Torque training in cycling: the never ending debate

Historically, in science in general (and obviously in sports science) it was the person who presented a novel and groundbreaking finding that had to demonstrate that the innovation really had the advantages that were advertised. Curiously, with torque intervals (understood as low cadence at high power outputs) the opposite is happening: followers of this type of work have managed to reach a status quo in which, paradoxically, science must show over and over again that their hypothesis is weak. How have we reached the point where studies and more studies doubt something that is no longer empirical experience and yet the burden of proof falls on the side of the detractors? Honestly, I ignore it at the same time that I am surprised.

First of all, I want to make it clear that I am not a strong opponent of torque specific intervals. I think they have their uses, especially to improve gross efficiency, and that they can be used in specific cases. But of course, I do not consider (as I frequently read on social networks) that torque training is a panacea. These types of intervals carry a series of risks at the anatomical level, they are extremely demanding at the muscular level and it is also very difficult to isolate their real usefulness: have I improved this year because I have incorporated torque work or simply because of the accumulation of training load from previous years? My real opinion on this matter does not differ much from what I already wrote in an article a couple of years ago. The scientific evidence is still somewhat scant but points to a lack of utility of this type of intervals. My empirical experience remains similar: they are useful in highly selected cases but it is necessary to study the risk/benefit ratio individually. Am I at risk of producing a recurrence of a previous pathology? Am I going to be able to correctly manage the extra muscle fatigue they generate? Have I exhausted conservative options previously? Here are some questions to ask yourself before proceeding with torque intervals.

Getting to the point, the objective of this article is not to repeat what I already said previously. My purpose is to make a brief review of the literature to counteract the opinions that are being expressed in social networks, especially those that seek to discredit academic works (and their authors) that show the apparent uselessness of these intervals.

Starting in chronological order, Paton et al., (2009) observed improvements in peak power and power at 4mmol lactate levels in amateur cyclists after 4 weeks of training, which consisted of 30s intervals at high power (500-520w). and relatively low cadences (60-70rpm). These types of intervals do not coincide with those preached by today’s authors, who opt for longer intervals (4-10min) at even lower cadences. Following the time line, Koninckx et al., (2010) did not see differences between the torque training group and the one that performed conventional training in the gym, although I would like to underline that their definition of torque is somewhat lax (very high power during very short intervals at 80rpm). The first study that I would define as high quality, with a good sample, clearly established torque intervals, and a good control group was that of Nimmerichter et al., (2012). Here, one group performed 5-minute uphill intervals at 300w and 60rpm while the control group continued with their usual training. Interestingly, there were no clear differences in the results between the two groups.

The study by Kristoffersen et al., (2014) is especially interesting. Its design has been extremely criticized in social networks in recent days. Here, curiously, the group that trained at freely chosen cadences achieved better adaptations than the group that performed intervals at extremely low cadences (40rpm). Why has this study been criticized? Because they defined the intensity of the intervals with heart rate. Although I agree that this is not optimal from a methodological point of view (we are establishing peripheral work with a central marker), I totally disagree with those who discredit this study by saying that real torque work was not performed. I encourage all readers to perform some 6-minute intervals at 80% Hrmax and 40rpm and then observe the N·m values obtained… Of course it is torque specific work. The real thing here is that the results of this work refute the idea that some good training coaches (and better commercials) try to sell us…

Finally, the quality papers that have been published most recently are those by Ludyga et al., (2016) and Whitty et al., (2016). In the first case, no differences were observed between the torque group and the high cadence group. In the case of the Whitty study, there was a technical tie: best 15-minute power output for the torque group, same VO2max and power in the incremental test for both groups, and better gross efficiency for the high cadence group.

As we can see, there are neither one nor two studies that show an apparent lack of effect of this type of work. At this point I am going to return to what was mentioned at the beginning of the article: the followers of torque training are the ones who have to convince us of its usefulness, taking into account that these interventions may produce certain injuries that can tip the balance of the risk/benefit in favor of detractors. As we well know, for the moment they have only presented empirical experiences that do not coincide with what is reported by science. I hope that in the near future they will be able to support their claims with independent data and, if that is the case, I will be the first to be interested in delving into the matter.

To continue questioning the arguments posted online, fans of torque training criticized the studies because they had not been carried out at sufficiently low cadences. At that time they were talking about 60rpm. When the first studies with these cadences were published, they began to talk about 40rpm as the sweet spot. When new evidence has been published that rejects this point, the argument has changed directly to “they need to set the intervals in N·m”. I am convinced that in the near future we will have papers with this methodology, but before getting there I would like to underline something that I think is not being taken into account: We have already enough fights with many athletes when we propose intervals with a 30w range of freedom. I cannot imagine what would be to ask them to perform intervals according to torque values (or alternatively delimit watts even more while demarcating cadence as well). It must be very fun to push 400w while maintaining 40rpm at the same time…

Finally, a small note that further reinforces the position of the detractors: most of the quality studies carried out with torque intervals have a control group that does high cadence work (110 to even 130rpm). This seems to me to be a methodological flaw since the control group should always be the one that trains at freely chosen cadences. What interests me is comparing torque with doing the same intervals but with natural cadences. What kind of neural adaptations are we looking for at 130rpm? Will they really have a transfer to the power associated with the freely chosen cadences? Most likely not, the results will simply be improved at those same cadences. Future studies should take this into account to achieve fair comparisons between groups.

In summary:

1- There is no clear scientific evidence that shows the benefits of torque training. Interventions lasting 1-3 months at 40-60rpm and high power outputs with several different interval durations have been explored. To date, no study has prescribed intervals in N·m.

2- Torque training, due to its extreme demand at the muscular level, can lead to injuries: patellar tendonitis, aggravation of patellar chondromalacia or recurrence of Osgood-Schlatter symptoms. These setbacks can lead to a lack of adherence to training that would nullify the hypothetical advantages obtained by the intervention in the first place.

3- Future studies should not compare torque with high cadence, but freely chosen cadences for a fair comparison between interventions.

4- Apart from greater fatigue at the muscular level, torque entails greater complexity (need to guide intervals in torque values or, alternatively, power+cadence) that can affect adherence to training.

5- There is empirical evidence reported by various coaches regarding the apparent usefulness of this work. This evidence remains to be contrasted at a scientific level in future studies.

Sebastian Sitko

Professor at the Faculty of Health and Sports Sciences, University of Zaragoza. PhD in Sports Sciences, master’s degree in High Performance Sports by the French Olympic Committee, national cycling, triathlon and athletics coach and official coach for the American College of Sports Medicine.


Koninckx, E., Van Leemputte, M., & Hespel, P. (2010). Effect of isokinetic cycling versus weight training on maximal power output and endurance performance in cycling. European journal of applied physiology109(4), 699-708.

Kristoffersen, M., Gundersen, H., Leirdal, S., & Iversen, V. V. (2014). Low cadence interval training at moderate intensity does not improve cycling performance in highly trained veteran cyclists. Frontiers in physiology5, 34.

Ludyga, S., Gronwald, T., & Hottenrott, K. (2016). Effects of high vs. low cadence training on cyclists’ brain cortical activity during exercise. Journal of Science and Medicine in Sport19(4), 342-347.

Nimmerichter, A., Eston, R., Bachl, N., & Williams, C. (2012). Effects of low and high cadence interval training on power output in flat and uphill cycling time-trials. European journal of applied physiology112(1), 69-78.

Paton, C. D., Hopkins, W. G., & Cook, C. (2009). Effects of low-vs. high-cadence interval training on cycling performance. The Journal of Strength & Conditioning Research23(6), 1758-1763.

Whitty, A. G., Murphy, A. J., Coutts, A. J., & Watsford, M. L. (2016). The effect of low-vs high-cadence interval training on the freely chosen cadence and performance in endurance-trained cyclists. Applied Physiology, Nutrition, and Metabolism41(6), 666-673.