Ultrasound reveals negligible cocontraction during isometric plantar flexion and dorsiflexion despite the presence of antagonist electromyographic activity

Because of the approximate linear relationship between muscle force and muscle activity, muscle forces are often estimated during maximal voluntary isometric contractions (MVICs) from torque and surface electromyography (sEMG) measurements. However, sEMG recordings from a target muscle may contain cross-talk originating from nearby muscles, which could lead to erroneous force estimates. Here we used ultrasound imaging to measure in vivo muscle fascicle length (Lf) changes and sEMG to measure muscle activity of the tibialis anterior, medial gastrocnemius, lateral gastrocnemius, and soleus muscles during ramp MVICs in plantar and dorsiflexion directions (n = 8). After correcting longitudinal Lf changes for ankle rotation, the antagonist Lf at peak antagonist root-mean-square (RMS) amplitude were significantly longer than the agonist Lf at this sEMG-matched level. On average, Lf shortened from resting length by 1.29 to 2.90 mm when muscles acted as agonists and lengthened from resting length by 0.43 to 1.16 mm when muscles acted as antagonists (depending on the muscle of interest). The lack of fascicle shortening when muscles acted as antagonists indicates that cocontraction was likely to be negligible, despite cocontraction as determined by sEMG of between 7 and 23% MVIC across all muscles. Different interelectrode distances (IEDs) over the plantar flexors revealed significantly higher antagonist RMS amplitudes for the 4-cm IEDs compared with the 2-cm IEDs, which further indicates that cross-talk was present. Consequently, investigators should be wary about performing agonist torque corrections for isometric plantar flexion and dorsiflexion based on the antagonist sEMG trace and predicted antagonist moment.