REF-827 · Balducci et al., 2017 · International Journal of Sports Medicine, Vol. · Paper · PubMed

Performance Factors in a Mountain Ultramarathon

Blimited evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

In 24 male ultra runners tested before and immediately after the Interlacs Trail (75 km, 3 930/3 700 m of ascent/descent), with maximal aerobic speed (MAS) assessed one week before, race time correlated with MAS (r = -0.74), the fraction of MAS sustained during the race (r = -0.89), knee-extensor force (r = -0.51) and knee-extensor force loss (r = -0.51).

A multiple regression using the fraction of MAS, MAS and knee-extensor force explained race time almost completely (R2 = 0.98). Contrary to the authors' expectations, performance was not correlated with the energy cost of level or uphill running, nor with how those costs changed after the race.

The authors conclude that training for a mountain ultramarathon should take into account knee-extensor strength, MAS and the fraction of MAS a runner can sustain.

Participants

n=24 male ultra runners (endurance mountain runners). The abstract does not give age, training volume or finish times.

Protocol

Maximal aerobic speed (MAS) assessed one week before the race.
Before and immediately after the Interlacs Trail (75 km; 3 930 m of ascent, 3 700 m of descent): energy cost of level and graded (10%) running, stiffness, knee-extensor force (KEf) and countermovement-jump height.
Correlations with race time; multiple regression on performance time in minutes.

Key Results

VariableCorrelation with race time
Fraction of MAS sustained in the racer = -0.89, p < 0.001
Maximal aerobic speedr = -0.74, p < 0.001
Knee-extensor forcer = -0.51, p < 0.05
Knee-extensor force lossr = -0.51, p < 0.05
Energy cost of level or uphill running (and its change after the race)not correlated
Regression: T = -11.852 × FMAS - 37.195 × MAS - 0.118 × KEf + 2090.581 (R2 = 0.98).

Practical Application

In 24 male runners of a 75 km mountain ultra, race time was best explained by the fraction of maximal aerobic speed they could sustain (r = -0.89), their maximal aerobic speed (r = -0.74) and knee-extensor force (r = -0.51) — and not at all by the energy cost of level or uphill running. For a mountain ultra, build the aerobic engine and the ability to hold a high share of it for many hours, and keep strong quadriceps; running economy measured fresh did not separate the fast from the slow.
Knee-extensor force loss also correlated with race time (r = -0.51), but the abstract does not say whether the loss was expressed as a positive amount or a negative change, so the direction (did faster runners lose more or less force?) cannot be read from it. What is clear is that higher knee-extensor force went with faster times; inference, not tested here: that fits the descent-damage evidence in REF-662 and REF-660.
The regression's R2 = 0.98 is a fit to these 24 runners, not a validated prediction tool; do not use the equation to predict an athlete's time.
Not from this paper; the base's own practice: heavy strength work for the quadriceps (SYNTH-44) and repeated downhill exposure (REF-782) are the ways to train the strength side; the study did not test either.

Limitations

Men only (n=24), one race; results may not carry over to women, other courses or longer ultras.
Correlational: strength and aerobic speed were not trained or manipulated.
The fraction of MAS sustained is partly an outcome of the race itself (faster runners by definition hold a higher fraction), which inflates its correlation with race time.
The regression was not validated in a separate group.
Written from the abstract and record metadata only (full text not read).
How this entry was checked and changed
Why this level

correlational, n=24 men, one race