Trail Run Racing: Factors Affecting Race Performance

Trail Run Racing: Factors Affecting Race Performance

If we can agree that research on trail running has been victim to both design and sampling bias – then we know that what we know about running performance may not apply to trail running performance – this was the premise of my previous post. So what evidence do I have to support this claim of bias within running research?


Recently I did a library search (think search engine for academics like google scholar) with the following keywords; running, endurance, biomechanics, metabolism, mechanical, physiology, efficiency, performance. What you’ll probably not have picked up is that I specifically didn’t include the words, “trail” or “road” in my search. I was seeking to find out whether my thoughts about bias were true or not. In doing so, I had to make sure that I left out the two words that would skew the results to either particular sample – here’s what I found:



I took the first 30 articles that popped up in the library search engine and summarized them as you see above. 50% of the articles about running performance were done on treadmills, 23% on road/tar, 13% on tartan/track and 14% on trail. Additionally, these articles ranged in publication date from 2010-2018 – a period in which research into trail running has shot upwards exponentially. Lastly, two of the articles on trail running were done by the same researchers (Lazzer et al.) on the same population (multi-stage trail race runners in Italy) published in 2012 and 2014 respectively. This shows that at best, 14% of what we know about running performance can be applied to a trail running population, but, owing to how knowledge is gathered over long periods of time, I think it’s safe to suggest that if I included all dates to include the early, mid, and late 1900s – this trail running representation would be even worse. I am going to be doing a research study on this hypothesis next year and examine this potential bias further so I’ll be sure to keep you filled in on what I find. I’m also going to explain in a later post why the treadmill studies favour road runners and road running performance rather than trail running/ers.


Hopefully I’ve convinced you that at the very least we need to readdress what we know about trail running performance. So what do we know? For the remainder of this post I’m going to be discussing the wonderful work of Ehrstrom et al. (2018) – an article titled, “Short Trail Running Race: Beyond the Classic Model for Endurance Running Performance”. In this research study, a group of 9 highly trained trail runners (top 20 runners in national/regional competitions) who completed the following over a period of 21 days;


  • A VO2MAX treadmill running test with +10% slope (to determine VO2MAX and lactate threshold),
  • A submaximal treadmill test at both 0% and +10% slope (to determine running economy at different speeds and elevation profiles),
  • Muscle performance tests (to determine maximal voluntary contraction and muscle endurance of the quadriceps),
  • A treadmill run to exhaustion,
  • And a 27km trail race (with 1,400m positive elevation for an elevation-to-distance ratio of 51.9).


What they found was that the classic endurance performance model of VO2MAX, lactate threshold, and running economy did not explain performance in a trail race for these men. When the researchers compared the 9 participants’ 27km trail run rankings with their VO2MAX, lactate thresholds and running economies (0% incline), this classic endurance model was only able to explain 48.1% of the variance in race result. Ehrstrom et al. suggest that the power of the classic endurance model to predict running performance is likely limited to flatter running surfaces and predicting variation in heterogenous (largely different) running populations rather than the homogenous (largely similar) study sample they used. Interestingly, they found that incorporating local muscular endurance (measured through the use of a Biodex isokinetic dynamometer) into the predictive model significantly increased the power of the model by an additional 26.9%. They suggest that for trail running, having superior muscular endurance might delay any changes in muscle recruitment or coordination patterns which could regulate the development of fatigue. Additionally, this muscular endurance was significantly correlated with uphill running performance/time most likely because trail runners are specifically adapted to climbing mountains (regular hill training and training on mountains) which subsequently improves their muscular endurance capacity and subsequent race time in events requiring large elevation gains.


Lastly, Ehrstrom et al. found that if they replaced running economy (0% incline) with running economy (10% incline), they improved the predictive model by an additional 20%. This suggests that running economy (oxygen cost at any given speed) is better tested in conditions specific to race performance (uphill or downhill running) than tested on flat surfaces (0% incline) which are not specific to trail running. Once again, it is the runners’ ability to run uphill efficiently rather than on flat surfaces that will ultimately distinguish their race performances. In a similar study to the study by Ehrstrom et al., it was found that running economy was better determined for trail runners when running outside on forest floor and mountain trail and at 10% incline or decline. The suggestion here is that trail runners are not only adapted to incline/decline running, but that they are better adapted to running on surfaces with technical demands than their road running counterparts.   


This all suggests that trail runners should be looking to improve not only their classic endurance model variables, but also their muscular endurance and running economy while running uphill. These two factors are not as important in road running performance, and this is why I am such a huge advocate of strength training for trail runners (see featured pic). The evidence from this study suggests that for trail runners with similar VO2MAX, lactate threshold and running economy, it is their muscular endurance and efficiency running uphill that will affect their race performance. As such, if you’re not in the gym already, and if you’re not incorporating uphill sessions into your weekly program – start now! This also points me even more in the direction of having “elevation gain” targets in my training programs, as opposed to volume (distance/time) or intensity (pace/heart rate) alone. Not only should trail runners be looking to incorporate hills and resistance training into their routines, they should also be looking to include stability work in the gym, and hill specific training (on trails, not on road) in their running programs. Next week I’ll be presenting further evidence of how trail running and road running performance are different – and how trail runners should train differently to improve specific race performance.



Ehrstrom, S., et al. (2018). Short trail running race: beyond the classic model for endurance running performance. Medicine & Science in Sports & Exercise. Pg 580-588

1 Comment
  • […] Last post I discussed the factors affecting trail running race performance and how they differ to road running performance greatly. Today I’m going to present some more information about how trail running and road running are different to one another using the example of a series of VO2MAX tests done by Scheer, Ramme, Reinsberger and Heitkamp (2018) on 13 highly trained trail runners. […]

Post a Comment