Ask an ultra runner where the race fell apart and plenty will point to the descents, not the climbs. Running downhill makes the quads and calves brake every step while they lengthen under load, and that's where most of the muscle damage in mountain races comes from.
This guide covers what a single 5 km descent did to experienced trail runners, who coped best, how running style changed the damage, and how to build downhill resilience into training.
The main study here is new (January 2026) and exploratory. It shows who coped better, not a proven training dose, so we've flagged its limits as we go.
Why Descents Hurt
The braking problem
Every downhill step lands harder and has to be slowed. The quads and calves do that braking while they lengthen, which is called an eccentric contraction. Eccentric work is cheap for the heart and lungs but rough on muscle fibres, so a descent can feel easy and still do real damage.
Average creatine kinase after the Ultra-Trail du Mont-Blanc or Western States can top 15,000 U/L. Late in ultra-trails, runners tend to lose more speed on descents and flats than on climbs. The pacing research the study team draws on suggests the best finishers are the ones who hold their downhill speed late in a race.
Diagram
Same muscle, two jobs
Schematic. Lengthening under load costs the heart and lungs less, but it's the action that damages muscle fibres.
What gets measured
Damaged muscle leaks proteins into the blood. Researchers use three of them as markers.
Rises and peaks fastest, so it's the most sensitive marker in the first hour after exercise.
The marker most people have heard of. Climbs more slowly and peaks later.
Also rises more slowly. Usually shortened to LDH.
What One Descent Does
The 36 runners (25 men, 11 women, average age 45) were all entered in Spain's 106 km Penyagolosa Trails race, and most had been running for well over a decade.
The test
The pace was set so their heart rate matched their first ventilatory threshold on an uphill test, which worked out to about 13.3 km/h. For the heart and lungs, that's comfortable. Blood was taken before and 30 minutes after, and again before and after the race itself four to six weeks later.
Figure 1
Rise in muscle damage markers 30 minutes after a 5 km descent
Average change across 36 runners. All three rises were statistically significant. A 30-minute sample catches creatine kinase and LDH early in their rise, which is why myoglobin looks so much bigger. Source: Martinez-Navarro et al. (2026).
For scale, creatine kinase in the same runners went from 146 to 3,930 U/L across the 106 km race, which climbs 5,584 m and drops 4,369 m.
A possible readiness test
LDH after the 5 km descent tracked LDH after the race (a strong correlation, r = 0.64, in the 27 finishers). With two extreme values removed it fell to moderate (r = 0.43). The authors suggest a standard descent could one day help gauge whether an athlete's legs are ready for an ultra. It needs confirming in bigger groups before anyone relies on it.
Who Coped Best
The researchers split the runners by what they'd done in the three months before the race. Weekly downhill repetitions made a clear difference. Strength training at least twice a week didn't show one in this group.
Figure 2
Weekly downhill repetitions vs none
Squat strength change after the descent
Creatine kinase after the descent (U/L)
Squat strength was measured as a maximal isometric half squat before and after the descent. Both differences were statistically significant, with moderate to large effect sizes. Only about 8 of the 36 runners (22%) did weekly downhill repetitions, so that group is small. Source: Martinez-Navarro et al. (2026).
Calf strength dropped by about 12 to 13% in both groups, so downhill training didn't protect everything. The squat result matters, though, because the quads do most of the braking on a descent.
Coach the Descent
The runners wore a footpod during the descent. Those who ran lower to the ground held on to more strength afterwards.
Run lower to the ground
Lower vertical oscillation was linked to keeping more calf strength (a moderate correlation, r = 0.44).
A shorter step length was linked to keeping more squat strength (a moderate correlation, r = 0.37).
The authors call this a more "terrestrial" pattern: quicker, shorter, flatter steps rather than long, floating strides. It could also work the other way round. Runners who've been hurt by descents before may have learned to run them this way to limit pain. Gait was linked to strength loss, not to the blood markers.
Interactive 3D
Two ways down a 15% slope at 13.3 km/h, and a steep 30% descent
3D runners load here on the published page. The Squarespace editor doesn't run scripts, so they won't appear while you're editing.
Illustrative model, not motion capture. The planted foot is locked to the slope and the joint angles were set from typical running ranges, adjusted for downhill running (more knee bend and a more upturned ankle at landing). Pelvic tilt and side-to-side hip drop follow a 15% decline study; the forward trunk lean and knee drive are a coaching illustration. Cadence, step length and bounce are illustrative, not study values. The short-step runner sits about 6% above the long strider's cadence, close to the range a 2025 review linked to the lowest energy cost and impact loading downhill, and gets most of its shorter step from landing closer under the body. The third runner, on a 30% slope at 8 km/h and 185 steps a minute, is also illustrative: on steep ground runners slow down, shorten up and spend more of each step braking with a bent knee. The yellow line traces each hip, so you can compare the bounce.
Diagram
Two ways down the same slope
Schematic, not to scale. Ticks mark footfalls. In the study, less vertical bounce and shorter steps were each linked to keeping more strength after the descent. Source: Martinez-Navarro et al. (2026).
The cues
What to watch for
Look for the runner who bounds downhill early in a long run and then stiffens and slows on the later descents. Overstriding and a lot of up-and-down movement are the signs to coach out on long downhill sections, especially in the second half of a session or race.
Watch the days after a downhill session too. Soreness that lasts several days and a drop in jump or squat strength mean the legs haven't adapted yet. That's expected after the first few sessions.
All three studies used smooth treadmills. Real descents add loose rock, roots, mud and wet ground, which ask more of balance and ankle control. Build downhill work on smooth, runnable trails or fire roads first, and only add technical ground once the legs have adapted.
Build It Into Training
Muscle adapts quickly to downhill running. After one damaging session, the next one causes less damage. Researchers call this the repeated bout effect, and two recent studies show it in runners.
The repeated bout effect
Ten runners did two 30-minute downhill runs three weeks apart. After the second, soreness and the creatine kinase rise were smaller, and the nervous system held its drive to the muscle better, so they lost less strength. Running economy on the flat didn't change either way. Earlier research found the protection can last up to six weeks.
Ten trained female runners who'd never trained downhill did the same session twice, three weeks apart. Gluteal soreness was lower straight after the second session and at 24 and 48 hours. Blood lactate was lower during the second session too.
Diagram
First session vs second session, three weeks apart
Illustrative shape, not measured values. It shows the direction of the findings: in both repeated bout studies, the second session three weeks later caused less soreness, and in the men's study a smaller creatine kinase rise and less strength loss.
What the studies used
| Study | Who | Session | Main finding |
|---|---|---|---|
| Martinez-Navarro et al. (2026) | 36 experienced trail runners | 5 km at −15%, heart rate matched to uphill first threshold | All damage markers rose. Weekly downhill reps linked to less damage and kept squat strength. |
| Khassetarash et al. (2022) | 10 recreational male runners | 30 min at −20% and 10 km/h, repeated after 3 weeks | Second bout caused less soreness, a smaller creatine kinase rise and less strength loss. |
| Tallis et al. (2024) | 10 trained female runners, new to downhill | Six 5-min efforts alternating flat and −15% at 70% of speed at VO2peak, repeated after 3 weeks | Second bout caused less gluteal soreness for 48 hours. |
Swipe sideways to see the whole table.
Putting it into a programme
None of these studies tested a full training block, so there's no proven dose yet. Here's how we'd read them together.
- Start well before race season.The first downhill session hurts the most. Get it done months out, keep it short, and expect a few sore days.
- Repeat it within three weeks.Both repeated bout studies used a three-week gap and found clear protection. Don't leave it so long that the effect fades.
- Make it weekly.In the trail runner study, the protected group did downhill repetitions at least once a week in the three months before the race.
- Build the dose slowly.Add length or steepness one at a time, and only once the soreness after the last session has settled within a day or two.
- Keep it away from key sessions.Damaged legs lose strength for days. Don't put a downhill session the day before a hard workout or race.
Evidence and References
What the studies found
In the trail runner study, all three damage markers rose after one 5 km descent, weekly downhill repetitions were linked to less creatine kinase and better squat strength retention, and lower, shorter downhill strides were linked to keeping more strength. In the two repeated bout studies, a second downhill session three weeks after the first caused less soreness, and in the men's study less muscle damage and less strength loss.
How the studies were run
In the Khassetarash study, strength was measured as a maximal voluntary contraction of the knee extensors before and after each downhill run and at intervals over the following week. The smaller strength loss after the second run came with better-preserved neural drive, meaning the nervous system kept switching the muscle on more fully.
In the Tallis study, the women's oxygen uptake, heart rate and lactate were all lower running downhill than on the flat at the same speed. Effort ratings were higher on the first downhill effort, then lower for the rest.
What this research can't tell you
The trail runner study shows who coped better, not that downhill repetitions caused the difference, and the downhill group was small. All three studies ran on treadmills at fixed slopes, which is gentler on balance and footing than a rocky descent. The two repeated bout studies each used ten runners of one sex. Blood markers vary a lot between people and don't map neatly onto how sore someone feels or how they'll race. We still don't know the best slope, length or frequency for downhill training in experienced trail runners.
This is general guidance based on research. It doesn't replace individual advice from a coach or health professional. Warm up properly before any downhill session: see the RAMP Warm-Up guide.
References
- Khassetarash, A., Vernillo, G., Krüger, R.L., Edwards, W.B. and Millet, G.Y. (2022) Neuromuscular, biomechanical, and energetic adjustments following repeated bouts of downhill running. Journal of Sport and Health Science, 11(3), pp.319–329.
- Engeler, N., Lichtenstein, E., Faude, O. and Roth, R. (2025) How downhill and uphill running interfere with posture and muscle activity: a descriptive laboratory study. International Journal of Sports Physical Therapy, 20(8), pp.1186–1197.
- Lu, Z., Suo, B., Deng, L., Wang, J., Fu, W., Zhong, Y. and Jin, J. (2025) A review of uphill and downhill running: biomechanics, physiology and modulating factors. Frontiers in Bioengineering and Biotechnology, 13, 1690023.
- Martinez-Navarro, I., Vicente-Mampel, J., López-Grueso, R., Suarez-Alcazar, M.-P., Vilar-Fabra, C., Collado-Boira, E. and Hernando, C. (2026) Downhill running-induced muscle damage in trail runners: an exploratory study regarding training background and running gait. Sports, 14(1), 12.
- Tallis, J., McMorrow, C., Shelley, S.P. and Eustace, S.J. (2024) Repeated bout effect of downhill running on physiological markers of effort and post exercise perception of soreness in trained female distance runners. Sports, 12(6), 169.
Figures and citations checked against the published sources, October 2026. Next in the series: Hills, HIIT or the Gym?