Published November 16, 2026 09:53AM
Running a marathon beats up your body. Your leg muscles will be damaged on a cellular level, which will translate into soreness and loss of strength that will persist for a week or more. Now imagine how much worse it would be if you were running one of those crazy mountain marathons with rough terrain and steep climbs and descents. It would be awful… right?
Maybe not, according to a new study in Medicine and Science in Sports and Exercise led by Mélanie Metra and Guillaume Millet of France’s Université Jean Monnet Saint-Étienne. Millet, who once finished third at the absurdly difficult 200-mile Tor des Géants mountain ultra, is one of the world’s leading trail running researchers. I’ve previously covered his research on the differences between trail and road runners, how to run uphill, how to succeed at Ultra-Trail du Mont-Blanc, and other trail-centric topics.
This time, Millet and his colleagues explore the question of whether road or trail running beats you up more. It’s not a perfect study (for reasons we’ll come back to), but it has some very interesting—and somewhat surprising—insights.
What They Did
The researchers asked runners to undertake one of the following three challenges: a simulated road marathon of roughly 26 miles on a treadmill; a simulated trail marathon on a treadmill; or an actual trail marathon on a trail outside the lab.
They took a whole bunch of measures of “neuromuscular fatigue” before and after the marathon, then again one day, two days, and seven days later. These involved various measures of leg strength, maximal jumps, questionnaires about subjective pain, and blood tests to assess markers of muscle damage. The objective was to determine whether the fatigue and recovery profile would differ after the various types of running.
To make the three challenges roughly equivalent, they used the following rule of thumb: every 100 meters of uphill climbing is equivalent to an extra kilometer of flat running. With that in mind, the two trail marathons involved a total distance of 28K (just over 17 miles) with 1,400 meters (4,600 feet) of vertical ascent and descent, for a total “equivalent effort” of roughly 42K, which is the standard (flat) marathon distance.
They recruited a total of 46 runners, of whom 16 were primarily road runners and were assigned to the simulated road marathon. The rest were primarily trail runners, and they were divided equally between the two trail marathon groups. The fitness, age, and training characteristics of the three groups were fairly well matched. Everyone was instructed to run the marathon as fast as possible.
What They Found
The overall finding was that the road marathon beat the runners up far more than either of the two trail marathons.
Here, for example, is what the loss of knee extensor (i.e. quadriceps) strength looked like in the post-marathon period, with the road group in red, the treadmill trail group in light blue, and the outdoor trail group in green:
The road marathoners lost more strength, and were still showing a deficit two days later. There’s a similar pattern, albeit less pronounced, in the subjective ratings of soreness and pain. (Calf muscles, on the other hand, didn’t show much difference between the groups.)
The blood tests also showed a major difference in muscle damage. Here are the results for one of the standard proxies of muscle damage, creatine kinase:

What It Means
The simplest interpretation is that road running causes more muscle damage than trail running.
This could be because road running is more repetitive: every stride is the same, so the same muscle fibers are being activated over and over until they get damaged. Trail running has more variety, both in the “technicity” of the terrain and in the shifts between flats, uphills, and downhills. The lack of differences between the simulated and real trail groups in the new study implies that technicity doesn’t really matter (since the treadmill has a very regular surface while the outdoor trail has an irregular one), so it’s the uphills and downhills that matter.
Another possibility is that it’s not just the variability in muscle activation—it’s also that all those hills force you to slow down, and in some cases even walk, so the slower pace reduces the load on your muscles.
Both of these factors may help explain one of the big riddles in the results, which is why all the downhills in the trail runs didn’t cause more damage. After all, marathoners complain that the slight downhills in the opening miles of Boston wreak havoc on their legs. But Boston’s smooth and gradual descent lets runners lock into a fast (and damaging) pace, while the irregular ups and downs of the trail marathon lead to a slower pace and more variable stride.
There are also some possible explanations that have less to do with the inherent differences between road and trail and more to do with the study design. One is that the groups weren’t random: the road runners had trained primarily on roads, and the trail runners primarily on trails. Perhaps it’s training on trails that makes your legs more resistant to muscle damage, rather than racing on trails being less damaging. Or perhaps there are other general differences between people who choose road versus trail running. The only way to check that possibility would be to randomize people to different groups.
The more serious problem, in my view, is that the overall average completion time for the road marathon was 3:54, compared to just 3:02 in the simulated trail group and 2:59 in the outdoor trail group. This suggests that “100 meters of climbing is equivalent to an extra kilometer of flat running” is an overly generous conversion factor. It seems very likely to me that I’d be more sore after a four-hour run than after a three-hour run, regardless of what the terrain was like.
To be fair, it’s not automatically true that longer runs cause more soreness than shorter ones. Pace matters, too. Millet himself has previously found that leg muscles are less fatigued after 200 miles at the Tor des Géants than after shorter races, because you simply can’t sustain a hard-enough-to-be-damaging pace for that long. But the difference in durations between the trail and road groups raises questions about the conclusions.
One reason I’m tempted (for better or worse) to wave away those questions is that the results fit with my own intuitions and experiences. I’ve never been as sore as I was after a road marathon; I’ve often been pleasantly surprised at how quickly my legs bounce back after a long ramble in the forest or mountains. We can’t yet call it a scientifically proven fact, but call it a strong hunch.
For more Sweat Science, sign up for the email newsletter and check out my new book The Explorer’s Gene: Why We Seek Big Challenges, New Flavors, and the Blank Spots on the Map.
