How the Central Nervous System Recovers — A Primer on Neuroplasticity

A person whose arm was paralyzed by a stroke picks up a spoon again several months later. A person with a spinal cord injury takes a few steps again after a long period of training. Severed nerves cannot be glued back together, and dead brain cells do not come back to life. Yet function returns.

Briefly answered, the brain and spinal cord recover function by rewiring surviving pathways rather than simply reconnecting broken ones. That foundation is neuroplasticity. We first met this term in the primer "Neuroplasticity — A Brain That Builds New Roads Beside the Broken Ones". Neuroplasticity is the ability of the brain to reorganize itself by forming new neural connections. That piece covered the distinction between recovery and compensation, the timeline of the first few months after injury, and the content and amount of practice.

This article goes to the next depth: why the nerves of the limbs regrow while the brain and spinal cord rely on rewiring, when plasticity occurs in a direction that actually prevents recovery, and where common misconceptions about plasticity go wrong.

Why We Need to Know — The Timetable After Emergency Treatment

Thrombolysis and thrombectomy, which we saw in the primer "What a Stroke Actually Is — Blocked, Burst, and the Clock", are treatments aimed at clots blocking blood vessels to save threatened brain tissue, and the time available is measured in hours. Even after passing that critical hurdle, aftereffects such as paralysis or speech disorders may remain. Instead of preventing damage, rehabilitation promotes plasticity within the surviving brain tissue, and the time available runs from days to weeks, and sometimes longer (PMID 32224759). Stroke rehabilitation is organized as a process using three principles of recovery: adaptation, restitution, and neuroplasticity (PMID 28157752).

Data showing that the adult brain can substantially reorganize its structure and function even after a stroke has accumulated over the past 10 years. There is also some recovery that happens spontaneously regardless of treatment, occurring especially within the first month, but its extent is limited (PMID 21600588). The question in rehabilitation papers is how to draw out more of this reorganization.

The Big Picture — Nerves in the Limbs Grow Back, While the Brain and Spinal Cord Rewire

The recovery principles of the nervous system differ depending on where the injury occurred. The central nervous system is the brain and the spinal cord. The spinal cord is a central nerve extending from the brain down inside the spinal column, serving as the pathway through which commands and sensations travel between the brain and the body. Peripheral nerves are the nerves that branch out from the brain and spinal cord to reach limbs and various parts of the body, and these nerves together are called the peripheral nervous system. That is the distinction used in the primer piece 「Paralysis and Spasticity — Where the Command Breaks Shapes What Follows」.

An axon is a branch that extends far from a nerve cell to carry signals over a distance, and is commonly called a nerve fiber.

Axons in the peripheral nerves grow back. When a nerve in the arm is severed, the nerve fiber farther from the cut site breaks down and its debris is quickly cleared away. This process is Wallerian degeneration, and this cleanup creates an environment where new fibers can grow (PMID 10219748). Schwann cells, which wrapped around the axon, line up to form a tube and guide the newly growing axon along that tube (PMID 33212795, PMID 10219748). Reinnervation is the process by which the regrown axon reconnects to a muscle (PMID 33212795).

This regeneration is slow. It grows 2 to 3 mm per day after a nerve is cut, and 3 to 4 mm per day after a crush injury (PMID 10219748). This means the farther the distance from the injured site to the muscle, the longer it takes to make contact again. Fibers responsible for movement preferentially seek out the muscle side again (PMID 10219748), but the grown axons fail to return to the tube they originally traveled through and to their original targets, which is why full functional recovery is rare (PMID 33212795). Recovery of peripheral nerves is covered in detail in 「How Peripheral Nerves Recover — An Introduction to Nerve Regeneration」.

The central nervous system, meaning the long nerve pathways of the brain and spinal cord, cannot grow back like this. Nerve pathways are routes formed by bundles of axons carrying signals in the same direction. The connection network of the central nervous system evolved to support broad functions while maintaining stability, and it possesses several 「brakes」 that preserve that stability. These are suppression originating from the extracellular environment, the intrinsic properties of nerve cells, and regulation that governs plasticity. These brakes, which maintained stability when healthy, turn into obstacles after an injury (PMID 32527334). The peripheral nervous system conversely has 「promoters」 that encourage regeneration and plasticity (PMID 32527334).

When the central nervous system is injured, scar-like tissue forms at the site, and debris from the material that wrapped the axons remains longer than in peripheral nerves. Both are external environments that block axons trying to grow. Beneath the differences between the two systems lies a trade-off. Instead of acquiring broad functions, the central nervous system has less power to adapt by creating new connections and new nerve cells (PMID 32527334).

Therefore, recovery after a stroke or spinal cord injury relies primarily on reorganizing and strengthening surviving circuits rather than newly reconnecting severed axons—in other words, neuroplasticity. Neuroregeneration is the process of creating new nerve cells and new connections to replace what was injured. Plasticity rearranges surviving elements to quickly regain function in the short and medium term, whereas neuroregeneration replaces what was lost over a long period to achieve long-lasting recovery (PMID 32527334). Recovery from peripheral nerve injuries relies heavily on regeneration, while recovery from central nervous system injuries centers on plasticity. Thus, 「neuroplasticity」 as encountered in rehabilitation is not the recovery principle for all nervous system injuries, and the recovery of someone with a severed nerve in the wrist is a different story.

There is limited regeneration in the central nervous system as well. In the corticospinal tract, which is the bundle of nerves running straight from the motor cortex to the spinal cord, electrical stimulation of the motor area causes axons to sprout new collateral branches (PMID 35034745). However, complete regeneration of entire long nerve pathways rarely occurs, and the main driving force of recovery remains plasticity.

How the Surviving Circuits Change — Five Paths

There was a time when people believed plasticity existed only in childhood. Today, we understand that plasticity continues throughout life and underpins learning, memory, and recovery after injury (PMID 40280532). Rehabilitation papers call these changes by five distinct names.

① Synaptic Plasticity — The Strength of the Seam Changes

A synapse is the seam where one nerve cell passes a signal to the next. Synaptic plasticity is the phenomenon where the strength of this seam increases or decreases depending on activity. If a state where the seam passes signals better lasts a long time after activity, it is called long-term potentiation (LTP); if a state where it passes them less lasts a long time, it is called long-term depression (LTD). These two are the main forms of synaptic plasticity that occur in the brain according to activity, and they are considered to show the cellular-level mechanisms of specific learning and memory (PMID 8717637). However, while LTD in the cerebellum, which refines movement, may be important for motor learning, the role that LTD in the hippocampus, which is involved in memory, plays in behavior is still being clarified (PMID 8717637).

In classical rat experiments, when a drug that blocks a specific receptor (the NMDA receptor) was administered, long-term potentiation in the hippocampus was blocked, and spatial learning was also impaired (PMID 2552039). This means that when the mechanism changing the seam is blocked, specific learning is blocked along with it. As long-lasting changes at the seam accumulate, the network takes on the shape of a memory or a complex motor task (PMID 35034736), and this is what rehabilitation training ultimately aims for.

② Cortical Reorganization — Redrawing the Body's Map

The cerebral cortex is the thin outer layer of the brain, spread with regions that divide up movement, speech, and sensation. Inside the cortex lies a map where each region is in charge of a different part of the body. Cortical reorganization is the redrawing of this map so that surviving areas take over the work that the injured area used to do. The brain was born with the ability to rewiring surviving networks to patch over structural damage (PMID 23774218).

fMRI is an MRI scan that shows which parts of the brain are active when it works. In a small study, as hand function improved in patients with cerebral infarction—that is, a blocked-side stroke—activity increased in the injured-side sensorimotor cortex and decreased in the uninjured-side activity. The research team viewed the cortical reorganization as happening side-by-side with motor recovery (PMID 12876462).

TMS is a test that touches the brain from outside the head with a magnetic field to see how it reacts. In a study that used this test to measure the map moving the hand, patients whose signals made it back down properly to the spinal cord (10 out of 27) had stronger grip strength the more their map had shifted (r=0.79). Across the group as a whole, map area and grip strength moved together too (r=0.52) (PMID 15294217). The research team concluded that this reorganization is functionally meaningful and contributes to motor outcomes (PMID 15294217). However, what this study measured was correlation.

③ The Bypass — Going Around Through the Paths That Remain

The motor cortex can reach the spinal cord directly through the corticospinal tract, or indirectly by using the brainstem—the lower stem connecting the brain and spinal cord—as a relay station. When a spinal cord injury destroys the corticospinal tract, this indirect pathway can bypass the damage and play an important role in controlling self-directed movement (PMID 35034745).

Even if it does not show up in clinical tests, if surviving connections cross the injured area, there is also room to draw out remaining functions through those connections (PMID 26843089).

④ Activity-Dependence — Rewired as Much as It Is Used

Activity-dependence is the property by which circuits are rewired as much as they are used and experienced, and it is also called experience-dependence. Among the properties of plasticity, it is the most important for rehabilitation. In animal models, a time-restricted window of plasticity opens after a stroke, and the largest share of recovery happens within it. Its mechanisms involve activity-driven rewiring and synapse strengthening (PMID 19888284).

There is a clue that the sheer amount of use is not enough. In the motor cortex of healthy animals, map shifts are accompanied by changes in synaptic shape, and these changes follow learning a new skill rather than simply using it a lot (PMID 12625638). This means that repeatedly raising an arm and learning a new meaningful task are inscribed differently in the brain, and task-oriented training, which will appear later, is rooted here.

⑤ The Critical Period — Even Flexibility Has Its Season

A critical period is a time window when the developing brain becomes exceptionally sensitive to specific experiences. This window is shaped not only by age but also by experience, and the possibility that it might reopen in adulthood has also been raised (PMID 15217343). The plasticity of an injured adult brain bears a striking resemblance to the plasticity of the developmental critical period. Both go through two phases: first, inhibition drops and connections fade, followed by a phase where branches sprout and connections multiply. However, the degree to which an injured adult brain can actually rewire itself and the time it takes remain major limits on recovery (PMID 24791715).

The primer called the brief period after an injury when the brain responds exceptionally strongly to training the sensitive period, and introduced CPASS as its evidence. This trial divided the timing of adding task-oriented arm practice into the acute phase (≤30 days), subacute phase (2 to 3 months), and chronic phase (≥6 months), and compared them against a control group that received only standard rehabilitation. When comparing arm function recovery over the year following stroke onset using an assessment score (ARAT), the subacute group (2 to 3 months) scored higher than the control group (+6.87 points), and the acute group (≤30 days) also scored higher, though by a smaller margin (+5.25 points). The chronic group showed no significant difference from the control group (PMID 34544853). The researchers view this as empirical evidence that humans, too, have a sensitive period for motor recovery (PMID 34544853).

What this trial compared was when the exact same amount of practice was added. The results for the chronic group only mean that the practice made no difference compared to the control group, and do not imply that improvement is impossible beyond that point.

The Clinical Landscape — Therapies That Intentionally Push Principles to Extremes

Stroke Motor Recovery

As seen in the primer, compensation strategies do not reduce the underlying impairment, but they get people home as early as possible. Functional recovery, on the other hand, reduces impairment longer, improves quality of life, and can last. Neurorehabilitation aims to drive this recovery by eliciting experience-dependent plasticity (PMID 31920570).

Task-oriented training is an approach where meaningful movements used in real life are practiced repeatedly as whole tasks, tailored to the person and their context. The rationale lies in the psychology of motor skill learning and the experience- and learning-dependent plasticity seen in ④ above (PMID 19504501).

Constraint-Induced Movement Therapy (CIMT) is a treatment where the unaffected arm is restrained to prevent its use, and the paralyzed arm is used intensively. When the paralyzed arm does not move well and a person comes to rely entirely on the unaffected arm, a state sets in where they stop using even the movement capacity they have left; this is learned non-use. CIMT produces effects through the interplay of two mechanisms. These are breaking learned non-use and driving cortical reorganization through use (PMID 12625639). Even in chronic stroke patients, after CIMT, the motor map measured by TMS expanded and its excitability increased, with the center of gravity of the map shifting to neighboring regions (PMID 9696052). This is a trace of treatment actually altering the injured hemisphere. Such neuroplastic changes appear to be important for the therapeutic effects of CIMT (PMID 24309263). "Able But Unused — An Introduction to Learned Non-Use and CIMT" covers this next.

Plasticity is not limited to movement. Visual fields lost to occipital stroke were long considered fixed and permanent after a brief period of natural improvement, but evidence has accumulated that visual training can recover a portion of that vision long after onset. However, this usually required long training lasting several months, and the recovered vision was not completely normal. The same training started in the subacute stage produced faster and broader recovery than when started in the chronic stage, and the authors suggest there is a window of opportunity in the early period (PMID 35034749).

Spinal Cord Injury and Gait Recovery

For a long time, spinal cord injury rehabilitation focused on bridging deficits with braces and wheelchairs under the premise that "the nervous system has fixed wiring and cannot be repaired." As evidence accumulated that the injured central nervous system can also change and learn, a paradigm shift occurred that views rehabilitation as the driving force leading to walking recovery (PMID 17012645).

Animal studies show that the spinal cord itself has neural networks that generate walking-like rhythm outputs, and these circuits can be activated by functional training that provides appropriate sensory feedback (PMID 23336934). Gait training is training that supports body weight as needed and repeats the action of walking on a treadmill or over ground. Among individuals with motor incomplete spinal cord injuries, meaning those who retain some voluntary movement strength below the injury, some regained walking through this training even several years after injury (PMID 28628595).

The effect heavily depends on the degree of injury. In motor complete paralysis, meaning when no such strength remains, the motor function below the injury goes unused and neural function declines (PMID 23336934). Epidural electrical stimulation is a method that places electrodes outside the membrane enclosing the spinal cord and passes electricity to wake up spinal circuits. Cases have been reported where individuals with motor complete paralysis regained several voluntary movements and standing while receiving this stimulation (PMID 28628595). The success or failure of both training and stimulation appears to depend on whether the spinal network maintains an excitation state appropriate for the task (PMID 28628595).

Applying epidural electrical stimulation alongside rehabilitation to 9 chronic spinal cord injury patients restored walking. Using a mouse model, the researchers even pinpointed the population of neurons essential for that recovery (PMID 36352232).

In mice with incomplete spinal cord injuries, ground walking training increased plasticity-related proteins in the motor cortex (such as synaptophysin and BDNF), which was interpreted as supporting better gait recovery (PMID 31276690).

Common Misconceptions — Five Illusions About Plasticity

Plasticity is a fascinating concept, so it is easily exaggerated. We match five common illusions against the evidence.

① "Plasticity Always Happens in the Good Direction"

Maladaptive plasticity is plasticity that occurs in a direction that actually hinders recovery. After a stroke, three are representative (PMID 22792492).

The first is compensatory movement. Relying on the less-injured side helps manage daily life on one's own, but it creates mismatched movement patterns that ultimately limit normal movement. The second is ipsilateral motor pathways. One cerebral hemisphere mainly moves the opposite side of the body, but ipsilateral pathways activated after injury cannot adequately replace the severed corticospinal tract, leading to low motor capacity and subsequent abnormal movements. The third is interhemispheric competition. This competition creates abnormal interhemispheric inhibition that weakens motor function. When inhibition is broadly released, competition also arises between the hand and the upper arm, increasing the risk that recovery remains incomplete (PMID 22792492). The authors suggest that to reduce this, rehabilitation programs should be tailored to the patient's degree of motor impairment, and non-invasive brain stimulation may be useful for correcting it (PMID 22792492).

Animal experiments show one such mechanism. When the intact forefoot was made to learn a skill, plasticity in the motor cortex responsible for that forefoot increased, but functional impairment of the paralyzed forefoot worsened, and the effect of rehabilitation training was reduced. This detrimental effect depended on the connections linking the two hemispheres and the contralateral motor cortex, and was linked to reduced activity in the peri-infarct motor cortex during rehabilitation training (PMID 23709698). The authors concluded that learning to rely on the intact side does more than just cause the paralyzed side to be used less; it can disrupt the function of an area in the injured hemisphere. They added that whether this generalizes to other injury sites or other functions has not yet been verified (PMID 23709698). That does not mean compensation is useless. Compensation is also a way to send people home early (PMID 31920570).

In the spinal cord as well, plasticity is a double-edged sword. Spinal junctions share a significant portion of the brain's learning and memory mechanisms, changing in response to activity even after complete spinal cord injury where inputs from the brain are cut off. In an animal experiment by one research team, when stimulation was matched to leg posture or given at regular intervals, plasticity occurred in a direction that aided subsequent learning. Random electrical stimulation or out-of-sync stimuli like skin wounds produced maladaptive plasticity that degraded subsequent learning and walking recovery while increasing pain hypersensitivity (PMID 23087647). Following a spinal cord injury, as downward activity is cut off, spinal interneurons degenerate, leading to hyperactive reflexes, muscle spasms, and maladaptive changes that form the basis of spasticity (PMID 35034745). Spasticity, where resistance grows stronger the faster a muscle is stretched, is continued in "One Word, "Spasticity" — Reading the Upper Motor Neuron Syndrome Apart".

Plasticity has direction, meaning the exact same capacity can both aid and block recovery. Therefore, plasticity is something to be carefully tuned rather than simply unleashed (PMID 23087647).

② "The Sooner and Harder, the Better"

The answer is closer to "we don't know yet." There is an experiment where one side of a rat's brain was injured, a cast was put on the healthy front leg, and the paralyzed front leg was forced into overuse. When forced into overuse during the first week, even the surviving surrounding tissue was damaged, the lesion grew larger, and functional recovery was severely hindered. When forced into overuse in the second week, the lesion did not grow larger, but functional recovery was still hindered. This means that about 2 weeks after injury is a vulnerable period for prolonged, forced overuse (PMID 9507166). What this experiment tested is a condition where extremely high demands were forced onto the injured brain for a long time.

In humans, rehabilitation started at a very early stage has not clearly improved outcomes. Meanwhile, evidence is growing that shorter and more frequent treatments can be safely started within the first day or two after onset. Since treatments stronger than standard intensity have not been linked to better outcomes, the optimal treatment dose has not yet been established (PMID 28157752). There is evidence in the opposite direction as well. In activity-dependent therapies like CIMT, recent studies suggest that very high doses are needed to obtain the best functional gains (PMID 32224759), and aerobic exercise can boost plasticity after a stroke, with that response being more pronounced in moderate-to-high-intensity programs, but results vary across studies and measurement methods differ, making it still too early to conclude (PMID 34669820). The aforementioned CPASS showed a prime window of the subacute phase rather than the chronic phase (PMID 34544853).

What is certain is that blindly going fast and hard is not the answer, nor is it useless—the timing and method matter. The amount of practice is continued in "How Much Is Enough — An Introduction to Dose-Response in Rehabilitation".

③ "Just Move a Lot"

As we saw in ④ above, structural changes in the motor cortex follow learning new skills rather than simply using it a lot (PMID 12625638). This is why rehabilitation emphasizes meaningful tasks and goal-directed practice.

④ "Recovery Means Returning to How Things Were"

Whether an outwardly positive improvement is the recovery distinguished in the primer (the same way as before the injury) or compensation (a different way) is a long-standing core question in rehabilitation research. Behavioral changes and plastic changes in brain structure follow each other bidirectionally, and the intensity of rehabilitation experience can also affect recovery (PMID 28512585, PMID 31920570). In spinal cord injury rehabilitation, to evaluate new recovery treatments, a separate scale was created to measure performance achieved without compensation (PMID 28628595). The deeper story is in "How Deep Is the Recovery — Between Compensation and True Recovery".

⑤ "The Principles Are the Same Everywhere"

Recovery principles vary depending on the site, the degree of injury, and the species. In a model where the spinal cord of a cat was completely transected, standing and walking returned even without task-specific training. Cats that simply had their calf muscles rhythmically stimulated by hand or received no treatment at all stood and walked with full weight-bearing. The researchers concluded that the spinal cord networks governing standing and walking largely remain intact even after complete transection, and that recovery is the result of restored excitability in spinal circuits rather than task-specific plasticity (PMID 31825306).

This result also means that what works in animals may not transfer directly to humans. Neuroplasticity is not a master key, but a complex phenomenon that operates differently depending on the conditions.

To Go Deeper

• Principles of plasticity: A classic review outlining the 10 principles of experience-dependent plasticity (PMID 18230848) and 15 principles of neurorehabilitation based on motor learning (PMID 31920570) bridge the foundation and the clinic. In this series, "Why Therapy Is Designed That Way — An Introduction to Motor Learning Principles" takes up the thread, and "When Does Recovery Happen — What Time Did, and What Treatment Did" takes up the share of time.
• Neuromodulation technologies: Tools targeting plasticity such as TMS, transcranial direct current stimulation (tDCS), and epidural or vagus nerve stimulation are advancing, and research combining training and stimulation in particular is active (PMID 26843089, PMID 36352232).
• The boundary between central and peripheral: The question of why peripheral nerves regenerate while central nerves are difficult leads to research trying to understand and release the "brakes" that suppress regeneration (PMID 32527334).
• Limits and unresolved issues: The problem of translating from animals to humans, distinguishing between recovery and compensation, and controlling maladaptive plasticity are the current forefront of rehabilitation science (PMID 28512585, PMID 23087647).

Neuroplasticity overturned the old belief that "the brain is fixed," but that optimism has limits. Plasticity rides on conditions, rides on direction, and rides on timing. Knowing this flexibility and its limits together is the primer for properly reading a single rehabilitation paper.

This article is for informational purposes only and does not substitute for diagnosis or treatment. The rehabilitation direction for individual patients must always be discussed with the attending medical team.

Source Papers

The Big Picture of Recovery (Acute vs. Rehabilitation, Spontaneous Recovery)

• Stroke recovery review (PMID 32224759)
• Stroke rehabilitation review (PMID 28157752)
• Neurological foundations of recovery after brain injury (PMID 21600588)

CNS/PNS Boundary (Plasticity vs. Axonal Regeneration and Reinnervation)

• Neuroregeneration and plasticity — physiological mechanisms of functional recovery after injury (PMID 32527334)
• Peripheral nerve regeneration and muscle reinnervation (PMID 33212795)
• Basics of nerve injury, axonal degeneration, and regeneration (Wallerian degeneration) (PMID 10219748)
• Neuroplasticity and repair in spinal cord injury (PMID 35034745)

Core Concepts — Synaptic Plasticity, Cortical Reorganization, Activity-Dependence, Critical Periods

• Basic mechanisms of plasticity and learning (PMID 35034736)
• Long-term depression (LTD) — learning-related synaptic plasticity (PMID 8717637)
• NMDA, LTP, and spatial learning (AP5) (PMID 2552039)
• Cortical reorganization after stroke: how much and how functional (PMID 23774218)
• Motor recovery and cortical reorganization in hemiparetic patients (fMRI) (PMID 12876462)
• Subcortical stroke: correlation between motor prognosis and cortical reorganization (TMS) (PMID 15294217)
• Neuroplasticity — current breakthroughs (PMID 40280532)
• Ten principles of experience-dependent neuroplasticity (PMID 18230848)
• Plasticity in stroke recovery: from synapse to behavior (PMID 19888284)
• Regulation of critical periods (PMID 15217343)
• Plasticity after adult injury — recapitulation of critical period mechanisms? (PMID 24791715)
• CPASS clinical trial (sensitive periods in humans) (PMID 34544853)

Clinical — Stroke Motor Recovery, CIMT, Spinal Cord Walking

• Principles of neurorehabilitation based on motor learning and brain plasticity (PMID 31920570)
• Evidence and clinical application of task-specific training (PMID 19504501)
• CIMT — a treatment for movement disorders utilizing plasticity (PMID 24309263)
• Motor recovery and extensive cortical reorganization after CIMT (PMID 12625639)
• Motor cortex plasticity during CIMT (TMS mapping) (PMID 9696052)
• Functional and structural plasticity of the motor cortex (skill-dependence) (PMID 12625638)
• Visual rehabilitation in cortical blindness (extension of CNS plasticity) (PMID 35034749)
• Paradigm shift in spinal cord injury rehabilitation (PMID 17012645)
• Physiological basis of locomotor training in spinal cord injury (PMID 23336934)
• Activity-based therapy for spinal cord injury (PMID 28628595)
• Neurons that restore walking after paralysis (epidural stimulation) (PMID 36352232)
• Epidural stimulation of lumbar spinal cord circuitry (PMID 26843089)
• Locomotor training and cortical plasticity in incomplete spinal cord injury (animal) (PMID 31276690)
• Aerobic exercise and neuroplasticity after stroke (PMID 34669820)

Common Misconceptions — Maladaptive Plasticity, Timing, Skill-Dependence, Compensation

• Maladaptive plasticity after stroke (PMID 22792492)
• Motor plasticity in stroke models: inherently use-dependent, but labile (PMID 23709698)
• Maladaptive plasticity of the spinal cord (PMID 23087647)
• Exacerbation of injury by overuse during the vulnerable early post-injury period (PMID 9507166)
• Understanding mechanisms of recovery and compensation (PMID 28512585)
• Recovery of standing and walking does not require task-specific training (cat) (PMID 31825306)