Spastic co-contraction: A misdirected central command.

OBJECTIVE: Antagonist co-contraction triggered by voluntary agonist command in spastic paresis is poorly understood. Using an isometric ankle paradigm designed to minimize movement-related stretch reflexes, we aimed to characterize the severity, modulation, and neurophysiological origin of spastic co-contraction. METHODS: Thirteen subjects with chronic hemi/paraparesis (n = 14 limbs) were seated with the ankle at 90° dorsiflexion, and performed submaximal then maximal isometric voluntary ankle plantar- and dorsiflexion efforts, with the knee flexed then extended (gastrocnemii slack then stretched). Surface and intramuscular EMG recordings (the latter using hook wire electrodes) from agonist and antagonist muscles were combined with ankle torque measurements. Co-contraction severity was classified as absent, no antagonist EMG activity during agonist effort; mild, presence of antagonist EMG activity without torque reversal; or severe, antagonist EMG activity associated with torque in the unintended direction. RESULTS: Antagonist co-contraction occurred in all limbs during dorsi- and plantar flexor efforts. Its severity increased from submaximal to maximal efforts, whether in knee flexion (p = 0.046, Wilcoxon) or knee extension (p = 0.014). Severe plantar flexor co-contraction was more prevalent with the knee extended than flexed (64% vs 0%; p = 0.004, McNemar). Antagonist activity preceded agonist activation in 18% of efforts. In some intramuscular recordings, the first motor unit recruited in co-contractions was identified as that first-recruited when the same muscle was contracted voluntarily as agonist. CONCLUSIONS: Spastic co-contraction likely reflects misdirected descending supraspinal command, potentiated by stretch-related peripheral afferent signals.Significance. Spastic co-contraction is an effort-dependent, stretch-sensitive phenomenon distinct from spasticity, and can severely impair force generation in spastic paresis.