REF-1731 · Seynaeve et al., 2026 · Medicine and Science in Sports and Exercise · Paper · PubMed

After one night without sleep, 16 recreational runners ran to exhaustion at their second lactate threshold 13.2% sooner than after normal sleep, although their gross metabolic power at the same submaximal speeds was lower (better economy) and their gait changed (lower cadence, longer ground contact, lower leg stiffness, more knee and less ankle work), in a randomised crossover (abstract only)

Main caveat: One small laboratory study; the abstract gives P values and directions only, apart from the 13.2% time-to-exhaustion change.
limited evidenceMeasured on recreational to trained athletesRead from the abstract only

Seynaeve, Vanwanseele and de Beukelaar asked how a night without sleep changes running mechanics and economy.

Sixteen healthy recreational runners (7 men, 9 women) did the same treadmill protocol twice, in random order, once after a night of total sleep deprivation and once after normal sleep: three submaximal stages at 90% of the first lactate threshold (LT1), at LT1 and at 90% of the second (LT2), then a stage at LT2 run to exhaustion.

Gait was measured with an instrumented treadmill and 3D motion capture, metabolic rate by indirect calorimetry.

After the sleepless night the runners took fewer steps per minute, stayed longer on the ground, had a higher duty factor, lower peak vertical force and loading rate, more knee flexion and knee extensor moment, more positive knee power and less positive ankle power, and lower leg stiffness.

Gross metabolic power at a given speed was lower, which the authors read as better running economy, yet time to exhaustion at LT2 was 13.2% shorter. The authors conclude that sleep deprivation shifted work from the ankle to the knee and lowered leg stiffness, and that despite the apparent efficiency gain the runners fatigued sooner at LT2.

Participants

n = 16 healthy recreational runners, 7 M / 9 F.
Age, training volume, running performance, habitual sleep and how thresholds were set: not in the abstract.

Protocol

Randomised crossover: one night of total sleep deprivation (SDEP) versus normal sleep (CON).
Treadmill stages at 90% of LT1, LT1 and 90% of LT2, then an interval to exhaustion at LT2.
Measures: spatiotemporal, kinematic and kinetic data (instrumented treadmill, 3D motion capture); metabolic rate (indirect calorimetry, reported as gross metabolic power).
Not in the abstract: stage durations, time of day of testing, how wakefulness was supervised, food and caffeine overnight, the washout between conditions, perceived exertion, heart rate, and whether gross metabolic power fell at every stage.

Key Results

Measure, SDEP vs CONDirectionP
Cadencelower0.030
Ground contact timelonger< 0.001
Duty factorhigher0.012
Peak vertical ground reaction forcelower0.002
Loading ratelower0.011
Knee flexion angle in stancelarger0.015
Peak knee extensor moment in stancelarger0.037
Average positive knee powerhigher0.019
Average positive ankle powerlower0.019
Leg stiffnesslower0.028
Gross metabolic power at a given speedlower< 0.001
Time to exhaustion at LT213.2% shorter0.023
No means, standard deviations, effect sizes or intervals are given in the abstract beyond the 13.2%.

Practical Application

One night without sleep cut how long recreational runners could hold their second-threshold pace by 13.2%, even though easier running cost them less energy than after normal sleep. Only the final interval measured endurance; the submaximal stages measured energy cost and gait, not performance.
Against SYNTH-07: the guide keeps the plan after one short night and lets the warm-up decide the key session. This study tested a full night without sleep, not a short night, and did not test a warm-up decision. It agrees with the guide that sleep loss costs same-day performance (REF-204).Inference, not tested here: if the warm-up is judged by how easy submaximal running feels or costs, it may not reveal the loss, because submaximal metabolic cost here was lower while threshold endurance fell; the first hard reps are the better test, as the guide already says.
Inference, not tested here: a run after a sleepless night came with a shorter, softer stride (lower cadence and leg stiffness, more knee work); whether that changes injury risk was not studied.
Not from this paper: after partial sleep loss or several short nights, see REF-204 and REF-943.

Limitations

One small laboratory study; the abstract gives P values and directions only, apart from the 13.2% time-to-exhaustion change.
Total deprivation for a night is a stronger exposure than the short nights most athletes have.
Lower gross metabolic power at the same speed is an unusual finding after sleep loss; the abstract gives no explanation (for example, a lower resting metabolic rate or substrate use), and whether it held at every stage is not stated.
The time to exhaustion came after three submaximal stages, so it measures threshold endurance in a pre-fatigued state, not a race.
Perceived exertion, motivation and heart rate are not in the abstract.
Written from the abstract and the PubMed record; full text not read.
How this entry was checked and changed
Why this level

a randomised crossover of 16 (the rubric's B+ band for crossovers of 15 or more) held at B: the abstract gives no magnitudes except the 13.2% time-to-exhaustion change, no variability and no intervals; time to exhaustion, economy and force-plate mechanics are outcomes or direct measures, not markers