- Email: [email protected]
Neuronal function in chronic spinal cord injury: Divergence between locomotor and flexion- and H-reflex activity
Clinical Neurophysiology 117 (2006) 1499–1507 www.elsevier.com/locate/clinph
Neuronal function in chronic spinal cord injury: Divergence between locomotor and flexion- and H-reflex activity R. Mu¨ller *, V. Dietz Spinal Cord Injury Center, Balgrist University Hospital, Forchstrasse 340, 8008 Zurich, Switzerland Accepted 30 March 2006 Available online 11 May 2006
Abstract Objective: Knowledge about the long-term course of spinal neuronal function after spinal cord injury (SCI) is important if regeneration therapies become available in the future. The objective of this study was to examine the behavior of locomotor EMG activity and of spinal reflexes in patients with chronic motor-complete SCI. Methods: EMG activity from rectus femoris (RF), biceps femoris (BF), medial gastrocnemius (GM) and tibialis anterior (TA) of both sides was investigated during locomotor movements assisted by a robotic device in 10 chronic (O1 year after accident) complete SCI and 5 healthy subjects. H-reflexes (recorded from GM) were induced during the onset and flexion reflexes (recorded from BF and TA) at the end of the stance phase. Results: Only in the chronic SCI subjects an exhaustion of EMG activity—i.e. a decrease in amplitude—occurred within a few minutes in all leg muscles. The EMG exhaustion was not associated with a change in the H- or flexion reflex amplitude during a walking session. Conclusions: Exhaustion of neuronal function in chronic complete SCI might be restricted to unused motor tasks, i.e. locomotion. The fact that H- and flexion reflexes show a normal behavior might be due to the fact that they still become activated. Significance: Training/pharmalogical approaches may be required to maintain neuronal function of unused tasks as a basis for future successful regeneration therapies. q 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Locomotor activity; Reflex activity; Spinal cord injury
1. Introduction The isolated spinal cord contains neuronal circuits underlying the generation of basic complex movements such as locomotion and flexion reflexes in cats (de Leon et al., 1998) and humans (for review see Dietz, 2002). However, the human studies have largely been based on recordings taken early after a spinal cord injury (SCI) (Andersen et al., 2004; Dimitrijevic and Nathan, 1968, 1970), and little is known about the long-term adaptive changes of spinal neuronal circuits after a complete SCI. Knowledge about the course of long-term spinal neuronal function in humans is important if regeneration therapies become available in the future. This indeed seems to be * Corresponding author. Tel.: C41 44 386 37 30; fax: C41 44 386 37 31. E-mail address: [email protected] (R. Mu¨ller).
feasible in the coming years (Raineteau and Schwab, 2001; Raisman, 2003; Schwab, 2004). The benefits of any such therapies rely crucially on the remaining function of neuronal spinal circuits. There is evidence for a reduction of spinal locomotor activity within one session of assisted treadmill walking in chronic complete SCI (Dietz and Muller, 2004). This EMG exhaustion could not be reverted by locomotor training. Therefore, a degradation of neuronal circuits in chronic SCI patients was assumed to take place. The aim of this study was to evaluate whether a similar behavior of reflex activity, i.e. an exhaustion of spinal reflexes occurs in chronic complete SCI patients within a session of assisted walking. A modification of the tibial nerve H-reflex would reflect a change in the monosynaptic pathway of the leg extensor muscles. The flexion reflex function is of special interest as it has been suggested to be a part of the locomotor pattern generator (Bussel et al., 1989;
1388-2457/$30.00 q 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2006.03.022
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
1500
Table 1 Characteristics of SCI subjects included in the study Subject
Age (y)
Sex
Level of lesion
ASIA
Duration of lesion (y)
Anti-spastic medication
SCI-1 SCI-2 SCI-3 SCI-4 SCI-5 SCI-6 SCI-7 SCI-8 SCI-9 SCI-10
33 45 56 40 33 35 29 53 39 34
M M M M M M F M M M
T6 C4 T4 T6 T6 T5 T5 C7 T5 T5
A B B A A A A B A A
1 2 2 7 7 8 11 12 12 13
Yes Yes No No Yes Yes Yes Yes No No
Age, sex, neurological level of lesion (C, cervical; T, thoracic), classification according to ASIA (American Spinal Injury Association, A, sensorimotor complete; B, motor complete; sensory incomplete), duration of SCI at time of recording (in years) and use of anti-spastic medication.
Dietz, 2002). Therefore, we hypothesized that the flexion reflex would show a similar amplitude reduction as the locomotor activity, while the H-reflex remains constant.
neurological level of lesion of the subjects was between C4 and T6 (see Table 1). The duration of the lesion lasted from 1 to 13 years. For comparison, the same measurements were done in 5 healthy subjects. For assisted walking, the driven gait orthosis (DGO) ‘Lokomat’ (Hocoma AG, Volketswil) mounted on a treadmill was used. A detailed description of the device can be found elsewhere (Colombo et al., 2000). Briefly, the DGO provides drives for physiological hip and knee joint movements of each leg, whereas the dorsiflexion of the feet during swing phase is achieved by elastic straps. As shown previously, the locomotor pattern evoked in such a condition differs little from that induced with external assistance provided by physiotherapists (Colombo et al., 2000). Unloading is achieved by a parachute harness connected to counterweights. SCI and healthy subjects walked for 15 min within the orthosis with support of 70% of their body weight. Speed was kept constant for all subjects at 2.0 km/h (0.56 m/s) whereas stepping cadence had to be slightly adjusted according to the individual leg length of the different subjects. All subjects had some but not regular experience in walking within the DGO.
2. Methods 2.2. Stimulation procedures 2.1. Subjects and walking procedures With the approval of the local Ethics Committee and the informed consent of the volunteers, leg muscle electromyographic (EMG) activity was recorded and analyzed during assisted locomotion with and without reflex stimulation in 10 subjects with a motor complete spastic paraplegia or tetraplegia (ASIA A/B) (Maynard et al., 1997). The
To elicit the H-reflex, electric stimuli were applied over the proximal tibial nerve in the popliteal fossa of the right leg. Stimuli were delivered at 30 s intervals (every 12 steps) through an isolated stimulator with constant current output (AS100, ALEA Solutions GmbH, Zurich). The stimulus consisted of a single biphasic stimulus of 2 ms duration (see Fig. 1). Stimulation intensity was set to the intensity
Fig. 1. Experimental set-up. Locomotion on the treadmill within a driven gait orthosis (DGO). Indicated are the signals for H-reflex (above) and flexion reflex stimulation (below) as well as the corresponding EMG responses.
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
required to elicit a maximal H-reflex response according to a preceding recruitment curve assessment. Maximal response was chosen to see any decline in amplitude. The stimuli were delivered during the early stance phase of the corresponding leg, i.e. 100 ms after heel strike, to enhance the probability of successful activation of the H-reflex. The distal tibial nerve at the dorsal aspect of the left medial malleolus was stimulated to elicit a flexion reflex response (FRR). Stimuli were similarly delivered every 30 s, but time shifted by 15 s compared to the H-reflex stimulation. The stimulus consisted of a train of 8 single biphasic electrical stimuli (duration 2 ms, i.e. frequency of 200 Hz, total stimulus length 40 ms; see Fig. 1). Flexion reflex threshold (FRT) was defined by the first visible contraction of TA muscle in the EMG signal or by direct observation. This approach had to be taken because of a very high foot skin impedance in the whole SCI group which made the usual assessment of motor threshold via activation of abductor hallucis muscle impossible. Stimulation intensity was set to 3 times FRT in SCI patients (16– 32 mA; can be assumed to represent noxious stimuli); in healthy subjects intensity had to be limited to 1–2 times FRT (16–40 mA) due to pain complaints; the stimuli were delivered at pre-swing (100 ms after contralateral heel strike). Before and after DGO walking, the stimulus–response relationship was assessed for both reflexes in a relaxed upright position (mounted into the harness) but without bearing weight. Stimulation intensity was increased by steps of 2 mA up to a maximum of 40 mA.
1501
EMG signal within a time-frame of 30–60 ms after stimulation onset. If elicited during walking, M-wave amplitude was checked for constancy of stimulation strength (Pierrot-Deseilligny and Mazevet, 2000). Flexion reflex responses were analyzed in TA and BF muscles of the corresponding leg, as well as in the contralateral GM, by calculating the rms value 60–450 ms after stimulation onset. Such a large window was used because there were large latency differences between subjects, and sometimes also time shifts of the reflex responses within one trial of a subject. Subjects or muscles with no FRR were omitted from the analysis. Data were then normalized to the median of the first 5 reflex responses during DGO walking. For every time-point of 5, 10 and 15 min after the beginning of the walking trial, the median value of 5 consecutive reflex responses were calculated. The onset of each reflex response was defined by a threshold of 5 times the standard deviation of the background EMG amplitude. FRR components with latencies between 60 and 120 ms were defined as medium, and longer ones as late. 2.4. Statistics Reduction in reflex response amplitude as well as in locomotor activity rms value after 15 min of walking were statistically tested using the one-tailed Wilcoxon signed ranks test. Correlations between the latency of flexion reflex responses and the duration and the level of lesion, respectively, were statistically tested using the Spearman rank correlation test. Significance level was set at P!0.05.
2.3. Data acquisition and analysis Leg muscle EMG activity (rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA) and medial gastrocnemius (GM)) from both sides was recorded through bipolar Ag/AgCl surface electrodes with an inter-electrode distance of 25 mm. EMG recordings were amplified, filtered (bandpass 30–300 Hz) and sampled at 1000 Hz via a 12-bit A/D-converter. In addition, trigger signals provided by sensors in the Lokomat that identified the onset of the right and left stance phases were recorded. For stimulation signal generation, data recording and signal analysis, customdesigned programs, written using Soleasy Software (ALEA Solutions GmbH, Zurich), were used. Locomotor EMG activity was analyzed using the rms value per stride smoothed by a moving window average (window width: 25 strides, which is equivalent to about 1 min of walking; strides with stimulation were excluded from this calculation) and then normalized to the rms value of the first stride (100% therefore represents the average value of the first 30 s of walking) (Dietz and Muller, 2004). Leg muscles without visible locomotor EMG activity were excluded from further analysis. H-reflex responses were analyzed in GM muscle by calculating the peak-to-peak amplitude of the corresponding
3. Results H-reflex responses (HRR) could be elicited during continuous walking over 15 min in all but one healthy subject. Flexion reflex responses (FRR) could be elicited in the TA muscle in all subjects. However, only 4 out of 10 SCI and 3 out of 5 healthy subjects showed a FRR in the BF muscle. In addition, FRR appeared with inter- and intraindividually variable latencies. Fig. 2 shows a typical example of a FRR. As in most chronic complete SCI patients TA EMG activity could hardly be detected during walking. In contrast, a distinct flexion reflex response appeared in the TA in this subject (Fig. 2A; latency around 200 ms). While there were some small changes in reflex amplitude and latency, the FRR remained generally constant throughout the 15 min of continuous walking in this subject. Fig. 2B shows a FRR in the BF muscle in the same SCI subject. In this case, the decrease in BF EMG activity was quite pronounced and occurred within 5 min of walking, with only single EMG potentials appearing towards the end of the recording. In contrast, a strong FRR was present throughout the whole 15 min walking period. In this
1502
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
Fig. 2. Flexion reflex responses (FRR) during walking. Ipsilateral (A) tibialis anterior (TA), (B) biceps femoris BF and (C) contralateral gastrocnemius (GM) EMG activity of a chronic complete SCI subject (SCI-6) during walking within the DGO. Course of EMG without stimulation (left side) and with flexion reflex stimulation (right side) at the beginning (0 min) and the end (15 min) of continuous walking over a 15 min period (2 km/h, 70% body unloading). Note. TA activity is almost absent in chronic SCI patients.
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
example, an early response (latency around 100 ms) was prominent at the beginning of the trial. It became smaller later, but was then accompanied by a pronounced late component (latency around 180 ms), which remained unchanged in amplitude until the end of the recording. In 5 out of 10 SCI subjects (but in none of the healthy subjects), EMG responses were also elicited in the contralateral GM muscle by ipsilateral FRR stimulation. Fig. 2C shows an example of contralateral GM EMG activity during the stance phase of locomotion with and without ipsilateral flexion reflex stimulation. With stimulation, 3–5 repetitive clonus-like bursts of GM activation appeared with an amplitude larger than the background EMG. These responses did not change in amplitude, in contrast to the condition without stimulation where a GM EMG decrement was seen. Quantitative analysis of the subjects with such responses showed no decrement of these contralateral GM responses over the 15 min recordings. Fig. 3 shows the quantitative analysis of the time course of the FRR and HRR in addition to that of the locomotor EMG activity over all subjects. The FRR amplitude in TA of both SCI and healthy subjects did not change significantly during the time of measurement (Fig. 3A, right panel). TA EMG activity during locomotion in the SCI subjects was usually not present. This missing activity seems not to be caused by the foot lifters. TA activity was also absent in manually assisted walking of the same patients without foot lifters and it was present in most of acute complete SCI subjects seen in other studies independent of type of assistance. The FRR in BF did not change significantly in amplitude in the SCI subjects (Fig. 3B, right panel). However, this figure includes data from only 4 SCI and 3 healthy subjects. In contrast, BF EMG activity of SCI subjects showed an almost significant decline within 15 min (PZ0.06) (Fig. 3B, left panel). Taking together all data of SCI subjects showing a BF activity (7 out of 10), this decline was highly significant as it occurred in all SCI subjects (P!0.01). Therefore, the decline of locomotor EMG activity was not associated with a corresponding change in flexion reflex amplitude. Fig. 3C shows the quantitative assessment of GM locomotor activity (left panel) and GM H-reflex behavior (right panel) in all healthy and SCI subjects with HRR and GM locomotor activity present during the 15 min of walking. There was a slight decrease in H-reflex amplitude in the SCI subjects, which was associated with a large variability. However, this change was not statistically different from that seen in healthy subjects. In contrast, GM locomotor activity of SCI subjects showed a significant decrement over the 15 min walking session (PZ0.02). Thus, the decrease observed in locomotor activity was not associated with a corresponding change in the H-reflex amplitude. One SCI patient showed no drop in GM locomotor activity (SCI-2). This subject received a regular locomotor training once a week during a 18 months period
1503
before the measurements. Excluding this patient from the HRR analysis did not change the overall results. The H-reflex latency in all subjects was in the range of 35–45 ms and did not differ between the walking and nonwalking conditions. Table 2 lists the latencies of the FRR in each subject. The FRR latencies varied largely within SCI subjects. Some corresponded to the values described in healthy subjects for medium latencies (80–100 ms), while in others only FRR with long latencies (200–250 ms) occurred. In some subjects, there were medium and long, or even an additional third longer-latency response. In the walking condition, medium latencies were seen more frequently than in the relaxed control condition. The appearance of medium, late or mixed FRR components did not correlate with any of the clinical parameters analyzed (i.e. duration, or level of SCI, antispastic medication) except for the TA FRR. Subjects with short duration of SCI (as well as healthy subjects) showed only medium latencies, whereas subjects with longer duration of lesion showed more mixed or only long latencies in the TA FRR (rZ0.77, PZ0.01). In addition, the course of medium and late FRR components was analyzed with regard to a possible differential behavior during the walking session. No difference was found for the TA FRR. For the BF a tendency (not significant) for a reduced medium and increased late response amplitude with walking time was found (see Fig. 4). Overall, the variability of the latencies was high and no firm conclusions could be drawn.
4. Discussion The aim of this study was to investigate the behavior of both locomotor activity and spinal reflexes in chronic complete SCI patients. Our main findings are as follows: H-reflex amplitude elicited in the GM and flexion reflex amplitude elicited in the TA and BF muscles were found to remain constant, despite an ‘exhaustion’ (in a unspecific sense) of the locomotor EMG activity in these muscles over the course of stepping. We showed earlier that this exhaustion is unlikely to be related to a habituation to the afferent information provided by the DGO (Dietz and Muller, 2004). In earlier studies, short-latency reflexes were suggested to be involved in the activation of the leg extensor muscles during locomotion in healthy subjects (for review see Dietz, 2003). For the generation of a locomotor activity in complete SCI subjects, load- and hip-related afferent input was shown to be of major significance, while stretch reflex activity was of minor significance (Dietz et al., 2002). Nevertheless, the removal of pre-synaptic inhibition of group Ia afferents, described for complete SCI patients early after trauma (Faist et al., 1994), might be reverted late after SCI, as it was suggested to occur for some spinal reflex mechanisms in cats (Tillakaratne et al., 2002) and humans
1504
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
Fig. 3. Time-course of reflex responses. Flexion reflex response (FRR) values (in % of initial value) of (A) tibialis anterior (TA) and (B) biceps femoris (BF), and (C) H-reflex values of gastrocnemius (GM) after 5, 10 and 15 min obtained in 5 healthy and 10 chronic SCI subjects during continuous walking over 15 min within the DGO (right panel). For comparison, the course of EMG activity during non-stimulated locomotion is shown (left panel). Note: In none of the SCI subjects was a distinct TA EMG recorded during non-stimulated locomotion. In (B) and (C) only subjects showing both locomotor and reflex activity are displayed. Data show median (thick horizontal line), first and third quartile as well as minimum and maximum values (white bars, healthy subjects; grey bars, SCI subjects); outliers are indicated by open circles. Statistically significant differences between values at 0 and 15 min are indicated by an asterisk (P!0.05).
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
1505
Table 2 Latencies of the flexion reflex responses of SCI and healthy subjects Subject
TA walking (ms)
BF walking (ms)
TA control (ms)
BF control (ms)
SCI-1 SCI-2 SCI-3 SCI-4 SCI-5 SCI-6 SCI-7 SCI-8 SCI-9 SCI-10 H-1 H-2 H-3 H-4 H-5
85–105/– 65–75/– 100–120/180–200 80–85/270–280 80–90/280–300 –/175–275 75–115/300–345 –/180–220 105–120/180–200 –/210–250 80–90/– 55–80/– 75–90/– 80–90/– 70–80/–
– 65–75/– –/190–240 – – – 95–120/250–280 – 70–85/140–175 – 70–85/– 65–80/– – – 60–70/115–180
130 160 185 200 215 180 120 260 160 205 80 80 80 85 75
– – 160 – – – 130 – 180 – 70 70 – – –
Latencies (in ms) of the medium/late components of the flexion reflex responses (FRR) during walking (range over 30 repetitions) and during the control condition (identical stimulation intensity as during walking) for individual subjects, recorded in the tibialis anterior (TA) and the biceps femoris (BF) muscles. SCI-x, SCI subject; H-x, healthy subject.
(Shefner et al., 1992). The present recordings show that in chronic SCI subjects, the H-reflex in the GM muscles remained constant in amplitude despite an exhaustion of leg extensor locomotor EMG activity. This discrepancy has to be seen in the light that a decrease in maximum HRR amplitude with time can be expected (Crone et al., 1999). This makes an involvement of the stretch reflex—we are aware that the H-reflex is not identical but part of this reflex (Pierrot-Deseilligny and Mazevet, 2000)—in the exhaustion rather unlikely. The flexion reflex or withdrawal reflex is of special interest in the context of the present study as it is assumed to be inter-connected with, and/or eventually to be part of, the central pattern generator underlying locomotion (Bussel
et al., 1989; Sherrington, 1910). Accordingly, locomotor movements could be induced in complete SCI patients by bilateral reciprocal activation of the flexion reflexes (Malezic and Hesse, 1995; Nicol et al., 1998). The fact that TA activity was absent in the SCI subjects during locomotion might be due to a reduced common synaptic drive to TA motoneurons (Hansen et al., 2005). This confounds, in part, the impact of TA activation. Flexion reflex recordings over a time period of 6 months after complete SCI showed that this reflex declines in amplitude with time (Hiersemenzel et al., 2000). Nevertheless, this study showed that if a flexion reflex response was present (medium or long latency or both components, most frequently in the TA) the amplitude remained constant
Fig. 4. Course of medium and late flexion reflex response amplitude. Medium and late components of FRR in TA (left panel) and BF (right panel) are shown over the 15 min walking time. Median (thick horizontal line), first and third quartile as well as minimum and maximum values are displayed (light grey bars, medium component; dark grey bars, late components of FRR); outliers are indicated by open circles.
1506
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
throughout a recording session. This was even the case if the TA was silent during locomotion in SCI subjects. The low number of SCI subjects with both locomotor and reflex activity in BF prevents to draw definitive conclusions for this muscle. In SCI subjects higher stimulus intensities could be applied than in the control subjects to ascertain an activation of flexion reflex afferents (Hiersemenzel et al., 2000). A wind-up of the flexion reflex described earlier (Hornby et al., 2003), as well as a habituation of the reflex (Dimitrijevic and Nathan, 1970), can largely be excluded for the stimulus intervals used in the present study. The FFR amplitude might be preserved: (1) due to a dynamic activation of motoneurons not involved in the locomotor activity by synchronized afferent volley released by the electrical stimulus; (2) because the flexion reflex stimulus activates afferents (i.e. noxious) which are not activated during stepping; therefore, only part of the flexion reflex might be related to locomotion (Lundberg, 1979); (3) because this reflex system is still regularly activated, e.g. during muscle spasms that frequently occur in chronic complete SCI; we would favor this last assumption. The great variability of flexion reflex responses in SCI subjects is consistent with earlier studies in this patient group (Andersen et al., 2004; Dimitrijevic and Nathan, 1970). We found that with increasing chronicity of SCI, there was a shift in the TA FRR from medium to long latencies. This might indicate an impairment of fast interneuronal connections with time after lesion. However, further longitudinal examinations after SCI are required to obtain more robust evidence about this aspect. It cannot be decided from this study whether the difference in FRR latency within one trial reflects a switch to another neuronal circuit of flexion reflex afferents. Cat preparations showed that the ipsilateral flexion movement following a noxious stimulus is associated with extension in the contralateral leg (Hagbarth, 1952; Sherrington, 1910). Consistent with previous studies, we did not observe such a contralateral activation of the extensor muscles in the healthy subjects (Hagbarth, 1960). In complete SCI patients a facilitation of contralateral leg extensor muscle activation to ipsilateral stimulation could only indirectly be demonstrated by the H-reflex technique in a non-moving condition (Bussel et al., 1989). In half of our SCI subjects, leg extensor activity was significantly enhanced during the contralateral stance phase following ipsilateral tibial nerve stimulation, but not in the relaxed control condition (hanging without weight bearing). This stimulus-associated EMG response did not decline in amplitude, as was the case for the background GM activity during the non-stimulated steps. The facilitatory effect of the flexion reflex on the contralateral GM EMG activity might be due to the activation of load-sensitive muscle receptors during the stance phase. Load receptors have been suggested to have an excitatory effect on flexion reflex pathways (Schmit et al., 2000).
In conclusion, this study shows that locomotor activity and the two spinal reflex systems tested show a divergent behavior in chronic complete SCI patients. The exhaustion of leg muscle EMG during assisted locomotion was not associated with a decline in H- or flexion reflex amplitude. This is surprising with respect to both reflex systems. The short-latency stretch reflex activity is thought to be involved in the activation of the main anti-gravity muscles during locomotion (Dietz, 2003). Also the flexion reflex system is suggested to share the same spinal inter-neuronal circuits of the central locomotor pattern generator (Bussel et al., 1989). However, one may argue that both the flexion and stretch reflexes are still effective in chronic SCI, since flexion spasms and muscle stretches frequently occur after SCI. We thus would speculate that the degradation/impairment of spinal neuronal function after SCI is strictly restricted to spinal neuronal circuits subserving locomotor function or to the drive of these circuits.
Acknowledgements The authors thank David Czell and Monica Stu¨ssi for help with data collection and analysis. We are grateful to Rachel Jurd, Tania Lam and Bjo¨rn Zo¨rner for editing the manuscript. This work was supported by the International Institute for Research in Paraplegia (IFP) and the National Center of Competence in Research on Neural Plasticity and Repair (NCCR Neuro). Disclosure. The authors have reported no conflicts of interest.
References Andersen OK, Finnerup NB, Spaich EG, Jensen TS, Arendt-Nielsen L. Expansion of nociceptive withdrawal reflex receptive fields in spinal cord injured humans. Clin Neurophysiol 2004;115:2798–810. Bussel B, Roby-Brami A, Yakovleff A, Bennis N. Late flexion reflex in paraplegic patients. Evidence for a spinal stepping generator. Brain Res Bull 1989;22:53–6. Colombo G, Joerg M, Schreier R, Dietz V. Treadmill training of paraplegic patients using a robotic orthosis. J Rehabil Res Dev 2000;37:693–700. Crone C, Johnsen LL, Hultborn H, Orsnes GB. Amplitude of the maximum motor response (Mmax) in human muscles typically decreases during the course of an experiment. Exp Brain Res 1999;124:265–70. de Leon RD, Hodgson JA, Roy RR, Edgerton VR. Locomotor capacity attributable to step training versus spontaneous recovery after spinalization in adult cats. J Neurophysiol 1998;79:1329–40. Dietz V. Proprioception and locomotor disorders. Nat Rev Neurosci 2002; 3:781–90. Dietz V. Spinal cord pattern generators for locomotion. Clin Neurophysiol 2003;114:1379–89. Dietz V, Muller R. Degradation of neuronal function following a spinal cord injury: mechanisms and countermeasures. Brain 2004;127: 2221–31. Dietz V, Muller R, Colombo G. Locomotor activity in spinal man: significance of afferent input from joint and load receptors. Brain 2002; 125:2626–34.
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507 Dimitrijevic MR, Nathan PW. Studies of spasticity in man 3. Analysis of reflex activity evoked by noxious cutaneous stimulation. Brain 1968;91: 349–68. Dimitrijevic MR, Nathan PW. Studies of spasticity in man 4. Changes in flexion reflex with repetitive cutaneous stimulation in spinal man. Brain 1970;93:743–68. 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;117:1449–55. Hagbarth KE. Excitatory and inhibitory skin areas for flexor and extensor motoneurons. Acta Physiol Scand Suppl 1952;26:1–58. Hagbarth KE. Spinal withdrawal reflexes in the human lower limbs. J Neurol Neurosurg Psychiatry 1960;23:222–7. Hansen NL, Conway BA, Halliday DM, Hansen S, Pyndt HS, BieringSorensen F, Nielsen JB. Reduction of common synaptic drive to ankle dorsiflexor motoneurons during walking in patients with spinal cord lesion. J Neurophysiol 2005;94:934–42. Hiersemenzel LP, Curt A, Dietz V. From spinal shock to spasticity: neuronal adaptations to a spinal cord injury. Neurology 2000;54: 1574–82. Hornby TG, Rymer WZ, Benz EN, Schmit BD. Windup of flexion reflexes in chronic human spinal cord injury: a marker for neuronal plateau potentials? J Neurophysiol 2003;89:416–26. Lundberg A. Multisensory control of spinal reflex pathways. In: Granit R, Pompeiano D, editors. Progress in brain research. Reflex control of posture and movement, vol. 50. Amsterdam: Elsevier; 1979. p. 12–28. Malezic M, Hesse S. Restoration of gait by functional electrical stimulation in paraplegic patients: a modified programme of treatment. Paraplegia 1995;33:126–31.
1507
Maynard Jr FM, Bracken MB, Creasey G, Ditunno Jr JF, Donovan WH, Ducker TB, Garber SL, Marino RJ, Stover SL, Tator CH, Waters RL, Wilberger JE, Young W. International Standards for Neurological and Functional Classification of Spinal Cord Injury, American Spinal Injury Association. Spinal Cord 1997;35:266–74. Nicol DJ, Granat MH, Tuson SJ, Baxendale RH. Variability of the dishabituation of flexion reflexes for FES assisted gait in spinal injured man. Med Eng Phys 1998;20:182–7. Pierrot-Deseilligny E, Mazevet D. The monosynaptic reflex: a tool to investigate motor control in humans. Interest and limits. Neurophysiol Clin 2000;30:67–80. Raineteau O, Schwab ME. Plasticity of motor systems after incomplete spinal cord injury. Nat Rev Neurosci 2001;2:263–73. Raisman G. A promising therapeutic approach to spinal cord repair. J R Soc Med 2003;96:259–61. Schmit BD, McKenna-Cole A, Rymer WZ. Flexor reflexes in chronic spinal cord injury triggered by imposed ankle rotation. Muscle Nerve 2000;23:793–803. Schwab ME. Nogo and axon regeneration. Curr Opin Neurobiol 2004;14: 118–24. Shefner JM, Berman SA, Sarkarati M, Young RR. Recurrent inhibition is increased in patients with spinal cord injury. Neurology 1992;42: 2162–8. Sherrington C. Flexion reflex of the limb, crossed extension-reflex, and reflex stepping and standing. J Physiol 1910;40:28–121. Tillakaratne NJ, de Leon RD, Hoang TX, Roy RR, Edgerton VR, Tobin AJ. Use-dependent modulation of inhibitory capacity in the feline lumbar spinal cord. J Neurosci 2002;22:3130–43.
Neuronal function in chronic spinal cord injury: Divergence between locomotor and flexion- and H-reflex activity R. Mu¨ller *, V. Dietz Spinal Cord Injury Center, Balgrist University Hospital, Forchstrasse 340, 8008 Zurich, Switzerland Accepted 30 March 2006 Available online 11 May 2006
Abstract Objective: Knowledge about the long-term course of spinal neuronal function after spinal cord injury (SCI) is important if regeneration therapies become available in the future. The objective of this study was to examine the behavior of locomotor EMG activity and of spinal reflexes in patients with chronic motor-complete SCI. Methods: EMG activity from rectus femoris (RF), biceps femoris (BF), medial gastrocnemius (GM) and tibialis anterior (TA) of both sides was investigated during locomotor movements assisted by a robotic device in 10 chronic (O1 year after accident) complete SCI and 5 healthy subjects. H-reflexes (recorded from GM) were induced during the onset and flexion reflexes (recorded from BF and TA) at the end of the stance phase. Results: Only in the chronic SCI subjects an exhaustion of EMG activity—i.e. a decrease in amplitude—occurred within a few minutes in all leg muscles. The EMG exhaustion was not associated with a change in the H- or flexion reflex amplitude during a walking session. Conclusions: Exhaustion of neuronal function in chronic complete SCI might be restricted to unused motor tasks, i.e. locomotion. The fact that H- and flexion reflexes show a normal behavior might be due to the fact that they still become activated. Significance: Training/pharmalogical approaches may be required to maintain neuronal function of unused tasks as a basis for future successful regeneration therapies. q 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. Keywords: Locomotor activity; Reflex activity; Spinal cord injury
1. Introduction The isolated spinal cord contains neuronal circuits underlying the generation of basic complex movements such as locomotion and flexion reflexes in cats (de Leon et al., 1998) and humans (for review see Dietz, 2002). However, the human studies have largely been based on recordings taken early after a spinal cord injury (SCI) (Andersen et al., 2004; Dimitrijevic and Nathan, 1968, 1970), and little is known about the long-term adaptive changes of spinal neuronal circuits after a complete SCI. Knowledge about the course of long-term spinal neuronal function in humans is important if regeneration therapies become available in the future. This indeed seems to be * Corresponding author. Tel.: C41 44 386 37 30; fax: C41 44 386 37 31. E-mail address: [email protected] (R. Mu¨ller).
feasible in the coming years (Raineteau and Schwab, 2001; Raisman, 2003; Schwab, 2004). The benefits of any such therapies rely crucially on the remaining function of neuronal spinal circuits. There is evidence for a reduction of spinal locomotor activity within one session of assisted treadmill walking in chronic complete SCI (Dietz and Muller, 2004). This EMG exhaustion could not be reverted by locomotor training. Therefore, a degradation of neuronal circuits in chronic SCI patients was assumed to take place. The aim of this study was to evaluate whether a similar behavior of reflex activity, i.e. an exhaustion of spinal reflexes occurs in chronic complete SCI patients within a session of assisted walking. A modification of the tibial nerve H-reflex would reflect a change in the monosynaptic pathway of the leg extensor muscles. The flexion reflex function is of special interest as it has been suggested to be a part of the locomotor pattern generator (Bussel et al., 1989;
1388-2457/$30.00 q 2006 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2006.03.022
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
1500
Table 1 Characteristics of SCI subjects included in the study Subject
Age (y)
Sex
Level of lesion
ASIA
Duration of lesion (y)
Anti-spastic medication
SCI-1 SCI-2 SCI-3 SCI-4 SCI-5 SCI-6 SCI-7 SCI-8 SCI-9 SCI-10
33 45 56 40 33 35 29 53 39 34
M M M M M M F M M M
T6 C4 T4 T6 T6 T5 T5 C7 T5 T5
A B B A A A A B A A
1 2 2 7 7 8 11 12 12 13
Yes Yes No No Yes Yes Yes Yes No No
Age, sex, neurological level of lesion (C, cervical; T, thoracic), classification according to ASIA (American Spinal Injury Association, A, sensorimotor complete; B, motor complete; sensory incomplete), duration of SCI at time of recording (in years) and use of anti-spastic medication.
Dietz, 2002). Therefore, we hypothesized that the flexion reflex would show a similar amplitude reduction as the locomotor activity, while the H-reflex remains constant.
neurological level of lesion of the subjects was between C4 and T6 (see Table 1). The duration of the lesion lasted from 1 to 13 years. For comparison, the same measurements were done in 5 healthy subjects. For assisted walking, the driven gait orthosis (DGO) ‘Lokomat’ (Hocoma AG, Volketswil) mounted on a treadmill was used. A detailed description of the device can be found elsewhere (Colombo et al., 2000). Briefly, the DGO provides drives for physiological hip and knee joint movements of each leg, whereas the dorsiflexion of the feet during swing phase is achieved by elastic straps. As shown previously, the locomotor pattern evoked in such a condition differs little from that induced with external assistance provided by physiotherapists (Colombo et al., 2000). Unloading is achieved by a parachute harness connected to counterweights. SCI and healthy subjects walked for 15 min within the orthosis with support of 70% of their body weight. Speed was kept constant for all subjects at 2.0 km/h (0.56 m/s) whereas stepping cadence had to be slightly adjusted according to the individual leg length of the different subjects. All subjects had some but not regular experience in walking within the DGO.
2. Methods 2.2. Stimulation procedures 2.1. Subjects and walking procedures With the approval of the local Ethics Committee and the informed consent of the volunteers, leg muscle electromyographic (EMG) activity was recorded and analyzed during assisted locomotion with and without reflex stimulation in 10 subjects with a motor complete spastic paraplegia or tetraplegia (ASIA A/B) (Maynard et al., 1997). The
To elicit the H-reflex, electric stimuli were applied over the proximal tibial nerve in the popliteal fossa of the right leg. Stimuli were delivered at 30 s intervals (every 12 steps) through an isolated stimulator with constant current output (AS100, ALEA Solutions GmbH, Zurich). The stimulus consisted of a single biphasic stimulus of 2 ms duration (see Fig. 1). Stimulation intensity was set to the intensity
Fig. 1. Experimental set-up. Locomotion on the treadmill within a driven gait orthosis (DGO). Indicated are the signals for H-reflex (above) and flexion reflex stimulation (below) as well as the corresponding EMG responses.
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
required to elicit a maximal H-reflex response according to a preceding recruitment curve assessment. Maximal response was chosen to see any decline in amplitude. The stimuli were delivered during the early stance phase of the corresponding leg, i.e. 100 ms after heel strike, to enhance the probability of successful activation of the H-reflex. The distal tibial nerve at the dorsal aspect of the left medial malleolus was stimulated to elicit a flexion reflex response (FRR). Stimuli were similarly delivered every 30 s, but time shifted by 15 s compared to the H-reflex stimulation. The stimulus consisted of a train of 8 single biphasic electrical stimuli (duration 2 ms, i.e. frequency of 200 Hz, total stimulus length 40 ms; see Fig. 1). Flexion reflex threshold (FRT) was defined by the first visible contraction of TA muscle in the EMG signal or by direct observation. This approach had to be taken because of a very high foot skin impedance in the whole SCI group which made the usual assessment of motor threshold via activation of abductor hallucis muscle impossible. Stimulation intensity was set to 3 times FRT in SCI patients (16– 32 mA; can be assumed to represent noxious stimuli); in healthy subjects intensity had to be limited to 1–2 times FRT (16–40 mA) due to pain complaints; the stimuli were delivered at pre-swing (100 ms after contralateral heel strike). Before and after DGO walking, the stimulus–response relationship was assessed for both reflexes in a relaxed upright position (mounted into the harness) but without bearing weight. Stimulation intensity was increased by steps of 2 mA up to a maximum of 40 mA.
1501
EMG signal within a time-frame of 30–60 ms after stimulation onset. If elicited during walking, M-wave amplitude was checked for constancy of stimulation strength (Pierrot-Deseilligny and Mazevet, 2000). Flexion reflex responses were analyzed in TA and BF muscles of the corresponding leg, as well as in the contralateral GM, by calculating the rms value 60–450 ms after stimulation onset. Such a large window was used because there were large latency differences between subjects, and sometimes also time shifts of the reflex responses within one trial of a subject. Subjects or muscles with no FRR were omitted from the analysis. Data were then normalized to the median of the first 5 reflex responses during DGO walking. For every time-point of 5, 10 and 15 min after the beginning of the walking trial, the median value of 5 consecutive reflex responses were calculated. The onset of each reflex response was defined by a threshold of 5 times the standard deviation of the background EMG amplitude. FRR components with latencies between 60 and 120 ms were defined as medium, and longer ones as late. 2.4. Statistics Reduction in reflex response amplitude as well as in locomotor activity rms value after 15 min of walking were statistically tested using the one-tailed Wilcoxon signed ranks test. Correlations between the latency of flexion reflex responses and the duration and the level of lesion, respectively, were statistically tested using the Spearman rank correlation test. Significance level was set at P!0.05.
2.3. Data acquisition and analysis Leg muscle EMG activity (rectus femoris (RF), biceps femoris (BF), tibialis anterior (TA) and medial gastrocnemius (GM)) from both sides was recorded through bipolar Ag/AgCl surface electrodes with an inter-electrode distance of 25 mm. EMG recordings were amplified, filtered (bandpass 30–300 Hz) and sampled at 1000 Hz via a 12-bit A/D-converter. In addition, trigger signals provided by sensors in the Lokomat that identified the onset of the right and left stance phases were recorded. For stimulation signal generation, data recording and signal analysis, customdesigned programs, written using Soleasy Software (ALEA Solutions GmbH, Zurich), were used. Locomotor EMG activity was analyzed using the rms value per stride smoothed by a moving window average (window width: 25 strides, which is equivalent to about 1 min of walking; strides with stimulation were excluded from this calculation) and then normalized to the rms value of the first stride (100% therefore represents the average value of the first 30 s of walking) (Dietz and Muller, 2004). Leg muscles without visible locomotor EMG activity were excluded from further analysis. H-reflex responses were analyzed in GM muscle by calculating the peak-to-peak amplitude of the corresponding
3. Results H-reflex responses (HRR) could be elicited during continuous walking over 15 min in all but one healthy subject. Flexion reflex responses (FRR) could be elicited in the TA muscle in all subjects. However, only 4 out of 10 SCI and 3 out of 5 healthy subjects showed a FRR in the BF muscle. In addition, FRR appeared with inter- and intraindividually variable latencies. Fig. 2 shows a typical example of a FRR. As in most chronic complete SCI patients TA EMG activity could hardly be detected during walking. In contrast, a distinct flexion reflex response appeared in the TA in this subject (Fig. 2A; latency around 200 ms). While there were some small changes in reflex amplitude and latency, the FRR remained generally constant throughout the 15 min of continuous walking in this subject. Fig. 2B shows a FRR in the BF muscle in the same SCI subject. In this case, the decrease in BF EMG activity was quite pronounced and occurred within 5 min of walking, with only single EMG potentials appearing towards the end of the recording. In contrast, a strong FRR was present throughout the whole 15 min walking period. In this
1502
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
Fig. 2. Flexion reflex responses (FRR) during walking. Ipsilateral (A) tibialis anterior (TA), (B) biceps femoris BF and (C) contralateral gastrocnemius (GM) EMG activity of a chronic complete SCI subject (SCI-6) during walking within the DGO. Course of EMG without stimulation (left side) and with flexion reflex stimulation (right side) at the beginning (0 min) and the end (15 min) of continuous walking over a 15 min period (2 km/h, 70% body unloading). Note. TA activity is almost absent in chronic SCI patients.
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
example, an early response (latency around 100 ms) was prominent at the beginning of the trial. It became smaller later, but was then accompanied by a pronounced late component (latency around 180 ms), which remained unchanged in amplitude until the end of the recording. In 5 out of 10 SCI subjects (but in none of the healthy subjects), EMG responses were also elicited in the contralateral GM muscle by ipsilateral FRR stimulation. Fig. 2C shows an example of contralateral GM EMG activity during the stance phase of locomotion with and without ipsilateral flexion reflex stimulation. With stimulation, 3–5 repetitive clonus-like bursts of GM activation appeared with an amplitude larger than the background EMG. These responses did not change in amplitude, in contrast to the condition without stimulation where a GM EMG decrement was seen. Quantitative analysis of the subjects with such responses showed no decrement of these contralateral GM responses over the 15 min recordings. Fig. 3 shows the quantitative analysis of the time course of the FRR and HRR in addition to that of the locomotor EMG activity over all subjects. The FRR amplitude in TA of both SCI and healthy subjects did not change significantly during the time of measurement (Fig. 3A, right panel). TA EMG activity during locomotion in the SCI subjects was usually not present. This missing activity seems not to be caused by the foot lifters. TA activity was also absent in manually assisted walking of the same patients without foot lifters and it was present in most of acute complete SCI subjects seen in other studies independent of type of assistance. The FRR in BF did not change significantly in amplitude in the SCI subjects (Fig. 3B, right panel). However, this figure includes data from only 4 SCI and 3 healthy subjects. In contrast, BF EMG activity of SCI subjects showed an almost significant decline within 15 min (PZ0.06) (Fig. 3B, left panel). Taking together all data of SCI subjects showing a BF activity (7 out of 10), this decline was highly significant as it occurred in all SCI subjects (P!0.01). Therefore, the decline of locomotor EMG activity was not associated with a corresponding change in flexion reflex amplitude. Fig. 3C shows the quantitative assessment of GM locomotor activity (left panel) and GM H-reflex behavior (right panel) in all healthy and SCI subjects with HRR and GM locomotor activity present during the 15 min of walking. There was a slight decrease in H-reflex amplitude in the SCI subjects, which was associated with a large variability. However, this change was not statistically different from that seen in healthy subjects. In contrast, GM locomotor activity of SCI subjects showed a significant decrement over the 15 min walking session (PZ0.02). Thus, the decrease observed in locomotor activity was not associated with a corresponding change in the H-reflex amplitude. One SCI patient showed no drop in GM locomotor activity (SCI-2). This subject received a regular locomotor training once a week during a 18 months period
1503
before the measurements. Excluding this patient from the HRR analysis did not change the overall results. The H-reflex latency in all subjects was in the range of 35–45 ms and did not differ between the walking and nonwalking conditions. Table 2 lists the latencies of the FRR in each subject. The FRR latencies varied largely within SCI subjects. Some corresponded to the values described in healthy subjects for medium latencies (80–100 ms), while in others only FRR with long latencies (200–250 ms) occurred. In some subjects, there were medium and long, or even an additional third longer-latency response. In the walking condition, medium latencies were seen more frequently than in the relaxed control condition. The appearance of medium, late or mixed FRR components did not correlate with any of the clinical parameters analyzed (i.e. duration, or level of SCI, antispastic medication) except for the TA FRR. Subjects with short duration of SCI (as well as healthy subjects) showed only medium latencies, whereas subjects with longer duration of lesion showed more mixed or only long latencies in the TA FRR (rZ0.77, PZ0.01). In addition, the course of medium and late FRR components was analyzed with regard to a possible differential behavior during the walking session. No difference was found for the TA FRR. For the BF a tendency (not significant) for a reduced medium and increased late response amplitude with walking time was found (see Fig. 4). Overall, the variability of the latencies was high and no firm conclusions could be drawn.
4. Discussion The aim of this study was to investigate the behavior of both locomotor activity and spinal reflexes in chronic complete SCI patients. Our main findings are as follows: H-reflex amplitude elicited in the GM and flexion reflex amplitude elicited in the TA and BF muscles were found to remain constant, despite an ‘exhaustion’ (in a unspecific sense) of the locomotor EMG activity in these muscles over the course of stepping. We showed earlier that this exhaustion is unlikely to be related to a habituation to the afferent information provided by the DGO (Dietz and Muller, 2004). In earlier studies, short-latency reflexes were suggested to be involved in the activation of the leg extensor muscles during locomotion in healthy subjects (for review see Dietz, 2003). For the generation of a locomotor activity in complete SCI subjects, load- and hip-related afferent input was shown to be of major significance, while stretch reflex activity was of minor significance (Dietz et al., 2002). Nevertheless, the removal of pre-synaptic inhibition of group Ia afferents, described for complete SCI patients early after trauma (Faist et al., 1994), might be reverted late after SCI, as it was suggested to occur for some spinal reflex mechanisms in cats (Tillakaratne et al., 2002) and humans
1504
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
Fig. 3. Time-course of reflex responses. Flexion reflex response (FRR) values (in % of initial value) of (A) tibialis anterior (TA) and (B) biceps femoris (BF), and (C) H-reflex values of gastrocnemius (GM) after 5, 10 and 15 min obtained in 5 healthy and 10 chronic SCI subjects during continuous walking over 15 min within the DGO (right panel). For comparison, the course of EMG activity during non-stimulated locomotion is shown (left panel). Note: In none of the SCI subjects was a distinct TA EMG recorded during non-stimulated locomotion. In (B) and (C) only subjects showing both locomotor and reflex activity are displayed. Data show median (thick horizontal line), first and third quartile as well as minimum and maximum values (white bars, healthy subjects; grey bars, SCI subjects); outliers are indicated by open circles. Statistically significant differences between values at 0 and 15 min are indicated by an asterisk (P!0.05).
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
1505
Table 2 Latencies of the flexion reflex responses of SCI and healthy subjects Subject
TA walking (ms)
BF walking (ms)
TA control (ms)
BF control (ms)
SCI-1 SCI-2 SCI-3 SCI-4 SCI-5 SCI-6 SCI-7 SCI-8 SCI-9 SCI-10 H-1 H-2 H-3 H-4 H-5
85–105/– 65–75/– 100–120/180–200 80–85/270–280 80–90/280–300 –/175–275 75–115/300–345 –/180–220 105–120/180–200 –/210–250 80–90/– 55–80/– 75–90/– 80–90/– 70–80/–
– 65–75/– –/190–240 – – – 95–120/250–280 – 70–85/140–175 – 70–85/– 65–80/– – – 60–70/115–180
130 160 185 200 215 180 120 260 160 205 80 80 80 85 75
– – 160 – – – 130 – 180 – 70 70 – – –
Latencies (in ms) of the medium/late components of the flexion reflex responses (FRR) during walking (range over 30 repetitions) and during the control condition (identical stimulation intensity as during walking) for individual subjects, recorded in the tibialis anterior (TA) and the biceps femoris (BF) muscles. SCI-x, SCI subject; H-x, healthy subject.
(Shefner et al., 1992). The present recordings show that in chronic SCI subjects, the H-reflex in the GM muscles remained constant in amplitude despite an exhaustion of leg extensor locomotor EMG activity. This discrepancy has to be seen in the light that a decrease in maximum HRR amplitude with time can be expected (Crone et al., 1999). This makes an involvement of the stretch reflex—we are aware that the H-reflex is not identical but part of this reflex (Pierrot-Deseilligny and Mazevet, 2000)—in the exhaustion rather unlikely. The flexion reflex or withdrawal reflex is of special interest in the context of the present study as it is assumed to be inter-connected with, and/or eventually to be part of, the central pattern generator underlying locomotion (Bussel
et al., 1989; Sherrington, 1910). Accordingly, locomotor movements could be induced in complete SCI patients by bilateral reciprocal activation of the flexion reflexes (Malezic and Hesse, 1995; Nicol et al., 1998). The fact that TA activity was absent in the SCI subjects during locomotion might be due to a reduced common synaptic drive to TA motoneurons (Hansen et al., 2005). This confounds, in part, the impact of TA activation. Flexion reflex recordings over a time period of 6 months after complete SCI showed that this reflex declines in amplitude with time (Hiersemenzel et al., 2000). Nevertheless, this study showed that if a flexion reflex response was present (medium or long latency or both components, most frequently in the TA) the amplitude remained constant
Fig. 4. Course of medium and late flexion reflex response amplitude. Medium and late components of FRR in TA (left panel) and BF (right panel) are shown over the 15 min walking time. Median (thick horizontal line), first and third quartile as well as minimum and maximum values are displayed (light grey bars, medium component; dark grey bars, late components of FRR); outliers are indicated by open circles.
1506
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507
throughout a recording session. This was even the case if the TA was silent during locomotion in SCI subjects. The low number of SCI subjects with both locomotor and reflex activity in BF prevents to draw definitive conclusions for this muscle. In SCI subjects higher stimulus intensities could be applied than in the control subjects to ascertain an activation of flexion reflex afferents (Hiersemenzel et al., 2000). A wind-up of the flexion reflex described earlier (Hornby et al., 2003), as well as a habituation of the reflex (Dimitrijevic and Nathan, 1970), can largely be excluded for the stimulus intervals used in the present study. The FFR amplitude might be preserved: (1) due to a dynamic activation of motoneurons not involved in the locomotor activity by synchronized afferent volley released by the electrical stimulus; (2) because the flexion reflex stimulus activates afferents (i.e. noxious) which are not activated during stepping; therefore, only part of the flexion reflex might be related to locomotion (Lundberg, 1979); (3) because this reflex system is still regularly activated, e.g. during muscle spasms that frequently occur in chronic complete SCI; we would favor this last assumption. The great variability of flexion reflex responses in SCI subjects is consistent with earlier studies in this patient group (Andersen et al., 2004; Dimitrijevic and Nathan, 1970). We found that with increasing chronicity of SCI, there was a shift in the TA FRR from medium to long latencies. This might indicate an impairment of fast interneuronal connections with time after lesion. However, further longitudinal examinations after SCI are required to obtain more robust evidence about this aspect. It cannot be decided from this study whether the difference in FRR latency within one trial reflects a switch to another neuronal circuit of flexion reflex afferents. Cat preparations showed that the ipsilateral flexion movement following a noxious stimulus is associated with extension in the contralateral leg (Hagbarth, 1952; Sherrington, 1910). Consistent with previous studies, we did not observe such a contralateral activation of the extensor muscles in the healthy subjects (Hagbarth, 1960). In complete SCI patients a facilitation of contralateral leg extensor muscle activation to ipsilateral stimulation could only indirectly be demonstrated by the H-reflex technique in a non-moving condition (Bussel et al., 1989). In half of our SCI subjects, leg extensor activity was significantly enhanced during the contralateral stance phase following ipsilateral tibial nerve stimulation, but not in the relaxed control condition (hanging without weight bearing). This stimulus-associated EMG response did not decline in amplitude, as was the case for the background GM activity during the non-stimulated steps. The facilitatory effect of the flexion reflex on the contralateral GM EMG activity might be due to the activation of load-sensitive muscle receptors during the stance phase. Load receptors have been suggested to have an excitatory effect on flexion reflex pathways (Schmit et al., 2000).
In conclusion, this study shows that locomotor activity and the two spinal reflex systems tested show a divergent behavior in chronic complete SCI patients. The exhaustion of leg muscle EMG during assisted locomotion was not associated with a decline in H- or flexion reflex amplitude. This is surprising with respect to both reflex systems. The short-latency stretch reflex activity is thought to be involved in the activation of the main anti-gravity muscles during locomotion (Dietz, 2003). Also the flexion reflex system is suggested to share the same spinal inter-neuronal circuits of the central locomotor pattern generator (Bussel et al., 1989). However, one may argue that both the flexion and stretch reflexes are still effective in chronic SCI, since flexion spasms and muscle stretches frequently occur after SCI. We thus would speculate that the degradation/impairment of spinal neuronal function after SCI is strictly restricted to spinal neuronal circuits subserving locomotor function or to the drive of these circuits.
Acknowledgements The authors thank David Czell and Monica Stu¨ssi for help with data collection and analysis. We are grateful to Rachel Jurd, Tania Lam and Bjo¨rn Zo¨rner for editing the manuscript. This work was supported by the International Institute for Research in Paraplegia (IFP) and the National Center of Competence in Research on Neural Plasticity and Repair (NCCR Neuro). Disclosure. The authors have reported no conflicts of interest.
References Andersen OK, Finnerup NB, Spaich EG, Jensen TS, Arendt-Nielsen L. Expansion of nociceptive withdrawal reflex receptive fields in spinal cord injured humans. Clin Neurophysiol 2004;115:2798–810. Bussel B, Roby-Brami A, Yakovleff A, Bennis N. Late flexion reflex in paraplegic patients. Evidence for a spinal stepping generator. Brain Res Bull 1989;22:53–6. Colombo G, Joerg M, Schreier R, Dietz V. Treadmill training of paraplegic patients using a robotic orthosis. J Rehabil Res Dev 2000;37:693–700. Crone C, Johnsen LL, Hultborn H, Orsnes GB. Amplitude of the maximum motor response (Mmax) in human muscles typically decreases during the course of an experiment. Exp Brain Res 1999;124:265–70. de Leon RD, Hodgson JA, Roy RR, Edgerton VR. Locomotor capacity attributable to step training versus spontaneous recovery after spinalization in adult cats. J Neurophysiol 1998;79:1329–40. Dietz V. Proprioception and locomotor disorders. Nat Rev Neurosci 2002; 3:781–90. Dietz V. Spinal cord pattern generators for locomotion. Clin Neurophysiol 2003;114:1379–89. Dietz V, Muller R. Degradation of neuronal function following a spinal cord injury: mechanisms and countermeasures. Brain 2004;127: 2221–31. Dietz V, Muller R, Colombo G. Locomotor activity in spinal man: significance of afferent input from joint and load receptors. Brain 2002; 125:2626–34.
R. Mu¨ller, V. Dietz / Clinical Neurophysiology 117 (2006) 1499–1507 Dimitrijevic MR, Nathan PW. Studies of spasticity in man 3. Analysis of reflex activity evoked by noxious cutaneous stimulation. Brain 1968;91: 349–68. Dimitrijevic MR, Nathan PW. Studies of spasticity in man 4. Changes in flexion reflex with repetitive cutaneous stimulation in spinal man. Brain 1970;93:743–68. 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;117:1449–55. Hagbarth KE. Excitatory and inhibitory skin areas for flexor and extensor motoneurons. Acta Physiol Scand Suppl 1952;26:1–58. Hagbarth KE. Spinal withdrawal reflexes in the human lower limbs. J Neurol Neurosurg Psychiatry 1960;23:222–7. Hansen NL, Conway BA, Halliday DM, Hansen S, Pyndt HS, BieringSorensen F, Nielsen JB. Reduction of common synaptic drive to ankle dorsiflexor motoneurons during walking in patients with spinal cord lesion. J Neurophysiol 2005;94:934–42. Hiersemenzel LP, Curt A, Dietz V. From spinal shock to spasticity: neuronal adaptations to a spinal cord injury. Neurology 2000;54: 1574–82. Hornby TG, Rymer WZ, Benz EN, Schmit BD. Windup of flexion reflexes in chronic human spinal cord injury: a marker for neuronal plateau potentials? J Neurophysiol 2003;89:416–26. Lundberg A. Multisensory control of spinal reflex pathways. In: Granit R, Pompeiano D, editors. Progress in brain research. Reflex control of posture and movement, vol. 50. Amsterdam: Elsevier; 1979. p. 12–28. Malezic M, Hesse S. Restoration of gait by functional electrical stimulation in paraplegic patients: a modified programme of treatment. Paraplegia 1995;33:126–31.
1507
Maynard Jr FM, Bracken MB, Creasey G, Ditunno Jr JF, Donovan WH, Ducker TB, Garber SL, Marino RJ, Stover SL, Tator CH, Waters RL, Wilberger JE, Young W. International Standards for Neurological and Functional Classification of Spinal Cord Injury, American Spinal Injury Association. Spinal Cord 1997;35:266–74. Nicol DJ, Granat MH, Tuson SJ, Baxendale RH. Variability of the dishabituation of flexion reflexes for FES assisted gait in spinal injured man. Med Eng Phys 1998;20:182–7. Pierrot-Deseilligny E, Mazevet D. The monosynaptic reflex: a tool to investigate motor control in humans. Interest and limits. Neurophysiol Clin 2000;30:67–80. Raineteau O, Schwab ME. Plasticity of motor systems after incomplete spinal cord injury. Nat Rev Neurosci 2001;2:263–73. Raisman G. A promising therapeutic approach to spinal cord repair. J R Soc Med 2003;96:259–61. Schmit BD, McKenna-Cole A, Rymer WZ. Flexor reflexes in chronic spinal cord injury triggered by imposed ankle rotation. Muscle Nerve 2000;23:793–803. Schwab ME. Nogo and axon regeneration. Curr Opin Neurobiol 2004;14: 118–24. Shefner JM, Berman SA, Sarkarati M, Young RR. Recurrent inhibition is increased in patients with spinal cord injury. Neurology 1992;42: 2162–8. Sherrington C. Flexion reflex of the limb, crossed extension-reflex, and reflex stepping and standing. J Physiol 1910;40:28–121. Tillakaratne NJ, de Leon RD, Hoang TX, Roy RR, Edgerton VR, Tobin AJ. Use-dependent modulation of inhibitory capacity in the feline lumbar spinal cord. J Neurosci 2002;22:3130–43.