Pattern Generators in Locomotion: Implications for Recovery of Walking After Spinal Cord Injury

The role of the human spinal cord in complex processing of sensory information during locomotion has traditionally been thought to be minimal compared to other mammals. However, the evidence from experimental models of animals after spinal transection demonstrates that the spinal cord of at least some adult mammals can execute full weight bearing stepping independent of brain input. The interaction of sensory information derived from the moving limbs with central pattern generators in the spinal cord have been identified as important components of the neural control of locomotion. Some of the features of central pattern generation important for successful locomotion in animals include recognition of order of neural events, interpretation of patterns of sensory input in a state-dependent manner, and the availability of multiple neural circuits for successful stepping. If the human has similar potential to generate locomotion based on circuits within the spinal cord, then taking advantage of these mechanisms may enhance locomotor recovery after spinal cord injury (SCI). Recent evidence suggests that the human lumbosacral spinal cord can process complex sensory information during locomotion in the absence of detectable supraspinal influence. Also, activity-dependent plasticity is evident in spinal neural networks involved in the generation of stepping in humans. Locomotor training, a new approach to gait rehabilitation, is based on providing specific sensory information related to locomotion and repetitive practice usually on a treadmill with body weight support and manual assistance if needed. This approach takes advantage of the plasticity and motor learning capacity of the human spinal cord. Functional recovery of walking after SCI may be significantly improved by implementing locomotor training in new rehabilitative strategies that focus on enhancing recovery below the level of lesion.