REF-239 · Yamaguchi et al., 2023 · Journal of Biomechanics, Vol. 158, · Paper · PubMed

Power production strategy during steady-state cycling is cadence dependent

B+moderate evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

Participants

n=20 trained collegiate cyclists; 30-s steady-state pedaling at 50% of maximal anaerobic power at 90, 120, 150 and 180 rpm (pedal force measurement system + motion capture).

Protocol

Within-subject laboratory study: twenty trained collegiate cyclists, 30-s steady-state pedaling at 50% of maximal anaerobic power (from the abstract).

Four cadence conditions: 90, 120, 150 and 180 rpm.

Measures: pedal kinetics (pedal force measurement system) and limb kinematics (motion capture); positive and negative mechanical work of hip and knee and each joint's share of the total.

Key Results

Positive mechanical work of hip extension decreased significantly with increasing cadence (P<0.05)

Positive mechanical work of knee flexion increased significantly with increasing cadence (P<0.05)

At 150 rpm and above: the knee joint contributed >70% of total mechanical work, the hip <40%

At 180 rpm: positive hip work shifted to negative work — the authors call this a different power production strategy from general pedaling cadences

Mechanism

Inference, not tested: the abstract describes the shift in joint work with cadence; it does not name the muscles. Reading hip extension as glutes/hamstrings and knee flexion as hamstrings/gastrocnemius is anatomy, not a measurement from this paper. The authors' own conclusion is narrower: pedaling at 180 rpm uses a different strategy, so track cyclists need specific training at such excessive cadences.

Practical Application

n=20 trained collegiate cyclists, 30 s at 50% of maximal anaerobic power: as cadence rose from 90 to 180 rpm, positive hip-extension work fell and knee-flexion work rose (both P<0.05); at 150 rpm and above the knee delivered >70% of the work and the hip <40%, and at 180 rpm hip work turned negative — the same watts are produced by a different joint strategy at very high cadence.

Inference, not tested: in ERG mode the watts are fixed but the cadence is not — so the joint (and presumably muscle) strategy producing those watts still changes with the cadence the rider chooses. ERG standardizes output, not input.

Inference, not tested: if the target event needs hip-extension work (climbing, sustained TT at moderate cadence), high-cadence intervals train a different strategy; include cadence-specific work. The paper's own recommendation is the mirror image for track cyclists: train at 180 rpm because that strategy is specific.

Inference, not tested: for variable-terrain racing (gravel), covering the cadence range in training prepares both the hip-led and knee-led strategies.

Limitations

Very short efforts (30 s). Supranormal cadences (150-180 rpm) less applicable to typical training. Young collegiate cyclists. Single power level (50% of maximal anaerobic power). Joint work only — no EMG, so the muscle-level reading is inference.

How this entry was checked and changed

Content correction 2026-09-27

Journal line said Vol. 159, 111795 -> PubMed: J Biomech 158:111772.

Lead Practical bullet was an ERG inference -> now the study's finding (n, protocol, direction, >70% / <40%, 180 rpm sign change); the ERG, event-specificity and gravel bullets are labelled as inference.

Mechanism claimed which muscles take over -> the abstract reports joint work only; muscle attribution labelled as inference. Sport line kept as cycling.

2026-09-28: added Protocol from the abstract (Europe PMC, DOI 10.1016/j.jbiomech.2023.111772, PMID 37643551).