Brain, Vol. 124, No. 8, 1590-1600,
August 2001
© 2001 Oxford University Press
Effect of bilateral subthalamic nucleus stimulation on gait in Parkinson's disease
1 Department of Clinical Neurology and Neurophysiology, University of Freiburg, Germany, 2 Department of Clinical and Biological Neurosciences, University Hospital of Grenoble and 3 INSERM Unit 318, Joseph Fourier University, Grenoble, France
Correspondence to:
Dr Michael Faist, Department of Clinical Neurology and Neurophysiology, University of Freiburg, Breisacherstr 64, D-79106 Freiburg, Germany E-mail: faist{at}nz11.ukl.uni-freiburg.de
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The fundamental disturbance of the parkinsonian gait is the reduction in walking velocity. This is mainly due to reduction in stride length, while cadence (steps/min) is slightly enhanced. Treatment with L-dopa increases stride length while cadence is unchanged. Chronic stimulation of the thalamus has no effect on Parkinsonian gait. The efficacy of electrical stimulation of the subthalamic nucleus (STN) on gait in advanced Parkinson's disease has been clearly demonstrated clinically. The aim of the present study was to quantify the changes in gait measures induced by STN stimulation and L-dopa and to assess possible differential or additive effects. Eight Parkinson's disease patients (mean ± SD age 48.1 ± 7.3 years) with chronic bilateral STN stimulation (mean duration of disease 13.3 ± 2.4 years, mean stimulation time 15.4 ± 10.6 months) and 12 age-matched controls were investigated. Subjects walked on a special treadmill with a closed-loop ultrasound control system that used the subject's position to adjust treadmill speed continuously for the actual walking velocity. In an appropriate crossover design, spatiotemporal gait measures and leg joint angle movements were assessed for at least 120 stride cycles in four treatment conditions: with and without stimulation and with and without a suprathreshold dose of L-dopa. With STN stimulation, there were increases of almost threefold in mean walking velocity (from 0.35 to 0.96 m/s) and stride length (from 0.34 to 0.99 m). Cadence remained constant. The range of motion of the major leg joints also increased. L-Dopa alone had a slightly weaker effect, with an increase in walking velocity to 0.94 m/s and in stride length to 0.92 m at a similar cadence. These increased values were in the range of those for healthy age-matched subjects performing the same task. The combination of both treatments further increased the mean walking velocity to 1.19 m/s and stride length to 1.20 m at an unchanged cadence. However, not all patients receiving STN stimulation improved further when they also received L-dopa. These results demonstrate that chronic bilateral STN stimulation, like treatment with L-dopa, improves walking velocity by increasing stride length without changing cadence. STN stimulation almost exclusively affects mechanisms involved in the control of spatial gait measures rather than rhythmicity. The gait measures obtained with STN stimulation alone are in the range of control subjects.
gait disorders; movement disorders; Parkinson's disease; gait analysis; STN stimulation
GPi = internal globus pallidus; STN = subthalamic nucleus; UPDRS = Unified Parkinson's Disease Rating Scale
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
A typical symptom of Parkinson's disease is a short-stepped, shuffling gait with slow walking velocity and decreased amplitude of the segmental movements. The problem in this gait disturbance is the regulation of stride length (Morris et al., 1994a
, b). Healthy subjects achieve an increase in walking velocity by increasing both stride length and cadence to a similar extent. Cadence regulation is not disturbed in Parkinson's disease. The slope of the stride length–cadence relationship is similar in healthy subjects and Parkinson's disease patients, whereas the intercept for stride length is reduced (Morris et al., 1998). This means that Parkinson's disease patients are able to increase their stride length by amounts similar to those achieved by healthy subjects for any given increase in cadence, but that the stride length is preset at a lower level for the whole stride length–cadence relationship in Parkinson's disease. Furthermore, the usable range of this relationship is reduced, which means that the `break point', i.e. the point at which stride length cannot be increased any more, occurs at a lower cadence. Treatment with L-dopa increases stride length and kinematic measures (swing velocity, peak velocity), whereas temporal measures, such as cadence, swing duration, double limb support time (both feet in contact with the ground) and, in one study, also cadence variability, have been found to be resistant to L-dopa (Blin et al., 1991; O'Sullivan et al., 1998). Treatment with L-dopa increases the usable range of the stride length–cadence relationship and also the intercept for stride length (Morris et al., 1998).
Chronic thalamic stimulation has no effect on parkinsonian gait (Defebvre et al., 1996). Bilateral posteroventral pallidotomy results in a twofold increase in stride length without a change in cadence (Siegel and Metman, 2000). The efficacy of electrical stimulation of the subthalamic nucleus (STN) in advanced Parkinson's disease has clearly been demonstrated clinically by the use of qualitative scales (Limousin et al., 1998; Yokoyama et al., 1999). However, quantitative assessment of gait in Parkinson's disease with chronic STN stimulation has not been published. Preliminary reports suggest a tremendous effect of STN stimulation on walking velocity (Allert et al., 2000; Faist et al., 2000; Krystkowiak et al., 2000; Melnick et al., 2000). However, results concerning the strategies by which this effect is achieved are divergent. Two of these studies suggest that the improvement is due exclusively to an increase in stride length (Faist et al., 2000; Melnick et al., 2000), similar to what was found for L-dopa (Blin et al., 1991; Morris et al., 1998; O'Sullivan et al., 1998). The other two studies report also an increase in cadence (Allert et al., 2000; Krystkowiak et al., 2000). The questions arise whether changes in measures of gait induced by STN stimulation are different from those induced by pallidotomy or L-dopa and whether the effects of L-dopa and STN stimulation are differential or additive. Deep-brain stimulation offers the exceptional possibility of studying the same patient with and without `symptoms' within the same experimental session. The aim of the present study was to quantify, in an appropriate crossover design, changes in the gait pattern of Parkinson's disease patients receiving chronic bilateral STN stimulation in four treatment conditions: with and without STN stimulation and with and without a suprathreshold dose of L-dopa. Also, a group of healthy age-matched control subjects was investigated. Some of these results have been published in abstract form (Faist et al., 2000).
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Eight Parkinson's disease patients (three women and five men) with a mean age of 48.1 years were investigated. These patients represent a rather young group, with disease onset at 34.8 years compared with the average onset of Parkinson's disease at 55 years. All had been operated at the Grenoble University Hospital and received chronic bilateral STN stimulation according to criteria reported by Limousin and colleagues (Limousin et al., 1998
). The position of the electrode was checked by intraoperative teleradiography and postoperative MRI. All the implanted electrodes were in the STN area. The stimulation was monopolar and used only one contact for each of the 16 implanted electrodes. The rate (mean ± SD) was 148 ± 24 Hz for the left side and 139 ± 19 Hz for the right, and the voltage 3.0 ± 0.4 V for the left side and 2.8 ± 0.4 V for the right. The pulse width was 60 µs for 15 electrodes and 90 µs for one. All patients had a clear response to L-dopa. We did not select the patients according to the degree of benefit induced by STN stimulation. Consecutive patients who were able to walk off-stimulation/off-medication and who agreed to travel to Freiburg entered the study. Detailed clinical data are given in Table 1. Five of the patients still required L-dopa after implantation (Table 1). Additional medication included amantadine in one patient, pergolide in two patients and bromocriptine in three patients. Additionally, 12 age-matched control subjects, with a mean age of 51.1 ± 9.1 years, were investigated. The study was approved by the ethics committee of the University of Freiburg according to the declaration of Helsinki, and the patients and healthy control subjects gave their written informed consent.
|
General experimental procedure
The gait analysis and clinical assessment were performed in the gait laboratory of the Department of Neurology and Clinical Neurophysiology of Freiburg University Hospital in four conditions: (i) off-stimulation/off-medication (–stim/ –dopa); (ii) on-stimulation/off-medication (+stim/–dopa); (iii) off-stimulation/on-medication (–stim/+dopa); and (iv) on-stimulation/on-medication (+stim/+dopa). One experimental session took ~1 h for each of the four conditions, including attachment of electrodes and goniometers and interruptions for rest. Two conditions were performed per day (early morning off-medication, late morning on-medication); 2 days were required per patient for the gait experiments (1 day off- and 1 day on-stimulation). The order of the off- and on-stimulation conditions was randomized; thus, four patients started in the off-stimulation condition and four patients started in the on-stimulation condition. The stimulation condition was set at least 1 h before the experiment by a person independent of the experimenters, who were blinded to the stimulation condition and to the patients. Despite this blinded condition, all patients identified the stimulation condition within a few minutes because of the magnitude of the effect. Before they were assessed off-medication, patients fasted and drugs were withdrawn overnight. The assessment on-medication was performed when the patient was fully `on', ~40 min after the administration of a suprathreshold dose of 200–300 mg liquid L-dopa and 50–75 mg benserazide (dispersible Madopar) and a light breakfast to minimize variability in the absorption of L-dopa and to speed up the absorption of L-dopa. Thus, the experimental procedures on a single day took ~4 h, including ample time for rest. Clinical assessment was performed with the Unified Parkinson's Disease Rating Scale (UPDRS) motor score and the Tinetti gait and balance score (Tinetti, 1986
). Healthy control subjects performed the experimental procedure only once.
Gait analysis: recording methods
For the quantitative analysis of gait, subjects walked on a self-driven treadmill with separate belts for the right and left feet. Each of the two belts was placed separately over a force-measuring system (Kistler, Winterthur, Switzerland). Thus, we were able to determine the force exerted on each belt and to assess heel contact and toe off in the step cycle from the output of four piezo elements fixed at the corners of each belt. The left leg can be placed on the right belt and vice versa, especially during fast walking, so that not all strides could be determined correctly by the force-measuring system. To determine spatiotemporal measurements for these strides, subjects wore on both feet a specially designed shoe that contained circuit breakers to determine heel strike and toe off. We recorded continuously the treadmill velocity and the position of the subject in an anterior or posterior direction on the belt. Angular displacements of the hip, knee and ankle joint of the right leg were measured using laterally placed goniometers. To adjust the treadmill velocity to the subject's chosen walking velocity, a closed-loop control system was used in which measurement of the subject's position on the treadmill gave the information required. The position was determined by a distance measurement (Headtracker ultrasonic system; Logitech, Fremont, Calif., USA) between a fixed emitter placed behind the treadmill and a receiver placed on the subject's back. When the subject tended to speed up or slow down, the measured distance increased or decreased and the computer control system accelerated or decelerated the treadmill in order to return the subject to the calibrated position. The speed of the treadmill was recorded continuously. This arrangement allowed the continuous monitoring of walking velocity for an unlimited walking distance. Freely chosen walking velocity was adjusted automatically to the intended walking velocity. Subjects were instructed to keep their walking velocity constant and at the maximum comfortable speed. For safety reasons, all subjects were secured with a safety belt without weight support; the belt was suspended from the ceiling to prevent injuries in case of a fall. After 5 min of training with stimulation on, all patients were able to walk freely and comfortably with this system. For gait assessment, at least 3 min of walking in a steady state for each of the four conditions was recorded in the crossover design described above. Two patients were unable to walk the required 120 stride cycles in the off-medication and off-stimulation conditions; therefore, only some of the biomechanical measures could be assessed in this single condition. One of these patients had started off-stimulation and the other had started on-stimulation, so the crossover design was not spoiled.
Data analysis
Spatiotemporal and velocity recordings were amplified (floating input amplifier, bandwidth 3–200 Hz, sensitivity 2 mV/V) and transferred on-line to a computer system (IBM-compatible Pentium) via an analogue–digital converter sampling at 500 Hz. The data were analysed off-line with software developed by our gait laboratory and programmed in LabView, National Instruments, Munich, Germany. The onset of the force signal was used as a trigger for selecting successive stride cycles. Stride duration was defined as the duration of the period between two successive impacts of the right leg. Walking velocity, stride and step length, cadence, swing time, swing velocity, stance time, double-limb support time and the range of motion for the hip, knee and ankle joint angles were assessed for at least 120 stride cycles. Further analysis was performed using Microsoft Excel and StatView; SAS Inc., Cary, NC, USA. Swing time, stance time and double-limb support time were expressed as a percentage of stride. All available steps were used to calculate the intrasubject variability of walking velocity, stride length, double support time and stride duration (which corresponds to that of cadence) using a variation coefficient (VC) calculated as
 |
Healthy subjects increase walking velocity by increasing both cadence and stride length in a stable linear relationship. As the observed changes in velocity in Parkinson's disease patients were due mainly to an increase in stride length, we calculated this contribution of stride length (Csl) as a percentage of the velocity increase according to the formula given in Ferrandez and Blin (1991):
 |
L is the difference between L and the stride length in the respective remaining condition (on medication or on stimulation or both), D is the stride duration during walking off medication and off stimulation, and D is the difference between D and the respective remaining condition. For all gait measurements, mean values were generated for each subject and each condition. Subject means were then used to generate group means and differences between therapeutic conditions. Each patient served as his own control using the –stim/–dopa condition as the baseline. To assess the effect of STN stimulation alone, the baseline values (–stim/–dopa) were subtracted from the +stim/–dopa values for each patient, and the mean differences and their 95% confidence intervals were calculated. Corresponding differences and confidence intervals were calculated to assess the effects of L-dopa and the additive and differential effects of STN stimulation. |
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Clinical data
The UPDRS motor scores of each patient for the different conditions are shown in Table 1
. The score (mean ± standard deviation) in the baseline condition of off-stimulation/off-medication was 49.4 ± 15.8, which is comparable with the preoperative value of 49.8 ± 14.6 in the off stage. With STN stimulation alone, the score improved to 7.4 ± 2.7, which is similar to the preoperative value of 7.6 ± 2.6 in the on stage. With L-dopa alone, the value was 13.6 ± 6.9. The combination of stimulation and L-dopa further improved the UPDRS motor score in all patients to 4.1 ± 1.8. The Tinetti gait score, which assesses gait clinically on a scale from 0 to 12, changed dramatically: off-stimulation/off-medication the value (mean ± SD) was 5.6 ± 2.7, whereas it was at its maximum value of 12 in all but two patients during all therapeutic conditions.
Walking velocity, stride length and cadence
After a short training period of <5 min, healthy subjects and all patients with stimulation on were able to perform free walking on the treadmill. The average walking velocity of the 12 age-matched healthy subjects was 1.07 m/s with a mean stride length of 1.15 m and a mean cadence (steps/min) of 112. The eight Parkinson's disease patients achieved a mean walking velocity of 0.35 m/s with a stride length of 0.34 m and a consecutive relative increase in cadence to 122 in their baseline off-stimulation/off-medication condition. The walking velocities of each of the eight Parkinson's disease patients for all four conditions are illustrated in Fig. 1.
|
Table 2
summarizes the mean values and standard deviations of all gait measures for healthy subjects and the patients in the different study conditions. Off-medication/off-stimulation, two patients were not able to walk the 120 stride cycles required for the full protocol. In this single condition for these two patients, only walking velocity, stride length and cadence could be assessed reliably from a few steps. To allow comparison between all four conditions, the mean values of the remaining measures are given in Table 2 only for those six patients for whom all data for all four conditions were available. The two patients who were virtually unable to walk off-stimulation/off-medication reached values similar to those for the other six patients in the three remaining conditions and are therefore not shown separately. For all eight patients, STN stimulation alone increased the mean walking velocity considerably from 0.35 to 0.96 m/s, which is almost in the range of healthy subjects. In most patients, this increase was achieved solely by an increase of almost threefold in stride length from 0.34 to 0.99 m, whereas there was a slight decrease in cadence from 122 to 118. L-Dopa alone had an effect similar to that of STN stimulation alone, with an increase in walking velocity to 0.94 and an increase in stride length to 0.92 m at a similar cadence of 121. STN stimulation together with administration of L-dopa further increased the mean walking velocity to 1.19 m/s and stride length to 1.20 m at an unchanged cadence of 119. Our Parkinson's disease patients improved their walking velocity mainly by increasing stride length in all therapeutic conditions. Some showed a decrease in cadence. This stride length–cadence relationship is illustrated in Fig. 2 for each patient.
|
|
Individual changes induced by the combination of STN stimulation and L-dopa are shown together with the values for the control subjects in Fig. 2
. As can be seen in the figure, the patients changed their stride length–cadence relationship so that it came within the range of the control subjects. The calculated contribution (mean ± standard deviation) of stride length (Csl) to the increase in walking velocity was 106 ± 22% for STN stimulation, 105 ± 25% for L-dopa treatment and 101 ± 10% for the combination of both treatments. Values above 100% are explained by a decrease in the mean cadence despite an increase in walking velocity. Thus, like L-dopa, STN stimulation is highly effective in increasing stride length.
To illustrate the amounts of change induced by stimulation and L-dopa, the mean differences between the three therapeutic conditions and the baseline (–stim/–dopa) are given in Table 3A–C. Absolute values and the differences as percentages of the baseline are shown together with the 95% confidence intervals. When they do not include zero, the data are set in bold type. The results for STN stimulation alone were similar to those for L-dopa alone. The combined effect was more pronounced than that of either treatment alone. Additional effects of either stimulation or L-dopa are shown in Table 3D–E. It should be noted that the data for the effect of adding STN stimulation to L-dopa treatment (Table 3D, difference +stim/+dopa versus –stim/+dopa) suggest a mild additional effect, whereas the confidence intervals for the effect of adding L-dopa treatment to stimulation always include zero (Table 3E, difference +stim/+dopa versus +stim/–dopa). This reflects the fact that all patients but one showed an additional effect of stimulation, whereas only three out of eight patients showed an additional improvement with L-dopa, to a walking velocity that was considerably higher than with STN stimulation alone (Patients 1, 5 and 7). As can be seen in Fig. 1, these three patients were also the best responders to L-dopa and showed the smallest effects of STN stimulation alone, and consequently the highest additional effects of L-dopa (Table 1 and Fig. 1). Also, the chronic use of L-dopa after surgery was greatest in Patients 1 and 7, who received a daily dose of 600 and 450 mg, respectively. The three best responders to STN stimulation (Patients 2, 3 and 4) did not require any L-dopa to reach a walking velocity similar to that of healthy subjects and they did not show better responses to L-dopa than the remaining patients. Patients 2 and 4 did not use any L-dopa after implantation. Thus, the chronic dose of L-dopa after surgery seems to depend on the improvement resulting from STN stimulation alone. There was no correlation between the general improvement after stimulation and the effect on gait.
|
Stance and swing phase, joint movements and intrasubject variability
With STN stimulation or L-dopa treatment, only slight changes were seen in cadence, whereas the other kinematic measures showed a twofold increase. Other kinematic measures changed secondarily to the increases in walking velocity and stride length. The clear reduction in the time spent in double support and the increase in the time spent in swing were explained sufficiently by the higher walking velocities in the therapeutic conditions. Swing velocity and the range of motion of the major leg joints also increased considerably for all treatment conditions (Table 2
). The changes in the range of motion of the leg joints throughout the stride cycle are illustrated for a representative patient in Fig. 3. Each line represents the mean value of ~120 strides. There was a marked increase in the range of motion for all joints in all therapeutic conditions towards the pattern of healthy control subjects.
|
STN stimulation reduced intrasubject variability, as expressed by the decreases of ~50% in mean variation coefficients for velocity, stride length and cadence. Again, the extent of improvement was similar for L-dopa. The combination of stimulation and L-dopa further decreased variability to the value for healthy subjects (Table 2
). The variability of the double-limb support time was less disturbed than in healthy subjects and therefore showed only moderate changes. The mean differences between the three therapeutic conditions and the baseline, together with their respective 95% confidence intervals (Table 3A–C), clearly demonstrate that STN stimulation and L-dopa induced general improvements in gait measures that were similar in extent. The best results were obtained by a combination of both treatments, with which the patients reached the level seen in the healthy subjects, although the difference between stimulation alone and stimulation with L-dopa suggests that not all patients require additional L-dopa (Table 3E). |
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
In the present study, STN stimulation induced an almost threefold increase in walking velocity by increasing stride length, while cadence was virtually unchanged. Walking at the same velocity, patients with Parkinson's disease make shorter strides than normal subjects (Morris et al., 1998
). However, cadence regulation in patients with Parkinson's disease is considered to be nearly normal as it does not differ from that seen in control subjects during natural walking (Ferrandez and Blin, 1991). There is a close link between walking velocity and kinematic measures that improve with STN stimulation and a consistent inverse relationship has been demonstrated between velocity and double-limb support time, i.e. the time spent with both feet on the ground (Blin et al., 1991; Elble et al., 1991). Regarding the minimal differences in cadence, most of the changes we observed with STN stimulation in double support time, swing phase, swing velocity and joint angle movements can be attributed to the increase in stride length (Elble et al., 1991). A low degree of variability in gait measures indicates high precision of the locomotor movement. According to Gabell and Nayak, the variabilities of stride length and cadence are related to gait patterning mechanisms and must be distinguished from the variability of the double-limb support time, which, they suggest, may be related to balance control mechanisms (Gabell and Nayak, 1984). In our study, the variabilities of stride length and cadence were clearly increased in Parkinson's disease patients compared with control subjects. Both measures improved considerably with STN stimulation, whereas the variability of double-limb support time was less increased and consequently showed a smaller change. The decrease in variability measures related to gait patterning mechanisms with STN stimulation indicates an increase in the stability of the walking pattern. Regarding the different gait measures, STN stimulation induced changes towards a normal gait pattern.
Comparison between STN stimulation, L-dopa treatment and other surgical procedures
We found similar results with almost identical mean values for a suprathreshold dose of L-dopa and STN stimulation. The combination of both treatments further improved the gait measures. This suggests a synergistic effect, which will be discussed below. Blin and colleagues and O'Sullivan and colleagues reported a 30% increase in walking velocity after administration of 250 mg L-dopa (Blin et al., 1991; O'Sullivan et al., 1998). This effect is much smaller than in the present study. In line with our findings, they reported an unchanged cadence. Laitinen found an increase in walking velocity of 29% after unilateral pallidotomy (Laitinen, 1994). After bilateral posteroventral pallidotomy, Siegel and Metman found a twofold increase in walking velocity (Siegel and Metman, 2000). About 80% of this increase was calculated to be due to changed stride length. In two preliminary reports, much weaker effects were described for chronic stimulation of the internal globus pallidus (GPi). After unilateral GPi stimulation, Melnick and colleagues found a velocity increase of 10–20% and no change in cadence (Melnick et al., 2000). After bilateral GPi stimulation, Allert and colleagues reported a velocity increase of ~50% (Allert et al., 2000); as far as can be judged from the data given, ~80% of this increase was due to the increase in stride length. For a group of Parkinson's disease patients receiving bilateral STN stimulation, Allert and colleagues reported an improvement in walking velocity of 120% (Allert et al., 2000). This improvement was due mainly to an increase in stride length, but the authors attributed ~10% of the improvement to an increase in cadence. Although detailed data from these studies are not available yet, this may point to a better efficacy of STN stimulation compared with other neurosurgical treatment options with respect to gait. Although relevant changes in cadence are suggested to be present after lesions or bilateral stimulation of the GPi, cadence is not affected by L-dopa or STN stimulation. An alternative explanation for these changes in cadence may be that the Parkinson's disease patients in these GPi studies were walking comparatively slowly and therefore were not yet at the break point of their linear cadence–stride length relationship (Morris et al., 1998). Consequently, some of the increase in velocity could still have been achieved by an increase in cadence without a consecutive decrease in stride length, as stride length does decrease (with increasing cadence) only beyond the break point. This cannot be judged from the abstracts cited above as not all the data required are given.
The contribution of stride length depends on the strategy used to increase walking velocity
In Parkinson's disease, different strategies can be used to increase walking velocity. A voluntary increase from a comfortable to a fast walking velocity in Parkinson's disease is achieved by increasing both stride length and cadence, provided the patient is not yet walking at his maximum stride length, i.e. if he is below the break point of his individual linear stride length–cadence relationship. Ferrandez and Blin reported that, without L-dopa, changes in stride length contributed 58% to a voluntary velocity increase from a comfortable speed to fast walking (Ferrandez and Blin, 1991). With L-dopa, stride length contributed 34% to a similar voluntary increase in velocity. When the subject was walking at a comfortable speed, stride length contributed 91% to the L-dopa-related increase in velocity. For the fast walking velocity, the contribution of stride length to the L-dopa-related increase in velocity was 76%. Again, it has to be noted that during fast walking the patients were probably walking beyond their break point, so that they were unable to increase their stride length further. Thus, the study of Ferrandez and Blin does not exclude the possibility that L-dopa merely affects stride length. It can be argued that there is a training effect if gait testing is repeated within a few hours. In Parkinson's disease, stride length can be increased to values similar to those seen in healthy subjects by visual cues or attentional strategies for at least 2 h, but the gait performance reverts to baseline within 1 day. Cadence is not affected by these cues (Morris et al., 1996; Prokop and Berger, 1996; Azulay et al., 1999). In view of the crossover design used in the present study, we can exclude the possibility that these strategies influence the effect seen for STN stimulation, because starting with and without stimulation was randomized and performed on different days. However, the effect of L-dopa may have been overestimated in our study design because all conditions including L-dopa treatment were performed after the corresponding trials without L-dopa, on the same day. The potential learning effect could have contributed to the changes observed after administration of L-dopa. Taking account of our study design, we can conclude that STN stimulation is at least as effective in improving gait as a suprathreshold dose of L-dopa, and is probably more effective. The effects of L-dopa we found were much stronger than those reported by other groups (Blin et al., 1991; O'Sullivan et al., 1998). Possible reasons for this difference in effectiveness may be the above-mentioned overestimation of the effect of L-dopa in the present study and the fact that we used a suprathreshold dose of L-dopa. Furthermore, one of our inclusion criteria for the implantation of STN stimulation electrodes was a good response to L-dopa. Finally, our patients were considerably younger than the populations taking part in other studies, in which the mean age was usually above 60 years, so that in our population less physical restraint could be expected.
Do STN stimulation and L-dopa have synergistic effects on gait?
Bejjani and colleagues used the UPDRS motor score in 10 Parkinson's disease patients, and suggested a synergistic effect of STN stimulation and L-dopa for axial parkinsonian symptoms (Bejjani et al., 2000). For the gait subscore, they found no significant additional effect of the combination. However, gait in the UPDRS subscore is assessed from 0 to 4, which might not be sensitive enough. In the present study, all patients except one showed additional improvement with the combination of treatments compared with a suprathreshold dose of L-dopa alone. Taking into account the overestimation of the effect of L-dopa compared with STN stimulation in our study, this may point to a synergistic effect. Furthermore, our results indicate that the effect of STN stimulation on gait is not necessarily closely related to the effect of L-dopa. The three best responders to STN stimulation alone did not show better responses to L-dopa alone than the remaining patients and the three worst responders to L-dopa did not show worse responses to STN stimulation. In selected patients, STN stimulation alone could improve gait to the level of healthy subjects without additional L-dopa, probably because of optimal electrode positioning and stimulation parameters. The fact that these patients were not necessarily the best L-dopa responders off stimulation also supports the possibility of a synergistic effect. In view of the pathophysiological considerations discussed below, STN stimulation may not influence the same pathways as L-dopa.
Pathophysiological considerations
The pathophysiology of the disturbance of stride length regulation and its modification by dopaminergic medication or deep-brain stimulation remains speculative. The symptoms involved in Parkinson's disease include bradykinesia, akinesia, rigidity, tremor, festinating gait and masked facies. The primary cause of these deficits is the degeneration of neurones in the substantia nigra. However, this wide range of deficits appears to reflect disorders in the function of several separate pathways within the basal ganglia. As far as walking is concerned, it has been proposed from animal work that the motor cortex and basal ganglia may undertake a series of functions which provide the spatial framework necessary for the performance of sequences of voluntary movements involved in locomotion (Garcia-Rill, 1986). Morris and colleagues hypothesized that there are two key roles of the basal ganglia in motor control. First, they provide a phasic cue to the supplementary motor area that is timed to terminate set-related activity in this area; therefore, they play a role in the maintenance of adequate movement preparation for submovements performed in a sequence. Secondly, the basal ganglia contribute to motor set for whole movement sequences and the absence of their contribution may lead to inadequate gain in motor execution mechanisms, which could result in the underscaling of successive steps during gait (Morris et al., 1996). The fact that stride length responds to attentional strategies indicates that stride length deficits in Parkinson's disease relate to problems in motor set rather than a deficit in the internal cueing of human locomotion (Morris et al., 1996). The existence of spinal locomotion oscillators also in humans has been deduced from studies in paraplegic patients (Dietz et al., 1994). The pedunculopontine nucleus (PPN) as a site of termination for basal ganglia outputs was proposed to be equivalent to the mesencephalic locomotor region as the area that modulates spinal locomotion oscillators. It was suggested that the descending output of the PPN is cholinergic and may be under GABAergic control, which, in turn, may be under subthalamic influence (Garcia-Rill, 1986). It was argued that afferents to the PPN from the motor cortex and the basal ganglia are disturbed in Parkinson's disease, and lack of regulation is then reflected in an inability to start or stop locomotion, whereas the stepping mechanism as such is not affected in the disorders of locomotion associated with diseases of the basal ganglia (Garcia-Rill, 1986; Morris et al., 1994b). It is assumed that stride length is controlled supraspinally by phasic output from the basal ganglia to the supplementary motor area, whereas spinal and brainstem mechanisms may play an important role in cadence (Morris et al., 1994a, 1996). A further argument supporting the supraspinal regulation of stride length comes from results in patients suffering from normal pressure hydrocephalus. After a CSF test or after shunting, these patients increased their walking velocity. As in Parkinson's disease patients, this was achieved almost exclusively by increasing stride length while cadence remained unchanged (Faist et al., 1997; Stolze et al., 2000). The fact that attentional strategies and visual cues can elicit normal stride length in Parkinson's disease (Morris et al., 1996; Prokop and Berger, 1996; Azulay et al., 1999) supports the view that alternative pathways that can influence spatial parameters in locomotion are still intact. STN stimulation may inhibit the disturbances in these alternative pathways and enable the patients to regain their ability to modulate stride length. It has to be kept in mind that STN stimulation and other neurosurgical interventions in Parkinson's disease do not re-establish disturbed pathways directly but rather block overactive connections. The original function of these blocked pathways is therefore disturbed. The long-term effects of STN stimulation are not yet known. The basal ganglia play a role in allowing movement performance to shift from conscious to automatic control. In dual-task conditions, Parkinson's disease patients also show a decline in walking performance if they have to perform a second task at the same time, such as reciting sentences (Morris et al., 1996). Future studies with STN should be directed to revealing possible deficits induced by STN stimulation. It would be of interest to discover whether, in such dual-task conditions, Parkinson's disease patients behave differently with STN stimulation than with L-dopa treatment. This information may shed light not only on locomotor mechanisms.
In conclusion, the results presented here demonstrate that chronic bilateral STN stimulation improved gait in a manner similar to L-dopa, by increasing stride length while cadence is unchanged. Excursion of the lower extremity joints in the sagittal plane also increased to normal values. Measurements associated with the timing of gait including the duration of the subphases were unchanged when the increase in stride length and velocity was taken into account. The present results confirm that locomotion disorders in Parkinson's disease affect mechanisms involved in the control of amplitude rather than rhythmicity. The gait measures obtained with STN stimulation alone are in the range of control subjects. Thus, chronic STN stimulation is effective in improving Parkinsonian gait and in some patients no additional L-dopa is required.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
The authors wish to thank all of the patients who gave their time for these experiments, U. Römmelt for excellent technical assistance and R. Rossner for statistical analysis. The Grenoble team was supported by INSERM.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionAcknowledgementsReferences |
|---|
Allert N, Volkmann J, Sturm V, Hefter H, Freund HJ. Improvement of gait in Parkinson patients treated by bilateral stimulation of subthalamic nucleus or internal pallidum [abstract]. Mov Disord 2000; 15 Suppl 3: 51.
Azulay JP, Mesure S, Amblard B, Blin O, Sangla I, Pouget J. Visual control of locomotion in Parkinson's disease. Brain 1999; 122: 111–20.
Bejjani BP, Gervais D, Arnulf I, Papadopoulos S, Demeret S, Bonnet AM, et al. Axial parkinsonian symptoms can be improved: the role of levodopa and bilateral subthalamic stimulation. J Neurol Neurosurg Psychiatry 2000; 68: 595–600.
Blin O, Ferrandez AM, Pailhous J, Serratrice G. Dopa-sensitive and dopa-resistant gait parameters in Parkinson's disease. J Neurol Sci 1991; 103: 51–4.[ISI][Medline]
Defebvre L, Blatt JL, Blond S, Bourriez JL, Gieu JD, Destee A. Effect of thalamic stimulation on gait in Parkinson disease. Arch Neurol 1996; 53: 898–903.[Abstract]
Dietz V, Colombo G, Jensen L. Locomotor activity in spinal man. Lancet 1994; 344: 1260–3.[ISI][Medline]
Elble RJ, Thomas SS, Higgins C, Colliver J. Stride-dependent changes in gait of older people. J Neurol 1991; 238: 1–5.[ISI][Medline]
Faist M, Schubert M, Borremans J, Krauss JK, Berger W, Lücking CH. Changes in gait after shunt operation in idiopathic normal pressure hydrocephalus (NPH) [abstract]. Electroencephalogr Clin Neurophysiol 1997; 103: 132.
Faist M, Xie J, Berger W, Pollak P, Lücking CH. Effect of chronic subthalamic nucleus stimulation on gait in Parkinson's disease [abstract]. Mov Disord 2000; 15 Suppl 3: 69.
Ferrandez AM, Blin O. A comparison between the effect of intentional modulations and the action of L-dopa on gait in Parkinson's disease. Behav Brain Res 1991; 45: 177–83.[ISI][Medline]
Gabell A, Nayak US. The effect of age on variability in gait. J Gerontol 1984; 39: 662–6.[ISI][Medline]
Garcia-Rill E. The basal ganglia and the locomotor regions. [Review]. Brain Res 1986; 396: 47–63.[Medline]
Krystkowiak P, Defebvre L, Blatt JL, Bourriez JL, Blond S, Guieu JD, et al. Influence of subthalamic nucleus stimulation on gait in Parkinson's disease: a study using the optoelectronic VICON system [abstract]. Mov Disord 2000; 15 Suppl 3: 48.[ISI][Medline]
Laitinen LV. Ventroposterolateral pallidotomy. Stereotact Funct Neurosurg 1994; 62: 41–52.[Medline]
Limousin P, Krack P, Pollak P, Benazzouz A, Ardouin C, Hoffmann D, Benabid AL. Electrical stimulation of the subthalamic nucleus in advanced Parkinson's disease. N Engl J Med 1998; 339: 1105–11.
Melnick M, Radtka S, Alsbury N, Peterson J, Pisenti M, Christine C, et al. The effect of unilateral deep brain stimulation on gait in patients with Parkinson's disease [abstract]. Mov Disord 2000; 15 Suppl 3: 64.
Morris ME, Iansek R, Matyas TA, Summers JJ. The pathogenesis of gait hypokinesia in Parkinson's disease. Brain 1994a; 117: 1169–81.
Morris ME, Iansek R, Matyas TA, Summers JJ. Ability to modulate walking cadence remains intact in Parkinson's disease. J Neurol Neurosurg Psychiatry 1994b; 57: 1532–4.
Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson's disease. Normalization strategies and underlying mechanisms. Brain 1996; 119: 551–68.
Morris M, Iansek R, Matyas T, Summers J. Abnormalities in the stride length–cadence relation in Parkinsonian gait. Mov Disord 1998; 13: 61–9.[ISI][Medline]
O'Sullivan JD, Said CM, Dillon LC, Hoffman M, Hughes AJ. Gait analysis in patients with Parkinson's disease and motor fluctuations: influence of levodopa and comparison with other measures of motor function. Mov Disord 1998; 13: 900–6.[ISI][Medline]
Prokop T, Berger W. Influence of optic flow on locomotion in normal subjects and patients with Parkinson's disease [abstract]. Electroencephalogr Clin Neurophysiol 1996; 99: 370.
Schwab RS, England AC. Projection technique for evaluating surgery in Parkinson's disease. In: Gillingham FJ, Donaldson MC, editors. Third symposium on Parkinson's disease. Edinburgh: ES Livingston; 1969. p. 152–7.
Siegel KL, Metman LV. Effects of bilateral posteroventral pallidotomy on gait of subjects with Parkinson disease. Arch Neurol 2000; 57: 198–204.
Stolze H, Kuhtz-Buschbeck JP, Drücke H, Jöhnk K, Diercks C, Palmié S, et al. Gait analysis in idiopathic normal pressure hydrocephalus—which parameters respond to the CSF tap test? Clin Neurophysiol 2000; 111: 1678–86.[ISI][Medline]
Tinetti ME. Performance-oriented assessment of mobility problems in elderly patients. J Am Geriatr Soc 1986; 34: 119–26.[ISI][Medline]
Yokoyama T, Sugiyama K, Nishizawa S, Yokota N, Ohta S, Uemura K. Subthalamic nucleus stimulation for gait disturbance in Parkinson's disease. Neurosurgery 1999; 45: 41–9.[ISI][Medline]
Received January 31, 2001. Revised April 18, 2001. Accepted April 18, 2001.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Crenna, I. Carpinella, L. Lopiano, A. Marzegan, M. Rabuffetti, M. Rizzone, M. Lanotte, and M. Ferrarin Influence of basal ganglia on upper limb locomotor synergies. Evidence from deep brain stimulation and L-DOPA treatment in Parkinson's disease Brain, December 1, 2008; 131(12): 3410 - 3420. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Chastan, G. W. M. Westby, J. Yelnik, E. Bardinet, M. C. Do, Y. Agid, and M. L. Welter Effects of nigral stimulation on locomotion and postural stability in patients with Parkinson's disease Brain, November 11, 2008; (2008) awn294v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Sailer, D. I. Cunic, G. O. Paradiso, C. A. Gunraj, A. Wagle-Shukla, E. Moro, A. M. Lozano, A. E. Lang, and R. Chen Subthalamic nucleus stimulation modulates afferent inhibition in Parkinson disease Neurology, January 30, 2007; 68(5): 356 - 363. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Colnat-Coulbois, G C Gauchard, L Maillard, G Barroche, H Vespignani, J Auque, and P. P Perrin Bilateral subthalamic nucleus stimulation improves balance control in Parkinson's disease J. Neurol. Neurosurg. Psychiatry, June 1, 2005; 76(6): 780 - 787. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Sturman, D. E. Vaillancourt, L. V. Metman, R. A. E. Bakay, and D. M. Corcos Effects of subthalamic nucleus stimulation and medication on resting and postural tremor in Parkinson's disease Brain, September 1, 2004; 127(9): 2131 - 2143. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Maurer, T. Mergner, J. Xie, M. Faist, P. Pollak, and C. H. Lucking Effect of chronic bilateral subthalamic nucleus (STN) stimulation on postural control in Parkinson's disease Brain, May 1, 2003; 126(5): 1146 - 1163. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Krystkowiak, J.-L. Blatt, J.-L. Bourriez, A. Duhamel, M. Perina, S. Blond, J.-D. Guieu, A. Destee, and L. Defebvre Effects of Subthalamic Nucleus Stimulation and Levodopa Treatment on Gait Abnormalities in Parkinson Disease Arch Neurol, January 1, 2003; 60(1): 80 - 84. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||