The Science of Sport, Training and Performance

Intensity, Training, and Load Characteristics of Professional Road Cycling

The Studies

Intensity and Load Characteristics of Professional Road Cycling: Differences Between Men’s and Women’s Races (2019)

By Dajo Sanders, Teun van Erp and Jos J. de Koning

Full text article available here.

Training Characteristics of Male and Female Professional Road Cyclists: a 4-Year Retrospective Analysis (2020)

By Teun van Erp, Dajo Sanders and Jos J. de Koning

Full text article available here.


Effective training periodization is often a direct consequence of having a better understanding of the demands of competition in terms of exercise intensity and training load. The demands of professional road cycling vary between males and females due to the nature of competitions (e.g. longer one-day and multiple-stage races in men’s cycling compared to women’s).

As such, training characteristics of male and female cyclists may have to differ to account for these differences, which ultimately leads to different training prescriptions and preparation for races.

These two studies evaluated the intensity and load demands of professional road cycling races (first study) and how this information influences training prescription to male and female professional cyclists (second study) in a 4-year retrospective analysis.


Data from 20 male and 10 female professional road cyclists of a World-Tour Team and UCI elite women’s team, respectively, were analysed.

Rate of Perceived Exertion (RPE), heart rate (HR) and power output data were collected during both single-day and multi-day stage races. The same variables were collected during training days. If a cyclist was absent from training for a period of 3 or more months, the data set for that year was excluded from the analysis.

Training intensity distribution was quantified using a 5-zone model for both HR and power output. HR zones were based on percentages of maximal HR (zone 1: 50-59% HRmax; zone 2: 60-69% HRmax; zone 3: 70-79% HRmax; zone 4: 80-89% HRmax; zone 5: 90-100% HRmax). Power output zones were based on percentages of Functional Threshold Power (FTP) (zone 1: <55% of FTP; zone 2: 56-75% of FTP; zone 3: 76-90% of FTP; zone 4: 91-105% of FTP; zone 5: >105% of FTP). FTP was determined as 95% of the highest 20-min mean maximal power output obtained for that specific season and adjusted accordingly every season.

Internal training load was quantified using Edwards training impulse (eTRIMP) based on arbitrary weighting factors (zone 1 = 1; zone 2 = 2; and so on) and session RPE (borg scale value x session duration in min). External training load was quantified using kJ (mechanical energy) and TSS (training stress score). These load metrics were also expressed relatively per kilometer. 

Effect sizes were calculated as Cohen’s d: 0 to 0.19, trivial; 0.2 to 0.59, small; 0.6 to 1.19, moderate; 1.2 to 1.99, large; >2.0, very large.

Practical Applications: What can we learn from the study?

1) Intensity and relative training load of women's races is higher than men's

As expected, duration, distance, absolute mean power output, total work (kJ), TSS and eTRIMP were significantly higher in men’s races compared to women’s. The race format differs between men’s and women’s cycling, so this is not surprising at all. Men’s races can often be two times the distances covered by females during competition (maximum 160 km), in both 1-day and multi-stage races. Consequently, the differences seen in absolute values of work, TSS and eTRIMP are almost certainly related to differences in volume-based metrics.

When we consider, load metrics expressed relative to distance (load per km) we can quickly see that these are substantially higher in women’s races, with large to very large effects for TSS and eTRIMP (relatively speaking) and a small higher effect for sRPE (see above). Similar results were found when TSS and eTRIMP were expressed relative to duration (load per min), with large to very large effects.

Again, percentage of time spent in zone confirms these findings. During racing women spent on average 42% of time in zone 4 and 21% in zone 5, whereas men spent 24% in zone 4 and 6% in zone 5 (large effect). Similarly, intensity factor (IF), mean HR and mean HR as a percentage of HRmax were also largely higher in women’s races versus men’s. This is true for both single day and multiday stage races.

The intensity of women’s races is substantially higher compared with men’s races. As such, approaches to training and preparation should not be used interchangeably between male and female cyclists given the different demands of racing.

2) Women spend less time in the low-intensity zones and more in the high-intensity zones during training

Time spent at low-intensities (zones 1 and 2) in both HR- and power-based training zones was higher in men’s training (small to moderate effect). In contrast, time spent in high-intensity zones (i.e. zone 4 and 5) was higher in women compared with men (albeit a small effect). % Time spent in zones 4 and 5 was also significantly higher in women’s training (small to moderate effect).

Women not only train less, but also spend more of that time training at higher intensities. However, the differences are not as pronounced in training as they are in racing. Women spend ~25% more time in high intensity zones (i.e. zone 4 and 5, HR-based) during racing, whereas that difference is only ~13% in training. Women train at higher intensities because they race at higher intensities. Another possible explanation is that not all women are full-time cyclists, which means they may have limited time to train and the way they compensate for this is by adding more high-intensity sessions for a given (feasible) training volume. Similarly, the racing calendar for women contains less competitions days, which means they can afford to do more high-intensity work throughout the season.

3) Is a pyramidal training intensity distribution more appropriate to elite cycling?

One of the major limitations of these studies was the fact that HR- and power-based training zones were not derived from physiological thresholds (i.e. LT1/VT1 and LT2/VT2/CP/OBLA/MLSS). Training zones were calculated from arbitrary percentages of HRmax and FTP, respectively, which makes it difficult to analyse the training load and intensity distribution with a high level of confidence. 

Nonetheless, we can (more or less) see from the figure above that a pyramidal training intensity distribution was used over the polarized approach, particularly in men’s training. A big emphasis is put on zone 1 and zone 2 training (~86% of training in men’s), with ~8% in the intermediate zone and ~6% of training spent in zones 4 and 5.

In races, cyclists spend considerable time at or around threshold intensity, particularly in stage races involving mountain climbing, where they are required to perform repeated efforts at this intensity to pace themselves over the climbs. Could this be the reason behind the preference for a pyramidal training intensity distribution as opposed to the polarized model? In cycling, it would certainly appear that the pyramidal approach is just as popular as the polarized training approach.


Share on facebook
Share on twitter
Share on linkedin

Related Articles

Training Support

How can I help you?


Bernardo Norte

BSc, Applied Sport Science