Energy Cost of Running at Extreme Uphill and Downhill Slopes
The basis of grade-adjusted pacing
Protocol & Participants
n=10 runners; treadmill inclined from −45% to +45% (−0.45 to +0.45) at several speeds; cost of walking (Cw) and running (Cr) measured in J/kg/m
Key Results (abstract)
Running, level: Cr = 3.40 ± 0.24 J/kg/m, independent of speed
Running, downhill: Cr falls to a minimum of 1.73 ± 0.36 J/kg/m at −20%; on steeper descents it rises again, to 3.92 ± 0.81 J/kg/m at −45%
Running, uphill: Cr rises to 18.93 ± 1.74 J/kg/m at +45% (≈5.6× the level cost)
Walking: minimum Cw 1.64 ± 0.50 J/kg/m at 1.0 ± 0.3 m/s on the level; 0.81 ± 0.37 at −10%; 3.46 ± 0.95 at −45%; 17.33 ± 1.11 at +45%
Above +15% and below −15% the mechanical efficiency of both gaits reached that of pure concentric and eccentric muscle contraction, respectively
Optimum gradient for mountain paths ≈ 20-30% for both gaits; downhill Cr was some 40% lower than literature values for sedentary subjects
Estimated maximal running speeds on positive gradients matched those adopted in uphill races; on negative gradients they were well above the speeds attained in downhill competitions
Practical Application
In 10 runners on a treadmill from −45% to +45%, the energy cost of running was 3.40 J/kg/m on the level, fell to a minimum of 1.73 J/kg/m at −20% and rose again to 3.92 at −45%; uphill it reached 18.93 J/kg/m at +45% (≈5.6× level). This curve is the physiological basis of grade-adjusted pace: judge effort off-road by grade-adjusted pace, not raw pace
Moderate descents (around −20%) are where running is cheapest — "free" speed; steeper than that, braking makes each metre cost more again
On steep climbs walking is cheaper than running (at +45%: 17.33 vs 18.93 J/kg/m); the switch point is individual — walk the steep pitches without guilt
Plan nutrition/effort by vertical gain, not just distance. Inference from the cost curve
Limitations
n=10, treadmill (not real technical ground); extreme gradients model mountains but not all conditions
Trained runners: downhill cost was ~40% lower than reported for sedentary subjects — the curve is population-dependent
The walking/running cost crossover is individual
How this entry was checked and changed
Why this level
regraded 2026-09-27 from A: descriptive measurement study, n=10
Content correction 2026-09-27
"Cr minimum at ~−10% downhill (~1.7-2.6 J/kg/m)" -> abstract: running minimum 1.73 J/kg/m at −20% (the −10% value, 0.81 J/kg/m, is for WALKING)
"At −45% Cr rises to ~3.5" -> 3.92 J/kg/m (3.46 is the walking value)
Added the walking values, the ±15% efficiency finding, the 20-30% optimum path gradient and the 40%-below-sedentary comparison
Level A -> B (measurement study, n=10); Sport line "trail-running, running" -> "trail running, running"