Enhanced lateral premotor activity during paradoxical gait in Parkinson\'s disease

Enhanced Lateral Premotor Activity During Paradoxical Gait in Parkinson’s Disease Takashi Hanakawa, MD,* Hidenao Fukuyama, MD,* Yukinori Katsumi, MD,† Manabu Honda, MD,* and Hiroshi Shibasaki, MD*

Parkinson’s disease (PD) patients often show marked improvement of hypokinetic gait when exposed to special stimuli. To investigate physiological mechanisms underlying this “paradoxical gait” induced by visual cues in PD patients, we examined regional cerebral blood flow changes during gait on a treadmill guided by two different visual cues, the lines oriented transversely to the direction of walk (TL) and the lines parallel to it (PL). Ten PD patients and 10 age-matched controls received injections of 99mTc-hexamethylpropyleneamine oxime twice, once during each walking condition. Brain perfusion images were obtained by single-photon emission computed tomography. When affected by TL, PD patients showed marked improvement of gait parameters, mainly reduction of cadence. In regional cerebral blood flow analysis, when TL was compared with PL, both groups had common activation in the posterior parietal cortex and cerebellar hemispheres. Especially in the right lateral premotor cortex, PD patients showed enhanced activation induced by TL to a significantly greater degree than the controls. The present study indicates that the network dedicated to visuomotor control, particularly the lateral premotor cortex, plays an important role in the development of the paradoxical gait induced by special visual stimuli in PD patients. Hanakawa T, Fukuyama H, Katsumi Y, Honda M, Shibasaki H. Enhanced lateral premotor activity during paradoxical gait in Parkinson’s disease. Ann Neurol 1999;45:329 –336

Patients with Parkinson’s disease (PD), suffering from considerable akinesia, can often perform surprisingly quick movements when exposed to special sensory or emotional stimuli, the phenomenon called “kine´sie paradoxale” or paradoxical movement.1 Among impressive anecdotes about paradoxical movement,2 the best documented example is a marked improvement of parkinsonian gait when the patients step across the lines placed transversely to the walking direction at appropriate intervals, whereas the lines parallel to the walking direction are totally ineffective.3 The effect of the transversely oriented visual cues on parkinsonian gait was verified by gait analysis4,5 and by clinical usefulness of an L-shaped walking stick.6 However, as few investigators have studied this issue, we do not know how the visual cues induce paradoxical gait. Single-photon emission computed tomography (SPECT) with the split-dose method of 99mTchexamethylpropyleneamine oxime (99mTc-HMPAO) detects regional cerebral blood flow (rCBF) changes during activation tasks.7–9 Based on characteristics of 99m Tc-HMPAO,10 we can obtain “snapshot” rCBF images that reflect neural activity over a period of several minutes after the tracer administration.7 This allows us to study human brain functions without constraining the subjects while performing tasks.11,12 From the Departments of *Brain Pathophysiology and †Neurology, Graduate School of Medicine, Kyoto University, Kyoto, Japan. Received Jul 6, 1998, and in revised form Oct 13. Accepted for publication Oct 14, 1998.

If specific brain areas actively function to develop paradoxical gait, those areas would show greater rCBF increase during paradoxical gait than during nonenhanced gait. In this regard, the role of the lateral premotor cortex (PMC) is of particular interest, because the PMC is preferentially active during externally cued movement relative to noncued movement.13,14 Furthermore, the PMC and parietal overactivity was reported while PD patients performed sensory-cued motor tasks.15 Another line of evidence suggests that the cerebellum, receiving visual information from the posterior parietal cortex (PPC), may subserve paradoxical movement.16 Therefore, we hypothesized that a brain circuit for visuomotor control, including the PMC, cerebellum, and PPC as the key nodes, might function in the development of visually cued paradoxical movement. Based on this hypothesis, we compared rCBF change between PD patients and age-matched controls during treadmill walk cued by two different visual stimuli, one inducing paradoxical gait and the other not. Subjects and Methods Subjects Ten patients with clinically diagnosed PD17 (6 men and 4 women; age, 68 6 6 years [mean 6 SD]; height, 155 6 9 cm; weight, 51 6 11 kg) and 10 matched controls (5 men Address correspondence to Dr Shibasaki, Department of Brain Pathophysiology, Graduate School of Medicine, Kyoto University, 54 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606-8507, Japan.

Copyright © 1999 by the American Neurological Association

329

and 5 women; 67 6 6.0 years; height, 156 6 12 cm; weight, 51 6 9.4 kg) participated in this study. Patients showing on–off phenomenon, dementia, or orthostatic hypotension were excluded before registration. Hoehn and Yahr score and Unified Parkinson’s Disease Rating Scale,18 and the degree of paradoxical gait, were assessed for each patient (Table 1). Antiparkinsonian medications were not interrupted for this study. All control subjects were neurologically normal without any previous history of neuropsychiatric disorders. All subjects gave written informed consent based on the study protocol approved by the ethics committee in our institute.

Performance Evaluation The walking performance was recorded on videotapes for subsequent review. Cadence (steps/min) was measured for 2 minutes while the treadmill kept moving at the steady speed. Consequently, variance of cadence was systematically different between the two conditions (see Fig 1B). Thus, the difference in cadence between the two conditions (TL effect; cadence PL 2 cadence TL) was calculated for each subject and compared between the groups with nonparametric statistics. Mean stride length for each subject was obtained by dividing the walking speed by half the mean cadence. After completing both tasks, the subjects evaluated effects of task condition on their gait subjectively. Mean blood pressure and pulse rate, measured before and after each task, were analyzed by analysis of variance.

Tasks Studied were two task conditions in which the subjects walked on a treadmill guided by two different visual cues, the lines transversely oriented to the walking direction (TL) and those parallel to it (PL) (Fig 1A). In TL, the subjects stepped across each line, which was 60 cm in length and placed at regular intervals (30 cm). In PL, they walked with their preferred stride length between the two lines placed 30 cm apart from each other. The lines were made of white cloth tape (3 cm in width) fixed on a black artificial rubber floor of the treadmill. The total length of the lines on the treadmill was identical (6 m) in the two conditions. All subjects walked on the treadmill at a steady speed of 13 m/min for 5 min. The walking speed and the intervals of the lines were determined so that the most hypokinetic patient in this series could walk without much difficulty after a brief practice. While walking, the subjects gently held side bars of the treadmill for safety. To balance the intensity of visual stimuli and the attention to the floor between the two tasks, the subjects were instructed to pay attention to both kinds of lines equally.

Image Acquisition Each subject underwent two consecutive SPECT scans.11,12 The order of task conditions was counterbalanced within each group to reduce the sequence effects, especially the carryover effects of paradoxical gait.4 Thirty seconds after starting the first task, the subjects were injected with 259 MBq of 99m Tc-HMPAO through a venous line fixed on their left forearm. After lying supine on a scanner bed, the subject’s head was positioned by using a three-dimensional laserreference system. They were scanned for 30 minutes by using a triple-head SPECT scanner (PRISM3000, Picker, Cleveland, OH) with high-resolution fan-beam collimators. After the first scan that produced low-dose images, the subjects returned to the treadmill. The subjects received another 777 MBq of 99mTc-HMPAO 30 seconds after starting the second task. After reposition, they were scanned for 15 minutes by the same scanner, which generated high-dose images. The interscan interval was approximately 1 hour.

Table 1. Subject Profiles Subjects

Age (yr)

Parkinson’s disease 1 2 3 4 5 6 7 8 9 10 Mean (n 5 10) SD

62 71 67 74 60 77 71 71 66 62 68 6

Control Mean (n 5 10) SD

67 6

Sex M F M F F M M F F M

Duration (yr)

H–Y Score

UPDRS III(/108)

Paradoxical Gait

L-Dopa 1 DCI (mg)

4 3 10 3.5 1 5 8 10 5 5 5.5 3.0

2.5 3 2.5 4 3 4 4 3 2.5 2 3 0.7

12 14 56 59 32 33 22 40 16 15 30 17

6 1 1 11 11 11 11 11 1 6

200 300 500 300 300 400 200 500 300 300 330 105

DA Agonist (mg)

Bro 7.5; Per 0.4 Bro 7.5 Per 0.25 Bro 10; Per 0.1 Bro 2.5 Bro 10

H–Y Score 5 modified Hoehn and Yahr rating score; UPDRS III 5 Unified Parkinson’s Disease Rating Scale motor examination (full score, 108); Paradoxical Gait 5 improvement of gait by transverse lines on a floor as assessed by a neurologist; DCI 5 decarboxylase inhibitors of aromatic L-amino acids (benserazide or carbidopa); DA Agonist 5 dopamine agonists; Bro 5 bromocriptine; Per 5 pergolide; 6 5 equivocal change; 1 5 obvious improvement; 11 5 marked improvement.

330

Annals of Neurology

Vol 45

No 3

March 1999

Fig 1. Two different visual cues placed on the treadmill (A), and mean cadence and standard deviation in each condition in control subjects and Parkinson’s disease (PD) patients (B). (A) White cloth tape was fixed on the floor of a treadmill in two different orientations, transverse (TL) and parallel (PL) to the walking direction. (B) The two groups showed almost identical cadence with little variation in TL (open column). Although both groups had higher cadence in PL (filled column) compared with TL, the difference of cadence between the two conditions was greater in PD patients than in the control subjects (see text). On a workstation (ODYSSEY; Picker, Cleveland, OH), projection data of the scans collected in a 128 3 128 matrix were prefiltered with a Butterworth filter. Then, transaxial images approximately parallel to the intercommissural line were reconstructed by using a Ramp filter with the postreconstruction attenuation correction. Finally, reconstruction yielded approximately 2-mm cubic voxels, with a 128 3 128 matrix and 64 slices. In-plane resolution was 7.8 mm full width at half maximum (FWHM) in the center of view.

Data Analysis After the reconstructed images were transferred to a SPARCstation20 (Sun Microsystems, Mountain View, CA), the images were analyzed by using Statistical Parametric Mapping

(SPM96; Wellcome Department of Cognitive Neurology, Institute of Neurology, London, UK)19 implemented on MATLAB (Math Works, Sherborn, MA). Each low-dose image was realigned by using the corresponding high-dose image as a reference to minimize the effect of reposition.20 Then both images were transformed to fit the standard stereotaxic space by using the mean image of the high-dose and low-dose images as a reference.20,21 The transformed images were smoothed by using an isotropic Gaussian kernel (FWHM 5 20 mm). Differences in global activity between scans were removed by scaling the activity in each pixel proportionally to the global activity. Mean global activity of each scan was adjusted to 50. Planned comparisons between conditions were performed on a pixel-bypixel basis by t statistics, generating SPM(t) subsequently transformed to the unit normal distribution [SPM(Z)]. First, TL compared with PL (TL 2 PL) was calculated separately within each group. The reverse contrast (PL 2 TL) was also tested within each group. Second, we tested an interaction of the task effects by the groups: (TL 2 PL in PD patients) 2 (TL 2 PL in controls). This comparison can reflect either “increase by TL in PD patients” or “decrease by TL in controls,” or both. To focus on the positive effects of TL, the analysis was performed in the selected voxels showing increased activity in TL 2 PL in PD patients (cutoff; Z 5 1.64). In a similar manner, we calculated the reverse comparison: (TL 2 PL in controls) 2 (TL 2 PL in PD patients) only in the voxels increasing activity in TL 2 PL in the controls. Our preexisting hypothesis was that areas showing overactivation during paradoxical gait would include the PMC, PPC, and cerebellum as described in the introductory section. In (TL 2 PL in PD patients) 2 (TL 2 PL in controls), we investigated those areas by using a height threshold at a Z score of 1.96, corresponding to an uncorrected p , 0.025. In other comparisons, a height threshold was set at a Z score of 3.09, corresponding to an uncorrected p , 0.001. Without a correction for multiple comparisons, the findings were somewhat descriptive. However, considering that the study was performed without repeated measures by SPECT whose sensitivity and spatial resolution are inferior to positron emission tomography,12 surviving areas are considered to be meaningful in the presence of preexisting anatomicophysiological background.

Results Performance Observation All subjects completed both tasks. In PD patients, preferred walking condition was TL in 7 patients, PL in 1, and equal in 2, and in the control subjects, TL in 2, PL in 2, and equal in 6, indicating that PD patients obviously preferred TL to PL ( p , 0.01, x2). On reviewing video, some PD patients dragged, shuffled, or just managed to keep up the pace in PL, although all the patients could walk smoothly in TL. One patient (PD 5) transiently showed a freezing phenomenon and required a temporal assistance in PL but walked fairly well throughout TL. No other patients showed freezing or festination during the tasks. The mean cadence was almost identical between the two groups in TL, indi-

Hanakawa et al: Paradoxical Gait in Parkinson’s Disease

331

cating good task performance in both groups (see Fig 1B). In PL, all the PD patients had higher cadence compared with TL, although all control subjects but 1 also showed similar tendency. The TL effect was significantly greater in PD patients (40.6 6 20.0) than in the controls (12.1 6 10.0) (Mann–Whitney’s U test, p , 0.005). A group average of the stride length in PL was shorter in PD patients (0.31 m) than in the controls (0.46 m), whereas that in TL was similar between the groups (PD patients, 0.59 m; control subjects, 0.58 m). Changes of mean blood pressure and pulse rate were not significantly different between the conditions or between the groups. rCBF Analysis In the within-group comparisons for TL 2 PL, the controls showed a significant rCBF increase in the bilateral PPC (Brodmann areas [BA] 7, 40, and 39), left dorsolateral prefrontal cortex (DLPFC; BA 10), left insula, and left cerebellar hemisphere (Table 2 and Fig 2A). In TL 2 PL in PD patients, brain activity was increased in the bilateral PPC (BA 7, 40), left cerebellar hemisphere, right PMC (BA 6), and right anterior cingulate gyrus (BA 32) (see Table 2 and Fig 2B). In PL 2 TL, there was no significant deactivation by TL in both groups, although compensatory activity due to higher cadence in PL than in TL was expected especially in PD patients.

Table 2. Areas Showing Significant Activation by Within-Group Analysis (TL Minus PL) Size (k)

Anatomical Location (Brodmann Area)

Control subjects 1 464 R precuneus (7) 2 449 L medial frontal gyrus (10) 3 406 L inferior parietal lobule (40, 39) 4 5

299 330

L insula L cerebellar hemisphere

Parkinson’s disease patients 1 402 L cerebellar hemisphere 2 1,801 R middle occipital gyrus (19) R inferior parietal lobule (7) 3 615 L precuneus (7) 4 841 R middle frontal gyrus (PMC; 6) 5 848 R anterior cingulate gyrus (32) 6 146 L inferior parietal lobule (40)

Coordinate Z Score

x

y

% of CBF Changesa

3.84 3.48

8 214

278 54

44 6

1.48 1.41

3.41

240

266

34

1.21

3.36 3.19 3.10

246 238 222

256 12 256

44 2 246

1.05 2.38 1.44

4.33

216

276

240

1.11

3.80

32

280

16

1.71

3.07

26

262

46

2.43

3.75 3.74

28 46

258 26

58 60

1.20 2.39

3.12

22

40

24

1.24

3.10

228

244

38

1.98

z

Numbers in the leftmost column correspond to the region numbers in Figure 2. a Mean regional cerebral blood flow (rCBF) increase in percentage in each corresponding voxel; TL 5 treadmill walk guided by transverse lines; PL 5 treadmill walk guided by parallel lines; R 5 right; L 5 left; PMC 5 lateral premotor cortex.

332

Annals of Neurology

Vol 45

No 3

March 1999

Fig 2. Brain activation maps during treadmill walk guided by the transverse lines (TL) compared with the parallel lines (PL). Depicted are three orthogonal projections of the areas showing a significant regional cerebral blood flow increase in TL compared with PL in controls (A) and in the Parkinson’s disease (PD) patients (B). We considered a Z score greater than 3.09 (uncorrected p , 0.001) to be significant, but in this figure the threshold was set at a Z score of .2.33 for display purposes only. In PD patients, the left prefrontal ( Z 5 2.77) and the right cerebellar ( Z 5 2.96) region did not reach statistical significance. Numbers correspond to the areas listed in Table 2.

In (TL 2 PL in PD patients) 2 (TL 2 PL in controls), the right PMC (BA 6) showed a significantly greater rCBF increase provoked by TL in PD patients than in the controls (Table 3 and Fig 3). This region, included in the PMC activation revealed by the withingroup analysis in PD patients (see Fig 2B), was the

Table 3. Areas Showing Significant Interaction by TL Between the Groups Size (k)

Anatomical Location (Functional Area; Brodmann Area)

PD . controlsb 154 R precentral gyrus (PMC; 6) R middle frontal gyrus (PMC; 6) Controls . PDc 120 L superior frontal gyrus (DLPFC; 9)

% of CBF Changesa

Coordinate Z Score

x

y

z

PD

Controls

2.51 2.22

60 44

2 22

38 58

1.19 2.19

20.89 0.12

3.67

212

54

8

20.79

1.30

a Mean regional cerebral blood flow (rCBF) increase in percentage in each corresponding voxel; PD 5 Parkinson’s disease; TL 5 treadmill walk guided by transverse lines; PL 5 treadmill walk guided by parallel lines; R 5 right; L 5 left; PMC 5 lateral premotor cortex. b (TL 2 PL in PD patients) 2 (TL 2 PL in controls). c (TL 2 PL in controls) 2 (TL 2 PL in PD patients).

DLPFC 5 dorsolateral prefrontal cortex.

patients), activation of the DLPFC was greater in the controls than in the PD patients (see Table 3).

Fig 3. Enhanced activation by TL in Parkinson’s disease (PD) patients compared with the controls: (TL 2 PL in PD patients) 2 (TL 2 PL in controls), superimposed on the surface of the standard brain. In the right lateral premotor cortex region (Brodmann area 6), PD patients had greater regional cerebral blood flow increases induced by TL compared to the controls. TL, treadmill walk guided by transverse lines; PL, treadmill walk guided by parallel lines.

only surviving blob beyond the criteria in the whole brain. In this PMC region where two local maxima existed, the PD patients revealed an increased activity in both the dorsal and the ventral parts, whereas the controls showed only slightly increased activity in the dorsal part but a slightly decreased activity in the ventral part (Fig 4). Among these, the dorsal part was located nearby the local maximum point in the PMC region in TL 2 PL in PD patients (see Tables 2 and 3). In (TL 2 PL in controls) 2 (TL 2 PL in PD

Discussion Parkinsonian gait is characterized by a short stride length compensated for by a high cadence,22–24 which was substantiated in the present PL paradigm, contrasted by the marked cadence reduction in TL. By analyzing the TL effect, the improvement of parkinsonian gait by the transverse lines was quantitatively demonstrated. The patients clearly recognized the gaitenhancing effects of the transverse lines. Thus, these observations allow us to interpret the present rCBF data in close relation to paradoxical gait. The present SPECT study disclosed rCBF changes reflecting the two different gait conditions. The PPC and cerebellum, both of which showed greater activity in TL than in PL commonly in the two groups, may represent the essential anatomical basis for performing the TL task. Associated with the gait improvement by TL, the PD patients showed unique activation in the PMC, which agreed with our initial hypothesis. Before discussing the functional relevance of the PMC, a confounding effect of the cadence changes in PD patients should be addressed. The apparent activation by TL could actually mean a greater rCBF decrease in PL than in TL, if the PMC activity were inversely related to cadence. However, this is unlikely because the controls, who had small but similar cadence change, did not show corresponding rCBF change (see Fig 4). We also reanalyzed the data from our previous SPECT study, which compared the simple treadmill walk with the rest.24 In PD patients, the gait-induced rCBF change was negligible in the dorsal PMC (mean rCBF change, 20.03%; Z score, 20.16), although that was somewhat reduced in the ventral PMC (mean rCBF change, 21.82%; Z score, 21.75). Therefore, at least in the dorsal PMC where lower limbs and trunk are represented,25 the brain activity changes in TL 2 PL

Hanakawa et al: Paradoxical Gait in Parkinson’s Disease

333

Fig 4. Brain activity changes in the ventral and dorsal regions of the right lateral premotor cortex (PMC) in Parkinson’s disease (PD) patients and control subjects. Adjusted brain activities, sampled in the representative voxels in ventral and dorsal parts of the PMC, are plotted. In PD patients, the mean regional cerebral blood flow (rCBF) is increased in the ventral as well as dorsal PMC in TL relative to PL. In the controls, rCBF is equivocally increased in the dorsal PMC, but decreased in the ventral PMC in TL relative to PL. TL, treadmill walk guided by transverse lines; PL, treadmill walk guided by parallel lines.

in PD patients cannot be explained simply by cadence reduction. Predictive sensory cues, such as serial transverse lines or regular sounds,26 may provide a trigger to switch from one step to the next during walk. Alternatively,

334

Annals of Neurology

Vol 45

No 3

March 1999

the transverse lines might help PD patients adequately scale the stride length, which is otherwise abnormally short, possibly due to underestimated muscle activity required to produce appropriate movement size.27 The PMC is known to function in these sensory-cued movements.13,14 Despite limited knowledge about human gait control, the PMC seems to be involved in bilateral coordination of axial and proximal limb muscles through projection to the brainstem reticular nuclei.28 In fact, the PMC lesion in humans disturbed performance of the bicycling movement.29 Functional underactivation of the cortical motor areas, through diminished thalamocortical facilitation, explains hypokinesia in PD patients.30 Among cortical motor areas, dysfunction of the supplementary motor area (SMA), strongly influenced by inputs from the basal ganglia, plays a substantial role in the pathophysiology of PD patients.31–33 By contrast, the PMC, mainly regulated by cerebellar inputs, is overactive during sensory-cued movement in PD patients, probably to compensate for the impaired SMA function.15 The basal ganglia–SMA system regulating elaboration of internally driven motor sequence34,35 might function to adequately scale the motor activity in normal noncued gait. However, PD fails in this process because of the basal ganglia–SMA dysfunction. Conversely, PD patients scale the motor activity relatively well when guided by visual cues.36,37 Because stride length is externally determined in TL but not in PL, the PMC, functionally better preserved than the SMA in PD patients, may take over the optimal scaling of the motor activity when cued by the task-relevant sensory information. Owing to the task relevancy, visual information in TL would be actively processed in the PPC that provides rich visual information to the PMC and cerebellum.38 – 40 In addition, the visual stimuli that can elicit paradoxical movement also effectively activate the pathway conveying visual information to the cerebellum,16 which in turn strongly influences the PMC. The DLPFC activation in TL 2 PL, possibly related to the spatial working memory function, was greater in the controls than in PD patients. It is interesting that although DLPFC underactivation in PD patients was reported previously,32 this abnormality remained despite the development of paradoxical gait. An important question remains unsolved concerning paradoxical movement elicited by emotional or attentional stimuli, although attention-induced paradoxical movement may partially share the underlying mechanisms with visually driven paradoxical movement.41 Of interest is what are shared and what are not. Possibly, visual cues also help patients concentrate on gait regulation. In this sense, as shown in TL 2 PL in PD patients, activation of the anterior cingulate cortex, which links emotion and behavior, could provide a

hint.42 These hypotheses should be validated in future studies using an emotional or attentional paradigm. 15. This study was supported in part by Grants-in-Aid for Scientific Research (A) 09308031, (A) 08558083, and (C) 08670705, and on Priority Areas 0827916 from the Japan Ministry of Education, Science, Sports and Culture, Research for the Future Program JSPSRFTF97L00201 from Japan Society for the Promotion of Science, and General Research Grants for Aging and Health “Analysis of aged brain function with neuroimaging techniques” and “Physiological parameters for evaluation of aging of brain” from the Japan Ministry of Health and Welfare.

16. 17.

18.

We thank the staffs of Department of Radiology, Kyoto University Hospital, especially Mr T. Fujita and Mr H. Kitano, for their technical help. 19.

References 1. Souques MA. Les syndromes parkinsoniens. Rev Neurol (Paris) 1921;1:543–573 2. Barbeau A. The clinical physiology of side effects in long-term L-DOPA therapy. Adv Neurol 1974;5:347–365 3. Martin JP. Locomotion and the basal ganglia. In: Martin JP, ed. The basal ganglia and posture. London: Pitman Medical, 1967:20 –35 4. Bagley S, Kelly B, Tunnicliffe N, et al. The effect of visual cues on the gait of independently mobile Parkinson’s disease patients. Physiotherapy 1991;77:415– 420 5. Forssberg H, Johnels B, Steg G. Is parkinsonian gait caused by a regression to an immature walking pattern? Adv Neurol 1984; 40:375–379 6. Dunne JW, Hankey GJ, Edis RH. Parkinsonism: upturned walking stick as an aid to locomotion. Arch Phys Med Rehabil 1987;68:380 –381 7. Oku N, Matsumoto M, Hashikawa K, et al. Carbon dioxide reactivity by consecutive technetium-99m-HMPAO SPECT in patients with a chronically obstructed major cerebral artery. J Nucl Med 1994;35:32– 40 8. Pantano P, Piero VD, Ricci M, et al. Motor stimulation response by technetium-99m-hexamethylpropylene amine oxime split-dose method and single photon emission tomography. Eur J Nucl Med 1992;19:939 –945 9. Marshall RS, Lazar RM, Heertum RLV, et al. Changes in regional cerebral blood flow related to line bisection discrimination and visual attention using HMPAO-SPECT. Neuroimage 1997;6:139 –144 10. Andersen AR, Friberg HH, Schmidt JF, Hasselbalch SG. Quantitative measurements of cerebral blood flow using SPECT and [99mTc]-d,l-HM-PAO compared to xenon-133. J Cereb Blood Flow Metab 1988;8(Suppl):S69 –S81 11. Fukuyama H, Matsuzaki S, Ouchi Y, et al. Neural control of micturition in man examined with single photon emission computed tomography using 99mTc-HMPAO. Neuroreport 1996;7: 3009 –3012 12. Fukuyama H, Ouchi Y, Matsuzaki S, et al. Brain functional activity during gait in normal subjects: a SPECT study. Neurosci Lett 1997;228:183–186 13. Mushiake H, Inase M, Tanji J. Neuronal activity in the primate premotor, supplementary, and precentral motor cortex during visually guided and internally determined sequential movements. J Neurophysiol 1991;66:705–718 14. Halsband U, Matsuzaka Y, Tanji J. Neuronal activity in the

20. 21. 22.

23.

24.

25.

26.

27.

28.

29. 30. 31.

32.

33.

34.

35.

primate supplementary, pre-supplementary and premotor cortex during externally and internally instructed sequential movements. Neurosci Res 1994;20:149 –155 Samuel M, Ceballos-Baumann AO, Blin J, et al. Evidence for lateral premotor and parietal overactivity in Parkinson’s disease during sequential and bimanual movements: a PET study. Brain 1997;120:963–976 Glickstein M, Stein J. Paradoxical movement in Parkinson’s disease. Trends Neurosci 1991;14:480 – 482 Hughes AJ, Daniel SE, Kliford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992; 55:181–184 Fahn S, Elton R, Members of the UPDRS Development Committee. Unified Parkinson’s disease rating scale. In: Fahn SMC, Calne DB, Goldstein M, eds. Recent developments in Parkinson’s disease. Florham Park, NJ: Macmillan Health Care Information, 1987:153–163 Friston KJ, Holmes AP, Worsley KJ, et al. Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Map 1995;2:189 –210 Friston KJ, Ashburner J, Frith CD, et al. Spatial registration and normalization of images. Hum Brain Map 1995;2:165–189 Talairach J, Tournoux P. Co-planar stereotaxic atlas of the human brain. New York: Thieme Medical, 1988 Wall JC, Turnbull GI. The kinematics of gait. In: Turnbull GI, ed. Physical therapy management of Parkinson’s disease. New York: Churchill Livingstone, 1992:49 – 67 Morris ME, Iansek R, Matyas TA, Summers JJ. The pathogenesis of gait hypokinesia in Parkinson’s disease. Brain 1994;117: 1169 –1181 Hanakawa T, Fukuyama H, Katsumi Y, et al. Cerebral blood flow activation study of gait disturbance in Parkinson’s disease. Neurology 1998;50(Suppl 4):A391 (Abstract) Preuss TM, Stepniewska I, Kaas JH. Movement representation in the dorsal and ventral premotor areas of owl monkeys: a microstimulation study. J Comp Neurol 1996;371:649 – 676 McIntosh GC, Brown SH, Rice RR, Thaut MH. Rhythmic auditory-motor facilitation of gait patterns in patients with Parkinson’s disease. J Neurol Neurosurg Psychiatry 1997;62: 22–26 Berardelli A, Dick JP, Rothwell JC, et al. Scaling of the size of the first agonist EMG burst during rapid wrist movements in patients with Parkinson’s disease. J Neurol Neurosurg Psychiatry 1986;49:1273–1279 Kuypers HG, Lawrence DG. Cortical projections to the red nucleus and the brain stem in the rhesus monkey. Brain Res 1967; 4:151–188 Freund HJ, Hummelsheim H. Lesions of premotor cortex in man. Brain 1985;108:697–733 DeLong MR. Primate models of movement disorders of basal ganglia origin. Trends Neurosci 1990;13:281–285 Georgiou N, Iansek R, Bradshaw JL, et al. An evaluation of the role of internal cues in the pathogenesis of parkinsonian hypokinesia. Brain 1993;116:1575–1587 Playford ED, Jenkins IH, Passingham RE, et al. Impaired mesial frontal and putamen activation in Parkinson’s disease: a positron emission tomography study. Ann Neurol 1992;32: 151–161 Samuel M, Ceballos-Baumann AO, Turjanski N, et al. Pallidotomy in Parkinson’s disease increases supplementary motor area and prefrontal activation during performance of volitional movements: an H215O PET study. Brain 1997;120:1301–1313 Mushiake H, Inase M, Tanji J. Selective coding of motor sequence in the supplementary motor area of the monkey cerebral cortex. Exp Brain Res 1990;82:208 –210 Deiber MP, Passingham RE, Colebatch JG, et al. Cortical areas

Hanakawa et al: Paradoxical Gait in Parkinson’s Disease

335

and the selection of movement: a study with positron emission tomography. Exp Brain Res 1991;84:393– 402 36. Flowers K. Lack of prediction in the motor behaviour of parkinsonism. Brain 1978;101:35–52 37. Jackson SR, Jackson GM, Harrison J, et al. The internal control of action and Parkinson’s disease: a kinematic analysis of visually-guided and memory guided prehension movements. Exp Brain Res 1995;105:147–162 38. Wise SP, Boussaoud D, Johnson PB, Caminiti R. Premotor and parietal cortex: corticocortical connectivity and combinatorial computations. Annu Rev Neurosci 1997;20:25– 42

336

Annals of Neurology

Vol 45

No 3

March 1999

39. Caminiti R, Ferraina S, Johnson PB. The sources of visual information to the primate frontal lobe: a novel role for the superior parietal lobe. Cereb Cortex 1996;6:319 –328 40. Stein JF, Glickstein M. Role of the cerebellum in visual guidance of movement. Physiol Rev 1992;72:967–1017 41. Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson’s disease. Normalization strategies and underlying mechanisms. Brain 1996;119:551–568 42. Devinsky O, Morrell MJ, Vogt BA. Contribution of anterior cingulate cortex to behaviour. Brain 1995;118:279 –306