Stretch versus shortening contractions subsequently decrease versus increase neural drive to the human tibialis anterior.

EMG-based muscle force predictions are often inaccurate following active muscle stretch or shortening because of residual force enhancement (rFE) or depression (rFD), respectively, which can alter the neural drive to a muscle. However, the extent of neural drive modulation attributable to rFE or rFD remains unknown owing to the limited spatial sampling of intramuscular EMG. Therefore, we combined two blind-source separation approaches with high-density surface EMG and mixed-effects modelling to assess changes in neural drive from more representative motor unit samples. Seventeen participants performed dorsiflexion contractions at 20% and 40% of maximum voluntary torque in three conditions: stretch-hold, shortening-hold and fixed-end reference (REF) conditions. The ankle dorsiflexion torques and angles were matched using dynamometry to the REF condition during a steady state following a 1-s 25° stretch or shortening, during which we decomposed tibialis anterior high-density EMG recordings. Normalised EMG amplitudes were ∼2% lower following stretch and 1% and 3% higher following shortening relative to REF at 20% and 40% of maximum voluntary torque (P ≤ 0.008), respectively. Discharge rates from 19 matched motor units per person on average obtained via DEMUSE and MUedit were similar (P = 0.666). Following stretch and shortening, discharge rates were ∼1 Hz lower (P ≤ 0.029) and 0 (P = 0.459) to 1 Hz higher (P < 0.001) relative to REF, respectively. More unique motor units were also detected following shortening. These findings indicate that to account for rFE or rFD, neural drive is respectively decreased or increased via reduced or additional motor unit recruitment and discharge rate modulation, with a contraction intensity-specific rate modulation following active shortening.