoff value for a podium performance.
The performance threshold as well as the positive, negative, and average likelihood for a podium performance are presented in Table 4.
Table 4Predictive likelihood and thresholds for a podium performance.
MMP – mean maximum power output, Comp – compound score; MMP2000 kJ – mean maximum power output after 2000 kJ accumulated work and intensity; n/a – not applicable; gray shaded cells represent the highest score for the level of significance as well as positive, negative, and average predictive likelihood.
⁎Significant (p < 0.05).
4. Discussion
To the best of the authors' knowledge, this is the first study to evaluate the relationship between race performance in cycling and combined absolute and relative power output characteristics. In this study, the only performance variables to display statistically significant effects for correlations and predictions with race performance scores were the 5-min field test (absolute MMP and allometric scaling value), the 5-min MMP (compound score and allometric scaling value) and 5-min MMP2000 kJ (absolute and relative power output, the compound score and the allometric scaling value). However, not all performance variables (e.g., laboratory measures) displayed such strong relationship and a predictive capacity to the race performance score as well as podium performance. Collectively the multi linear regression analysis could explain 55% of the variance in the race score. Therefore, more than half of the difference in performance could be explained by a performance physiological predictor despite the numerous other variables which influence performance such as terrain, aerodynamics, and team strategy.,
32
,33
Interestingly the absolute power in the 5-min field and the 5-min MMP2000 kJ compound score yielded the strongest predictive capacity for podium performances. For this reason, our hypothesis can only be partly accepted. One existing measure (absolute power 5-min MMP) was as predictive as the novel compound score. Absolute power of the 5-min field test resulted in the highest negative predictive capacity (82.4%) whereas the compound score of the 5-min MMP2000 kJ, resulted in the highest positive, and average (83.3 and 78.0% - respectively) predictive capacity for a podium performance. Laboratory derived absolute and relative power output characteristics yielded no significant relationship with nor a predictive capacity of race performance. This finding is in line with previous research where the absolute power output was demonstrated to be a strong performance differentiator.10
,34
Collectively, these results indicate that in order to achieve a podium in U23 single-day racing, it requires an absolute power output in the 5-min field of ˃448 W and a compound score of 5-min MMP2000 kJ of ˃2876 W2·kg−1 – for comparison see Table 4. However, it should be acknowledged that there was still an average likelihood of 28.1% in the absolute power of 5-min field test and 22.0% in the compound score of 5-min MMP2000 kJ that podium performances were achieved below these thresholds – as illustrated in Fig. 1. It is important to note here that the performance thresholds were statistically verified
30
and not picked by arbitrary selection, through ROC curve analysis.30
For this reason, practitioners should have confidence using the proposed methods for an U23 rider population in one day races. It might have been possible, for some performance metrics, to derive an arbitrary performance threshold value that gave a 100% negative or positive likelihood for a podium performance, however, a much greater dataset and therefore improved statistical power would have been needed.Nevertheless, caution is required when interpreting these results and trends. Combining positive and negative likelihood for podium performance in one measure – the average predictivity – the compound score of 5-min MMP2000 kJ recorded the highest (78.3%) followed by the compound score of the 5-min MMP (75.8%) and the absolute power in the 5-min field test (71.9%) – for comparison see Table 4. These findings underpin the robustness of the proposed new method – the compound score – as it combines absolute power output and power output to the mass of the rider in one metric. Researchers have previously attempted to scale power output to homogenize performance characteristics for cyclists of different masses. Swain
18
suggested that body mass should be adjusted to the power of 0.32 when predicting flat time trial performance. Subsequently, Mujika and Padilla35
used this allometric scaling method and showed that PPO adjusted for body mass (W·kg−0.32) was the strongest predictor of performance for time trial specialists and a mass exponent of 0.78 for climbing specialists (W·kg−0.78).14
,36
Likewise, Lamberts and colleagues- Impellizzeri F.M.
- Ebert T.
- Sassi A.
- et al.
Level ground and uphill cycling ability in elite female mountain bikers and road cyclists.
Eur J Appl Physiol. 2008; 102https://doi.org/10.1007/s00421-007-0590-9
23
demonstrated that allometric scaling of peak power output was the strongest predictor of 40 km laboratory time trial performance. However, these mass exponents mentioned in the literature have been chosen arbitrarily by fitting one specific dataset. Although allometric scaling accounts for different contributions from absolute and relative power output, it would require a constant update of the mass exponents based on each individual dataset. For example, the mass exponent for the present group of U23 cyclists would be 0.30 ± 0.03 for flat and 0.70 ± 0.03 for uphill terrain. Using the compound score approach instead already accounts for the individual contribution of absolute and relative power output in one measure and thus is clearly more advantageous for practitioners than the traditional allometric scaling method.The rationale for the development of the compound score came from the fact that professional road races take place over mixed terrain. Athletes are required to produce high velocities on flat terrain (often exceeding 50 km·h−1 average for many hours of racing) and in addition to be able to climb at high rates of altitude gain (1500–1850 vertical meters ascent each hour).
37
,38
In road cycling there are relatively limited means to reduce the frontal surface area of the athlete (a low handlebar position, forward saddle position and using the drop portion of the handlebar).32
As a result, absolute power is a key determinate of performance on flatter terrain. More accurately, power to drag coefficient (W·CdA−1) would be the key determinate however, as in the case for many coaches and practitioners, CdA data were not available for the use in this study. In contrast, climbing requires a high power to mass ratio to overcome the primary force of gravity as the velocity of the cyclist on climbs is lower and reduces the influence of drag as gradients steepen. However, on gradients of less than 6%, the velocities during races can reach as high as 35 km·h−1.33
,39
At these velocities the drag force is still a significant resistance to the cyclist. Therefore, racing over climbs with lower gradients requires a combination of both absolute power and power to mass. The compound score was therefore proposed to assess overall performance and it has been used by two of the authors for some time with anecdotal success.The present study confirms that the compound score is a valid and accurate predictor of overall race performance and may be more so than either absolute or relative power output information alone. However, our data also suggests that the use of the compound score alongside other more traditional metrics (e.g., absolute power) is advisable to predict podium performances over mixed terrain.
Lastly, the race winning performance is often required toward the latter stages of the race distance. Accumulated fatigue (or the ability to resist this) has been proposed as key performance determinate and is now referred to as “durability”. A growing body of literature supports this assessment.
20
,21
,- Spragg J.
- Leo P.
- Swart J.
The relationship between physiological characteristics and durability in male professional cyclists.
Med Sci Sports Exerc. 2022; (Published online)https://doi.org/10.1249/MSS.0000000000003024
40
In keeping with this literature, the fatigued 5 min MMP compound score provided the highest positive and average predictive capacity (83.3 and 78.0% - respectively).4.1 Limitations and future directions
The authors would like to acknowledge that the study is not without its limitations. Firstly, racing contained a variety of different race types (flat, semi mountainous and mountainous). It would have been preferable to determine the strongest predictors for each terrain. While this approach would be outside the scope of the present study and requiring a much more counterbalanced number of subjects for each terrain, future research is recommended to include measures of aerodynamic drag (W·CdA−1) to account for these differences in terrain. In addition, the contribution of absolute and relative power output in the compound score measure is not equal, given the differences in the scale of these variables. Caution should also be paid when replicating the results of the present study to other cycling populations (e.g., elite men, women's cycling) outside the men's U23 category.
5. Conclusion
Over mixed terrain, the ability to produce high power outputs to overcome drag, to produce high power to mass to overcome gravity and the ability to stave off the onset of fatigue are the three key determinants of success. The novel compound score provides an accurate assessment of these variables and can be used to predict subsequent likelihood of race success in an U23 single day racing season.
Funding information
No financial support or funding was received for this project.
Confirmation of ethical compliance
Ethical approval was provided by the University of Innsbruck and followed the guidelines as set out in the declarations of Helsinki.
CRediT authorship contribution statement
All authors have equally contributed to the production of the manuscript. Swart and Wakefield had the idea of the concept. Leo, Spragg and Swart conducted the data collection and analysis. Leo, Spragg, Wakefield and Swart drafted, edited, and revised the document.
Declaration of interest statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Acknowledgements
The authors would like to thank all participants and team staff of Tirol KTM Cycling Team. We would also like to thank Prof. Robert Chung, PhD for his statistical and methodological advice.
References
- Exercise intensity and load during mass-start stage races in professional road cycling.Med Sci Sports Exerc. 2001; 33: 796-802https://doi.org/10.1097/00005768-200105000-00019
- Intensity and load characteristics of professional road cycling: differences between men’s and women’s races.Int J Sports Physiol Perform. 2019; 14: 296-302https://doi.org/10.1123/ijspp.2018-0190
- Demands of professional cycling races: influence of race category and result.Eur J Sport Sci. 2020; 16: 1-12https://doi.org/10.1080/17461391.2020.1788651
- Cross-sectional differences in race demands between junior, under 23, and professional road cyclists.Int J Sports Perform Analys. 2022; (aop): 1-8
- Physiological and biomechanical factors associated with elite endurance cycling performance.Med Sci Sports Exerc. 1991; 23: 93-107
- Determinants of cycling performance: a review of the dimensions and features regulating performance in elite cycling competitions.Sports Med Open. 2020; 6: 23https://doi.org/10.1186/s40798-020-00252-z
- The power profile predicts road cycling MMP.Int J Sports Med. 2010; 31: 397-401https://doi.org/10.1055/s-0030-1247528
- Predicting cycling performance in trained to elite male and female cyclists.Int J Sports Perform Analys. 2014; 9: 610-614https://doi.org/10.1123/IJSPP.2013-0040a
- The record power profile to assess performance in elite cyclists.Int J Sports Med. 2011; 32: 839-844https://doi.org/10.1055/s-0031-1279773
- Training, performance, and physiological predictors of a successful elite senior career in junior competitive road cyclists.Int J Sports Physiol Perform. 2018; 13: 1287-1292
- Do race results in youth competitions predict future success as a road cyclist? A retrospective study in the Italian Cycling Federation.Int J Sports Physiol Perform. 2022; 17: 621-626
- Maintaining power output with accumulating levels of work done is a key determinant for success in professional cycling.Med Sci Sports Exerc. 2021; 53: 1903-1910
- Power profiling, workload characteristics and race performance of U23 and professional cyclists during the multistage race Tour of the Alps.Int J Sports Physiol Perform. 2021; 16: 1-7https://doi.org/10.1123/ijspp.2020-0381
- Level ground and uphill cycling ability in professional road cycling.Med Sci Sports Exerc. 1999; 31: 878-885https://doi.org/10.1097/00005768-199906000-00017
- Critical power and aerodynamic drag accurately predict road time-trial performance in British champion cyclists.Med Sci Sports Exerc. 2011; 43: 160-161
- Influence of body position when considering the ecological validity of laboratory time-trial cycling performance.J Sports Sci. 2008; 26: 1269-1278https://doi.org/10.1080/02640410802183585
- Allometric scaling of uphill cycling performance.Int J Sports Med. 2008; 29: 753-757
- The influence of body mass in endurance bicycling.Med Sci Sports Exerc. 1994; 26: 58-63
- The physical demands and power profile of professional men’s cycling races: an updated review.Int J Sports Physiol Perform. 2021; 16: 3-12https://doi.org/10.1123/IJSPP.2020-0508
- The relationship between training characteristics and durability in professional cyclists across a competitive season.Eur J Sport Sci. 2022; (AoP): 1-17https://doi.org/10.1080/17461391.2022.2049886
- The relationship between physiological characteristics and durability in male professional cyclists.Med Sci Sports Exerc. 2022; (Published online)https://doi.org/10.1249/MSS.0000000000003024
- Five-minute power-based test to predict maximal oxygen consumption in road cycling.Int J Sports Physiol Perform. 2021; 17: 9-15https://doi.org/10.1123/ijspp.2020-0923
- Allometric scaling of peak power output accurately predicts time trial performance and maximal oxygen consumption in trained cyclists.Br J Sports Med. 2012; 46: 36-41
- Preferred pedalling cadence in professional cycling.Med Sci Sports Exerc. 2001; 33: 1361-1366https://doi.org/10.1097/00005768-200108000-00018
- A static method for obtaining a calibration factor for SRM bicycle power cranks.Sports Eng. 2005; 8: 137-144
- Power profiling and workload characteristcs in U23 and professional cyclists during the multistage race “Tour of the Alps”.Int J Sports Med. 2020; 16: 1089-1095
- The record power profile of male professional cyclists: normative values obtained from a large database.Int J Sports Physiol Perform. 2022; 1: 1-10
- The record power profile of male professional cyclists: fatigue matters.Int J Sports Physiol Perform. 2022; 1: 1-6
- Morris A.S. Langari R.B.T.M. Third E. Statistical Analysis of Measurements Subject to Random Errors. Academic Press, 2020https://doi.org/10.1016/B978-0-12-817141-7.00004-9
- Receiver operating characteristic (ROC) curve analysis for medical diagnostic test evaluation.Caspain J Intern Med. 2013; 4: 627-635
- Likert scales, levels of measurement and the “laws” of statistics.Adv Health Sci Edu. 2010; 15: 625-632
- Aerodynamic drag in cycling pelotons: new insights by CFD simulation and wind tunnel testing.J Wind Eng Ind Aerodyn. 2018; 179: 319-337
- Mechanisms of performance improvements due to a leading teammate during uphill cycling.Int J Sports Med. 2018; 13: 1215-1222
- Elite versus non-elite cyclist–stepping up to the international/elite ranks from U23 cycling.J Sports Sci. 2022; (Published online): 1-11https://doi.org/10.1080/02640414.2022.2117394
- Physiological and performance characteristics of male professional road cyclists.Sports Med. 2001; 31: 479-487https://doi.org/10.2165/00007256-200131070-00003
- Level ground and uphill cycling ability in elite female mountain bikers and road cyclists.Eur J Appl Physiol. 2008; 102https://doi.org/10.1007/s00421-007-0590-9
- Climbing performance in U23 and professional cyclists during a multi-stage race.Int J Sports Med. 2021; 43: 161-167
- “Engine matters”: A first large scale data driven study on cyclists’ performance.in: 2013 IEEE 13th International Conference on Data Mining Workshops. IEEE, 2013: 147-153
- Aerodynamic analysis of uphill drafting in cycling.Sports Eng. 2021; 24: 1-11
- Durability and repeatability of professional cyclists during a Grand Tour.Eur J Sport Sci. 2021; (Published online): 1-8
- Accuracy of cycling power meters against a mathematical model of treadmill cycling.Int J Sports Med. 2017; 38: 454-461https://doi.org/10.1055/s-0043-102945
- Caveats and recommendations to assess the validity and reliability of cycling power meters: a systematic scoping review.Sensors. 2022; 22: 1-32https://doi.org/10.3390/s22010386
- UCI Constitution and Regulations.
Article info
Publication history
Published online: November 30, 2022
Accepted:
November 28,
2022
Received in revised form:
November 16,
2022
Received:
July 10,
2022
Identification
Copyright
© 2022 Sports Medicine Australia. Published by Elsevier Ltd. All rights reserved.
00496-0&id=gr1.jpg){kind=link}