What Trail Running and Hiking Ask of Your Joints

The people who come in here after a season on the trails tend to show up stiff in the same three places. Ankles and hips first, which nobody finds surprising. Then a lower back that has gotten quietly rigid, which most of them have not connected to anything, because their back does not hurt during the run. It shows up later, sitting at a desk on Monday, and they file it under something unrelated.

Around Austin the distinction between hiking and trail running has mostly collapsed anyway. People are out on the Greenbelt and Barton Creek in low-profile shoes moving at whatever pace the rock lets them, alternating between running and picking their way down a ledge. The demand is the same either way, and it is not the demand most trail prep articles describe.

The terrain does not ask for more range

Here is the finding that reorganized how I think about this. Voloshina and Ferris put twelve runners on a treadmill fitted with wooden blocks, height variation up to about 2.5 cm, which is nothing compared to actual trail (1). Modest bumps. Then they measured everything.

Mean joint angles, torques and powers were mostly unaffected by the uneven surface. Average hip flexion, average knee moment, average ankle power: largely the same as running on a smooth belt. If you only looked at the means you would conclude the terrain barely mattered.

The variability is where the whole story sits. Ankle and knee angle variability increased by roughly 25% and hip angle variability by 35%. Ankle moment variability went up about 60%. Knee and hip moment variability more than doubled. Joint power variability rose 50% at the ankle and around 70% at the knee and hip (1).

So the joint is being asked to arrive at a wide scatter of positions and produce force at each one, on a schedule set by the rock rather than by you. Total available range barely enters into it. A stretch adds a few degrees to an arc you enter slowly, with warning, in a position you chose; the capacity the trail taxes is a different one.

This is also why generic mobility prescriptions age badly for trail people. The relevant question is not how far the hip goes; it is how much of that arc the hip can produce force in, and how quickly it can get there. Controlled articular rotations exist for exactly this reason. Taking a joint slowly through its whole rotational envelope under your own control is a way of covering the scatter rather than one favored line through it.

The ankle goes quiet and the thigh picks it up

Two more numbers from the same study, and they surprised me the first time I read them.

Positive ankle work decreased by 22% on the uneven surface and negative ankle work by 18%. Leg stiffness increased about 20%, and runners contacted the ground with a shorter, more bent leg (1). The joint closest to the unpredictable surface gave up mechanical contribution and got stiffer. The authors read this as a stabilization strategy, which tracks. The ankle is the first thing to encounter the perturbation and the least able to plan for it.

Meanwhile, three thigh muscles showed increased EMG activity: vastus medialis by 7%, medial hamstring by 19%, and rectus femoris by 20%. Nothing in the lower leg changed significantly (1).

That reads to me like a tax. The ankle stiffens because it cannot predict the ground, so the work migrates up the chain to the thigh and hip. An ankle with a narrow usable range has less to give away before that migration starts, which is one plausible reason ankle restrictions on assessment so often travel with hip complaints in this population. I want to be careful here; the study measured healthy people on a treadmill, not injured trail runners, so treat that last step as a working hypothesis rather than a demonstrated chain. If you want to attack the ankle end of it directly, we built ankle mobility foundations for that.

The injury pattern at least sits in the same neighborhood. A systematic review of sixteen studies covering 8,644 trail runners found the foot was the most common site of injury, followed by knee, lower leg, thigh and ankle (2). Worth capping that honestly: incidence across those studies ranged from 1.6 to 4,285 injuries per 1,000 hours of running, which is wide enough that you should read it as a pattern of where things go wrong and not as a rate.

Descending is the part that gets people

Going up is a conditioning problem. Coming down is a tissue problem.

Downhill running shifts muscle function toward eccentric contraction, particularly at the knee extensors, to control the rate of descent and absorb impact energy (3). And high-intensity descending produces more muscular fatigue than climbing at a similar metabolic intensity, which points at eccentric loading rather than cardiorespiratory capacity as the actual limiter on a technical descent (3).

Rectus femoris deserves specific attention here because it keeps turning up. It was the muscle with the largest activity increase on uneven ground. It also crosses both the hip and the knee, and bi-articular muscles like it carry a higher proportion of fast-twitch fibers, are preferentially recruited during eccentric contractions, and are more prone to damage than the mono-articular quads sitting next to them (4). Uneven terrain loads it more than anything else in the leg, and then the descent loads it eccentrically. It is the same muscle we are chasing when someone cannot tolerate a hip flexor position, and it is a genuine pain to lengthen properly because shortening it at one end lets it go slack at the other.

The useful part of the eccentric literature is the repeated bout effect. When a downhill bout is repeated one to six weeks after the first, markers of muscle damage including isometric peak torque, creatine kinase, soreness and range of motion are meaningfully reduced (5). Protection from graded exposure, not from stretching beforehand. Which means the single most effective thing a trail runner can do about wrecked quads is run some downhill on purpose, at a manageable grade, well before the day that matters.

Why the lower back turns up stiff

This is the one people do not expect, and it is the one I see most consistently in this group.

Nothing in the Voloshina data isolates the trunk, so I am reasoning from mechanism and from what shows up on assessment rather than from a study that measured it. Hip moment variability more than doubled on uneven ground. When a hip has a narrow band it can produce force in, the demand does not disappear. It gets covered somewhere, and the somewhere available is the segment directly above.

I am not saying trail running causes back stiffness, and I am not saying a stiff back is coming from the hips. What I am saying is that hip restriction and lower back complaints travel together often enough in this population that treating them as separate problems wastes time. We wrote more about that relationship in why back pain is not always a back problem.

What actually changes in training

The training implication follows from the variability finding rather than from any of the usual advice.

Range without control in it is not useful to a trail runner, because the trail is going to demand force production at angles you did not choose. That is what PAILs and RAILs are for. You get to a position, hold it long enough to matter, then produce force into it under your own power so the nervous system has a reason to give you access under speed and load. Passive range that you have never contracted into is range the rock is going to take back from you at the worst possible moment.

Then the exposure piece. Graded downhill work, on the actual grade you are going to run, spread across the weeks before you need it.

And the shoe question, which comes up every time. Low-profile shoes on rock give you more information from the ground and more demand on the foot at the same time. That is a reasonable trade if the foot has been trained to handle it and a bad one if it has not been out of a cushioned trainer in five years. We have written about the barefoot side of that argument separately.

None of this tells you which of these applies to you specifically. Your ankle might be fine and your hip rotation might be the thing costing you on descents, or it might be the reverse, and there is no article that can sort that out from a distance. That is the entire reason we assess first.

References

1. Voloshina AS, Ferris DP. Biomechanics and energetics of running on uneven terrain. J Exp Biol. 2015;218(5):711-719. https://doi.org/10.1242/jeb.106518
2. Viljoen CT, et al. Epidemiology of Injury and Illness Among Trail Runners: A Systematic Review. Sports Med. 2021;51(5):917-943. https://link.springer.com/article/10.1007/s40279-020-01418-1
3. A review of uphill and downhill running: biomechanics, physiology and modulating factors. Front Bioeng Biotechnol. 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12592170/
4. Localization of damage in the human leg muscles induced by downhill running. Sci Rep. 2017;7:5769. https://www.nature.com/articles/s41598-017-06129-8
5. A single bout of downhill running attenuates subsequent level running-induced fatigue. Sci Rep. 2020;10:19004. https://pmc.ncbi.nlm.nih.gov/articles/PMC7606541/

---

Written by