Locomotion in patients with spinal cord injuries.

Key Words: Incomplete/complete paraplegia, Increased muscle tone, Monosynaptic/polysynaptic reflexes, Spinal locomotor centers, Training effects.

To control posture and gait, the central nervous system selectively utilizes afferent information from a variety of sources, Which then interact with central programs to adjust the movement to the actual requirements. Although a predominant working range exists for each receptor system, considerable overlap among systems is present. Thus, under normal conditions, stepping movements are not particularly impaired in the absence of one of the main afferent systems (ie, visual, proprioceptive, or labyrinthine).

There is general agreement that locomotor movements in mammals depend primarily on neuronal mechanisms within the spinal cord that can act in the absence of any afferent input. (See review by Griliner.(1)) Nevertheless, afferent information is essential for both bipeds and quadrupeds in order to adjust the central pattern to the external requirements.(1,2) Further refinement is achieved supraspinally, namely via the cerebellum. (See review by Armstrong.(3))

The neuronal regulation of human locomotion is similarly achieved by a complex interaction of spinal and supraspinal mechanisms.(4) The rhythmic activation of lower-extremity muscles by spinal interneuronal circuits is modulated and adapted accordingly by multisensory afferent input. Muscle activity, which results from the interaction of these different mechanisms, causes functionally modulated muscle tension by the mechanical muscle fiber properties.(5) The spinal programming as well as the reflex activity are under supraspinal control. When this supraspinal control is impaired (eg, in patients with spinal cord lesions), spasticity (hypertonicity) can develop. Spasticity produces numerous physical signs such as exaggerated reflexes, clonus, and muscle hypertonia. Spasticity has been defined as an increase in resistance to passive muscle stretch in a velocity-dependent manner following activation of tonic stretch reflexes.(6)

The degree of spasticity bears little relationship to the patient's disability, which is due to a movement disorder. In the past, a widely accepted premise was that exaggerated reflexes are responsible for spastic muscle hypertonia, and therefore for the movement disorder. The function of these reflexes during natural movements and the relationship between exaggerated reflexes and the movement disorder are not often considered. In patients with spinal cord lesions, a characteristic gait impairment is seen. This gait impairment can be evaluated by the recording of electrophysiological and biomechanical variables. There is some difference between spasticity of cerebral origin and spasticity of spinal origin, but the main features, such as the pattern of lower-extremity muscle activity during locomotion and the pathophysiology of increased muscle tone, are quite similar.(4)

Spastic Movement Disorder

Reflexes and Muscle Tone

Neuronal reorganization occurs following central lesions in both cats(7) and humans.(8) Novel connections (eg, sprouting, functional strengthening of existing connections, removed depression of previously inactive connections) may cause changes in the strength of inhibition. In addition, supersensitivity caused by the denervation may occur.(7) Although recent observations have indicated that spinal cord lesions do not cause sprouting of primary afferents in either cats(9) or humans,(10) changes in the reduction of presynaptic inhibition of group Ia fibers do occur,(11) which correlates with the enhanced excitability of tendon tap reflexes. In addition, reduction of presynaptic inhibition is stronger in patients with paraplegia compared with patients with hemiplegia.(12) No correlation, however, is seen between decreased presynaptic inhibition of la terminals and the degree of muscle tone measured by Ashworth's scale.(12) Nevertheless, treatment of spasticity is usually directed toward reducing stretch reflex activity, as it was thought that exaggerated reflexes were responsible for increased muscle tone and therefore spastic movement disorders.

The pattern of muscle activation and the development of increased muscle tone in people with spasticity, however, are basically different in the active motor condition compared with the clinical testing of the passive muscle.(13-15) During every movement, the muscle is active, and it is the movement disorder that hampers the patient. In contrast, during clinical examination of muscle tone (according to Ashworth's scale), the muscle is passive. Basically, the muscle behaves differently in response to stretch in the active and passive conditions. Extensive investigations on functional movements of lower-extremity(16,17) and arm(15,18,19) muscles have not shown any causal relationship between exaggerated reflexes and movement disorder. In adult patients with cerebral or spinal lesions, the reciprocal mode of lower-extremity muscle activation during gait is preserved in spasticity. Exaggerated stretch reflexes in persons with spasticity are associated with an absence or reduction of functionally essential polysynaptic (long-latency) reflexes.(16) Tension development during functional movements(16,20) does not depend on exaggerated stretch reflexes. The overall lower-extremity muscle activity is reduced in patients with spasticity.(16,18) For the assessment of changes in muscle electrical activity in subjects without neurological impairments, and to compare these changes with values obtained from patients with spasticity, the mean values of electromyographic (EMG) strength were determined for each group of subjects.(16,18) The sampling interval for the calculation of the mean values with standard deviations was 1/20 of one step cycle.(16,18)

Neuronal Regulation of Locomotion

During gait in patients with spasticity, the EMG activity in the calf muscles is smaller in amplitude and less well modulated compared with that of persons without neurological impairments (Fig. 1). This observation is most likely due to the impaired function of polysynaptic reflexes. Fast regulation of motoneuron discharge, which characterizes the normal muscle, is absent in patients with spasticity.(21,22) This fast regulation of motoneuron discharge corresponds to a loss of EMG modulation during gait. In patients with hemiparesis due to cerebral lesions, the strength of EMG activity in the affected lower extremity is often reduced compared with that of the unaffected lower extremity. This reduction in strength corresponds to the degree of paresis observed during both gait(16) and elbow movements.(18)

During gait in patients with spastic paresis (acquired at birth or later in life), a fundamentally different development of tension of the triceps surae muscle takes place during the stance phase.(16,17) For patients with spastic hemiparesis, tension development correlates with the modulation of EMG activity in the unaffected lower extremity (the same is true for individuals without neurological impairments), whereas tension development is correlated to the stretching period of the tonically activated (with small EMG amplitude) muscle in the spastic lower extremity (Fig. 2).(16,17) During gait, there is no visible influence of monosynaptic reflex potentials on the tension developed by the triceps surae muscle (for a description of tension recording, see Berger et al(16)). A similar discrepancy between the resistance to stretching and the level of EMG activity has been described for flexor muscles of the upper limb in patients with spasticity.(23,24)

Muscle tone during functional movements in patients with spastic paresis cannot be explained by increased activity of motoneurons, as assessed by EMG recordings. Instead, the motor units of the triceps surae muscle are transformed such that a higher-tension-to-EMG-activity relationship occurs during the stretching phase. Consequently, we believe that regulation of muscle tension takes place at a lower level of neuronal organization. Such a transformation is functionally meaningful because it enables the individual to support the body weight during gait. Fast active movements, however, become impossible.

Application of findings related to tone from studies of animals to humans must be done cautiously. An acute rigor appears in the decerebrate cat, whereas in individuals with paresis secondary to acute spinal or supraspinal lesions, muscle tone develops slowly over a period of weeks. Currently, no animal model exists to explain the development of hypertonicity in humans.

Motor Unit Transformation

Several findings support the suggestion that changes in the mechanical properties of muscle fibers occur in patients with spasticity. Contraction times of hand muscles(25) as well as of the gastrocnemius muscle(26) in patients with hemiparesis are prolonged. This is not true where there are spinal transections in cats(27) and humans.(28) Torque motor experiments applied to upper- and lower-limb muscles indicate a major, nonreflex contribution to the increased muscle tone in the antigravity muscles (ie, leg extensors and elbow flexors).(29,30) Histochemical and morphometric studies of spastic muscle have shown changes in the muscle fibers.(21,31) These changes include 1) increased levels of muscle fiber atrophy (especially of type II fibers), 2) a predominance of type I fibers in the gastrocnemius muscle 6 months after stroke, when spasticity of cerebral origin has been established (however, the soleus muscle of patients with long-standing paraplegia was shown to consist of type II fibers(28)), and 3) structural changes, such as the appearance of target fibers, mainly in type I fibers.

The alteration to a simpler regulation of muscle tension could be advantageous because it enables the person with paresis due to spinal or supraspinal lesions to support the body during gait and, consequently, to achieve mobility.(32) Rapid movements, however, are no longer possible due to the absence of modulation of muscle activity. Following a severe spinal or supraspinal lesion, these transformative processes can overshoot, with unwelcome consequences (ie, painful spasms and involuntarily movements).

Patients-with severe spinal or supraspinal lesions are usually much less able to perform stepping movements because the spinal locomotor centers do not provide sufficient basic activation of the leg muscles to perform functional movements such as gait. In recent years, it has been shown that specific locomotor training could enhance locomotor activity and thereby improve mobility. This approach will be discussed in the next section.

Locomotion in Patients With Paraplegia

Neuronal Capacity of Spinal Cord

For more than 30 years, functional electrical stimulation of paralyzed limb muscles was the only technique used to the improve the mobility of patients with paraplegia.(33) Despite technological developments, we believe that this method is still in at an experimental stage, and no breakthrough has occurred allowing for more extensive application of this technique.(34) The lack of a breakthrough has been due mainly to basic problems such as rapid development of muscle fatigue and adverse interactions with spinal reflexes, which cannot be overcome by technical means.(34) During the last few years, a new approach has been developed that improves the mobility of patients with paraplegia.(35-38) This improved mobility is achieved by external activation and training of spinal locomotor centers.(35-38)

In cats, recovery of locomotor function following spinal cord transection was believed to be restricted to immature animals.(39) The recovery of the locomotor patterns can be improved using regular training even in adult animals.(40) When spontaneous stepping was not stimulated, the cats lost their ability to step spontaneously.(40) During locomotor training, the animal was supported and thus only bore a part of its body weight. Locomotor movements of the hind limbs were induced by a treadmill while the forelimbs stood on a platform. With ongoing training, the body support was decreased in proportion to the animal's improving locomotor abilities. After several weeks of training, the cats were able to take more body weight and perform well-coordinated stepping movements.(35) The locomotor pattern at this stage closely resembles the pattern of the normal adult cat.(1) Thus, it can be concluded that the training represents an important factor for the recovery of locomotor function.

In humans, steplike movements are present at birth and can be initiated spontaneously or by peripheral stimuli.(41) The EMG activity underlying this newborn stepping has been shown to be centrally programmed and, as it has also been observed in children with anencephaly, it is likely that spinal mechanisms generate the EMG activity.(41) The apparent loss of locomotor movements in humans whose spinal cord was accidentally transected has been suggested to be due to a greater predominance of supraspinal over spinal neuronal mechanisms.(42) Nevertheless, there are indications that spinal interneuronal circuits exist that are involved in the generation of locomotor EMG activity in the leg muscles of humans,(3) similar to those described for cats.(35) Furthermore, involuntary steplike lower-extremity movements were recently described for a patient with an incomplete injury of the cervical spinal cord.(43)

Effect of Locomotor Training in Patients With Paraplegia

The aim of recent studies has been to determine the degree to which locomotor EMG activity and movements can be both elicited and trained in the leg muscles of patients with either complete or incomplete paraplegia.(36,37)

The induction of complex bilateral lower-extremity muscle activation combined with coordinated stepping movements in patients with incomplete and complete paraplegia was achieved by our approach of partially unloading (up to 60%) patients who were on a moving treadmill (Fig. 3).(36) To ensure consistent electrode placement over the course of training, initial electrode placement was marked by blue ink. The lower-extremity movements had to be assisted externally, especially during transitions from stance to swing, during the first phases of the training (usually 2-6 weeks, depending on the severity of paresis) in patients with incomplete paraplegia and during the whole training period (3-5 months) in patients with complete paraplegia. In comparison with subjects without neurological impairments, patients with paraplegia displayed a less dynamic mode of muscle activation (ie, it was less well modulated in amplitude) (Fig. 3). This finding may be due to the impaired function of polysynaptic spinal reflexes in patients with spinal lesions.(32) In other respects, such as the timing, the pattern of leg muscle EMG activity was similar to that seen in subjects without neurological impairments.(32) Different amplitude levels of leg extensor (the main antigravity muscle during gait) EMG activity occurred in the different subject groups, which largely exceeded the interindividual variability. The amplitude level of EMG activity was considerably smaller in patients with complete paraplegia compared with patients with incomplete paraplegia. Both patient groups had smaller EMG levels compared with those of subjects without neurological impairments (Fig. 3). Despite the reduced level of EMG activity, increased muscle tone and exaggerated reflexes were present in both patient groups. This finding supports earlier suggestions that alterations of mechanical properties of the tonically active muscle are mainly responsible for the clinical signs of spasticity.

The EMG patterns observed during locomotor training in patients with complete paraplegia may be due to rhythmic stretches of the leg muscles.(38) As shown in Figure 4, integrated EMG data for the tibialis anterior and gastrocnemius muscles were collected every 50 milliseconds over the whole step cycle and were related to the lengthening or shortening (determined by ankle and knee joint movements) of the respective muscle.(38) Leg muscle EMG activity was about equally distributed during muscle lengthening and shortening in both subjects without neurological impairments and patients with complete paraplegia during locomotion. This finding indicates that stretch reflexes are unlikely to play a major role in the generation of the leg muscle EMG pattern in these patients and that the locomotor pattern is programmed at a spinal level.

During the course of a daily locomotor training program, the amplitude of gastrocnemius muscle EMG activity increased during the stance phase, whereas an inappropriate tibialis anterior muscle activation decreased.(37) These training effects were seen in both patients with incomplete and complete paraplegia (Fig. 3). The training effects were related to a greater weight-bearing function of the extensors (ie, body unloading during treadmill locomotion could be reduced). The slope of the increase of gastrocnemius muscle EMG activity was similar in patients with incomplete and complete paraplegia. This finding indicates that the isolated human spinal cord contains the capacity not only to generate a locomotor pattern, but also "to learn." Only patients with incomplete paraplegia benefited from the waning program insofar as they learned to perform unsupported stepping movements on solid ground.(36,37) Patients with complete paraplegia experienced positive effects on the cardiovascular and musculoskeletal systems (ie, they had less severe spastic symptoms). In comparison with leg extensor EMG activity, only a small effect of training on leg flexor EMG amplitude was seen, which may be due to the different neuronal control of leg flexor and extensor muscles during locomotion.(44) The decreased unloading of the body (ie, successive reloading of the body) during the training may serve as a stimulus for extensor load receptors, which have been shown to be essential for leg extensor activation during locomotion in both cats(45) and humans.(46) The lower gain of extensor EMG activity (ie, the generally smaller EMG amplitude) in patients with complete paraplegia may be due to a loss of input from descending noradrenergic pathways to spinal locomotor centers.(35)

Practical Application of Locomotor Training

According to our experience over 4 years of locomotor training, the patient should be suspended by a parachute harness on the treadmill in a near-vertical body position to achieve both an axial direction of contact forces with the ground and extension of the hip during the stance phase. Depending on the severity of the spinal cord lesion, the pelvis may have to be fixed to avoid deviations in a lateral and posterior direction.

The amount of unloading should be maintained such that the maximal body load is taken over by the lower extremities but rhythmical stepping movements can still be performed. The locomotion speed should be adjusted such that an optimal rhythmicity occurs for each patient and also such that minimal assistance is required. For the assistance, an appropriate timing of stance and swing phases on either side should be achieved. Specific attention should be given to the timing of the double stance phase in order to achieve the optimal rhythmicity, which allows the patient to perform stepping movements with as little assistance as possible. The knee and hip joints should be fully extended sequentially during the stance phase to optimally support body weight. Nevertheless, hyperextension of the knees should be avoided to prevent damage of ligaments around the knee joint. The feet should pass, positioned closely together, under the vertical projection of the patient's center of gravity in order to obtain a full activation of the extensor muscles for push-off. The pelvis should be kept fixed so that loading of knee and ankle joints occurs in a physiological manner.

By unloading the body, the locomotor training can be started at an early stage of the rehabilitation program. The locomotion training with body unloading allows the performance of rhythmical stepping movements, which can easily be assisted and controlled by the attendant physical therapists. The unloading and assistance should be adapted to the actual condition and the severity of muscle paresis of the patient. A drawback of this training approach is that in patients with complete or almost complete paraplegia, the ergonomics necessary to assist the leg movements are difficult to perform on a long-term basis (ie, over weeks to months).

Pharmacological Influences on Locomotor Activity

A recent study(37) showed no training effects in about half of the patients with complete paraplegia who underwent daily locomotor training (ie, the amplitude of leg muscle EMC activity did not increase). The a posteriori analysis led to the suggestion that this finding might be due to the action of the drugs taken by the patients. Cannabinoids(47) and prazosin (an [alpha]-1 adrenoceptor antagonist(1)) are drugs that are known to inhibit spinal neuronal activity. In contrast, Barbeau et al(48) demonstrated in spinal cats that the administration of clonidine enhances locomotor activity. Nevertheless, a contrary effect was seen in two patients following intrathecal application of a low dose (25 [micro]g) of clonidine. This effect was also correlated with a loss of EMC activity and flaccid paresis.(37) Although a species-dependent action appears to be unlikely for this observation, there may be several explanations for the differential effects seen between cats and humans. For example, a dosage-dependent or "level of lesion"-dependent change of the drug action may explain these differential effects. The timing of drug application may also play a role because of changes of tile receptor. Clonidine was applied in the acute stage following spinal cord transection in cats,(48) whereas it was given several weeks posttrauma in human patients.(37) In contrast to clonidine, the sympathomimetic drug epinephrine was shown to have a positive effect on the locomotor pattern and performance.(37) This effect was correlated with an increase of gastrocnemius muscle EMC activity during the stance phase. This finding suggests that human spinal locomotor activity can be influenced pharmacologically. Further studies are needed to determine the degree to which human spinal locomotor activity can be positively influenced by other adrenergic agonists, thereby supporting the locomotor training. Further detailed and controlled studies are needed, particularly bearing in mind that neuronal changes may occur in untreated patients during the first months following spinal cord injury.

Outlook

The analysis of neuromuscular changes following cerebral or spinal motor lesions revealed that some of these changes (eg, development of increased muscle tone) can be advantageous if they provide body support during stepping movements. This knowledge could have the consequence that these neuromuscular changes could be targeted and become incorporated into physical therapy approaches for patients.(49) However, the efficacy of such approaches is lacking, and research is needed in riffs area.

In severely affected patients with spinal cord injury, the strength of leg muscle activation is not sufficient to build up enough underlying muscle tone or to control limb movements for locomotion. One approach to enhance spinal locomotor activity in the patients with incomplete and complete paraplegia represents the search for substances that influence the gain of leg extensor EMG activity. Although the substances that have been investigated to date have not been convincing in their action, other substances that exhibit a stronger effect might be explored in the future. Another approach concerns the release of spinal reflexes (eg, by the electrical stimulation of cutaneous nerves(50)), thereby physiologically enhancing and controlling leg muscle EMC patterns.(51) A promising approach for the future may be to induce partial regeneration of the lesioned spinal cord tract fibers. Recent experiments with rats have indicated that after inhibition of neurite growth inhibitors, partial regeneration can occur. (See review by Schwab.(52)) Combined with appropriate locomotor training, this approach may improve functional mobility, even that of patients with almost complete paraplegia.

References

1 Grillner S. Control of locomotion in bipeds, tetrapods, and fish. In: Brookhart M, Mountcastle VB, eds. Handbook of Physiology: The Nervous System; Volume II: Motor Control, Part 2. Washington, DC: American Physiological Society; 1981:1179-1935.

2 Gordon CR, Fletcher WA, Melvill Jones G, Block EW. Adaptive plasticity in the control of locomotor trajectory. Exp Brain Res. 1995; 102:540-545.

3 Armstrong DM. The supraspinal control of mammalian locomotion. J Physiol (Lond). 1988;405:1-37.

4 Dietz V. Human neuronal control of automatic functional movements: interaction between central programs and afferent input. Physiol Rev. 1992;73:33-69.

5 Gollhofer A, Schmidtbleicher D, Dietz V. Regulation of muscle stiffness in human locomotion. Int J Sports Med. 1984;5:19-22.

6 Lance JW. Symposium synopsis. In: Feldman RG, Young RR, Koella WP, eds. Spasticity: Disordered Motor Control. Chicago, Ill: Year Book Publishing Co Inc; 1980:485-495.

7 Mendell LM. Modifiability of spinal synapses. Physiol Rey. 1984;64: 260-324.

8 Carr LJ, Harrison LM, Evans AL, Stephens JA. Patterns of central motor reorganization in hemiplegic cerebral palsy. Brain. 1993;116: 1223-1247.

9 Nacimiento W, Mautes A, Topper R, et al. B-50 (GAP-42) in the spinal cord caudal to hemisection: indication for lack of intraspinal sprouting in dorsal root axons. J Neurosci Res. 1993;35:603-617.

10 Ashby P. Discussion I. In: Emre M, Benecke R, eds. Spasticity: The Current Status of Research and Treatment. Carnforth, United Kingdom: Parthenon; 1989:68-69.

11 Burke D, Ashby P. Are spinal "presynaptic" inhibitory mechanisms suppressed in spasticity? J Neurol Sci. 1972;15:321-326.

12 Faist M, Mazevet D, Dietz V, Pierrot-Deseilligny E. A quantitative assessment of presynaptic inhibition of Ia afferents in spastics: differences in hemiplegics and paraplegics. Brain. 1994;117:1449-1455.

13 Thilmann AF, Fellows SJ, Garms E. Pathological stretch reflexes on the "good" side of hemiparetic patients. J Neurol Neurosurg Psychiatry. 1990;53:208-214.

14 Thilmann AF, Fellows SJ, Garms E. The mechanism of spastic muscle hypertonus: variation in reflex gain over the time course spasticity. Brain. 1991;114:233-244.

15 Ibrahim IK, Berger W, Trippel M, Dietz V. Stretch-induced electromyographic activity and torque in spastic elbow muscles. Brain. 1993; 116:971-989.

16 Berger W, Quintern J, Dietz V. Tension development and muscle activation in the leg during gait in spastic hemiparesis: the independence of muscle hypertonia and exaggerated stretch reflexes. J Neurol Neurosurg Psychiatry. 1984;47:1029-1033.

17 Dietz V, Berger W. Normal and impaired regulation of muscles stiffness in gait: a new hypothesis about muscle hypertonia. Exp Neurol. 1983;79:680-687.

18 Dietz V, Trippel M, Berger W. Reflex activity and muscle tone during elbow movements in patients with spastic paresis. Ann Neurol. 1991;30:767-779.

19 Powers RK, Campbell DL, Rymer WZ. Stretch reflex dynamics in spastic elbow flexor muscles. Ann Neurol. 1989;25:32-42.

20 Berger W, Quintern J, Dietz V. Pathophysiology of gait in children with cerebral palsy. Electroencephalogr Clin Neurophysiol. 1982;53:538-548.

21 Dietz V, Ketelsen UP, Berger W, Quintern J. Motor unit involvement in spastic paresis: relationship between leg muscle activation and histochemistry. J Neurol Sci. 1986;75:89-103.

22 Rosenfalck A, Andreassen S. Impaired regulation of force and firing pattern of single motor units in patients with spasticity. J Neurol Neurosurg Psychiatry. 1980;43:907-916.

23 Powers RK, Marder-Meyer J, Rymer WZ. Quantitative relations between hypertonia and stretch reflex threshold in spastic hemiparesis. Ann Neurol. 1988;23:115-124.

24 Lee WA, Boughton A, Rymer WZ. Absence of stretch reflex gain enhancement in voluntarily activated spastic muscle. Exp Neurol. 1987; 98:317-335.

25 Young JL, Mayer RE. Physiological alterations of motor units in hemiplegia. J Neurol Sci. 1992;54:401-412.

26 Dietz V, Berger W. Interlimb coordination of posture in patients with spastic paresis: impaired function of spinal reflexes. Brain. 1984; 107:965-978.

27 Cope TC, Bodine SC, Fournier M, Edgerton VR. Soleus motor units in chronic spinal transected cats: physiological and morphological alterations. J Neurophysiol. 1986;55:1202-1220.

28 Shields RK. Fatigability, relaxation properties, and electromyographic responses of the human paralyzed soleus muscle. J Neurophysiol. 1995;73:1295-2205.

29 Hufschmidt A, Mauritz KH. Chronic transformation of muscle in spasticity: a peripheral contribution to increased tone. J Neurol Neurosurg Psychiatry. 1985;48:676-685.

30 Sinkjaer T, Toft E, Larsen K, et al. Non-reflex and reflex mediated ankle joint stiffness in multiple sclerosis patients with spasticity. Muscle Nerve. 1993;16:69-76.

31 Edstrom L Selective changes in the size of red and white muscle fibres in tipper motor lesions and parkinsonism. J Neurol Sci. 1970;11: 537-555.

32 Dietz V, Quintern J, Berger W. Electrophysiological studies of gait in spasticity and rigidity: evidence that altered mechanical properties of muscle contribute to hypertonia. Brain. 1981;104:431-449.

33 Kantrowitz A. Electronic Physiologic Aids: Report of the Maimonides Hospital. Brooklyn, NY: Maimonides Hospital; 1960:4-5.

34 Quintern J, Minwegen P, Mauritz K-H. Control mechanisms for restoring posture and movements in paraplegics. In: Allum JHJ, Hulliger M, eds. Afferent Control of Posture and Locomotion. Prog Brain Res. 1989;80(special issue):489-502.

35 Barbeau H, Rossignol S. Enhancement of locomotor recovery following spinal cord injury. Curr Opin Neurol. 1994;7:517-524.

36 Dietz V, Colombo G, Jensen L. Locomotor activity in spinal man. Lancet. 1994;344:1260-1263.

37 Dietz V, Colombo G, Jensen L, Baumgartner L. Locomotor capacity of spinal cord in paraplegic patients. Ann Neurol. 1995;37:574-582.

38 Dietz V. Locomotor training in paraplegic patients. Ann Neurol. 1995;38:965.

39 Forssberg H, Grillner S, Halbertsma J. Role in spinal cord locomotion of the low spinal cat, II: interlimb coordination. Acta Physiol Scand. 1980;108:238-295.

40 Barbeau H, Rossignol S. Recovery of locomotion after chronic spinalization in the adult cat. Brain Res. 1987;412:84-95.

41 Forssberg H. A developmental model of human locomotion. In: Grillher S, Stein PSG, Stuart DG, et al, eds. Neurobiology of Vertebrate Locomotion, Volume 45. London, United Kingdom: Macmillan Publishers Ltd; 1986:485-501.

42 Kuhn RA. Functional capacity of the isolated human spinal cord. Brain. 1950;73:1-51.

43 Calancie B, Needham-Shropshine B, Jacobs P, et al. involuntarily stepping after chronic spinal cord injury: evidence for a control rhythm generator locomotion in man. Brain. 1994;17:1143-1159.

44 Dietz V, Horstmann GA, Berger W. Interlimb coordination of leg-muscle activation during perturbation of stance in humans. J Neurophysiol. 1989;62:680-693.

45 Pearson KG, Collins DF. Reversal of the influence of group Ib-afferents from plantaris on activity in medial gastrocnemius muscle during locomotor activity. J Neurophysiol. 1993;70:1009-1017.

46 Dietz V, Gollhofer A, Kleiber M, Trippel M. Regulation of bipedal stance: dependency on "load" receptors. Exp Brain Res. 1992;89:229-231.

47 Meinck HM, Schonle PW, Conrad B. Effect of cannabinoids on spasticity and ataxia in multiple sclerosis. J Neurol. 1989;236:120-122.

48 Barbeau HJ, Julien C, Rossignol S. The effects of clonidine and yohimbine on locomotion and cutaneous reflexes in the adult chronic spinal cat. Brain Res. 1987;437:83-96.

49 O'Dwyer NJ, Ada L, Neilson PD. Spasticity and muscle contracture following stroke. Brain. 1996;119:1737-1749.

50 Tax AAM, van Wezel BMH, Dietz V. Bipedal reflex coordination to tactile stimulation of the sural nerve during human running. J Neurophysiol. 1995;73:1947-1964.

51 Jones CA, Yang JF. Reflex behaviour during walking in incomplete spinal cord injured subjects. Exp Neurol. 1994;128:239-248.

52 Schwab ME. Regeneration of lesioned CNS axons by neutralisation of neurite growth inhibitors: a short review. Paraplegia. 1991;29:294-298.

V Dietz, MD, is Professor, Doctor, and Chair/Head of Paraplegiology, Swiss Paraplegic Centre, University Hospital Balgrist, Forchstrasse 340, CH-8008 Zurich, Switzerland ([e-mail]). Address all correspondence to Dr Dietz.

M Wirz, PT, is Physiotherapist, Swiss Paraplegic Centre, University Hospital Balgrist.

L Jensen, DM, is Bioengineer, Swiss Paraplegic Centre, University Hospital Balgrist.

This work was supported by the Swiss National Science Foundation (No. 31-42.899.95) and the International Research Institute for Paraplegia (P16/93).