Multisite Hebbian Plasticity Restores Function in Humans with Spinal Cord Injury

Objective Spinal cord injury (SCI) damages synaptic connections between corticospinal axons and motoneurons of many muscles, resulting in devastating paralysis. We hypothesized that strengthening corticospinal‐motoneuronal synapses at multiple spinal cord levels through Hebbian plasticity (i.e., “neurons that fire together, wire together”) promotes recovery of leg and arm function. Methods Twenty participants with chronic SCI were randomly assigned to receive 20 sessions of Hebbian or sham stimulation targeting corticospinal‐motoneuronal synapses of multiple leg muscles followed by exercise. Based on the results from this study, in a follow‐up prospective study, 11 more participants received 40 sessions of Hebbian stimulation targeting corticospinal‐motoneuronal synapses of multiple arm and leg muscles followed by exercise. During Hebbian stimulation sessions, 180 paired pulses elicited corticospinal action potentials by magnetic (motor cortex) and/or electrical (thoracic spine) stimulation allowing volleys to arrive at the spinal cord 1–2 milliseconds before motoneurons were activated retrogradely via bilateral electrical stimulation (brachial plexus, ulnar, femoral, and common peroneal nerves) for biceps brachii, first dorsal interosseous, quadriceps femoris, and tibialis anterior muscles as needed. Results We found in our randomized study that participants receiving Hebbian stimulation improved their walking speed and corticospinal function to a greater extent than individuals receiving sham stimulation. In agreement, prospective study participants improved their grasping and walking, corticospinal function, and quality of life metrics, exhibiting greater improvements with more sessions that persisted 9‐month post‐therapy. Interpretation Our findings suggest that multisite Hebbian stimulation, informed by the physiology of the corticospinal system, represents an effective strategy to promote functional recovery following SCI. ANN NEUROL 2023;93:1198–1213