Motor Control & Sensory Integration

46 R. lansek et al. SMA I Prepare Prepare Prepare Prepare BG Mvmt X Mvmt X+ 1 Mvmt X+2 Mvmt X+3 J J J c., cue cue cue MC Execute Execute Execute Execute Execute M v m t X-] Mvmt X Mvmt X+I Mvmt X+2 vmt X+3 SMA = Supplementary motor area BG = Basal Ganglia MC = Motor Cortex Figure2. Automatictemporalsequencingof movementsby BG-SMAinteraction. It is evident from the Brotchie et al. (1991a, b, c) studies that the basal ganglia are not involved in the initiation of the movement sequence, as internal cues are not generated until the sequence has become established. However, the GP does generate premovement sustained activity for whole movement sequences, as does the SMA (Schultz & Romo, 1992). The premovement sustained activity generated in the GP may contribute to the same activity generated in the SMA in order to maintain the preparedness of whole movement sequences until execution. If this contribution is deficient then the whole movement sequence may not be properly initiated. The SMA has all the neural requirements to initiate a movement sequence (set and phasic activity) if appropriately prepared. In summary it is suggested that, at a neural level, the basal ganglia functions predominantly as the provider of an internal cue, represented as phasic neural activity. Effectively this cue acts to release the next submovement and in turn its execution results in the generation of another cue. The release of the upcoming movement is dependent upon the abrupt cessation of tonic SMA preparatory activity. Another function of the basal ganglia is to contribute to cortical motor preparedness for whole movement sequences. This function is represented in the GP as premovement sustained activity. It is evident that the single cell studies and cerebral blood flow studies are in agreement, that the basal ganglia and the SMA work in unison to run movement sequences which are well learnt and predictable. We next describe how this hypothesis

The B G and SMA in the Elaboration of Movement 47 of basal ganglia-SMA interaction helps to explain hypokinesia and akinesia in Parkinson's disease. 5. PARKINSON'S DISEASE Parkinson's disease (PD) is the most common disease affecting the basal ganglia in humans. As a consequence, patients With PD have been investigated extensively in an attempt to better understand basal ganglia function and the underlying mechanisms of hypokinesia. Parkinson's disease is characterized by loss of dopamine producing neurones which project from the substantia nigra to the striatum (Calabresi, Mercuri, Sancesario & Bernardi, 1993). As a consequence the level of dopamine is reduced, particularly in the putamen (Gingrich & Caron, 1993). The clinical consequence of this loss is reduced movement (hypokinesia) and difficulty in initiating movement (akinesia) (Hallet & Khoshbin, 1980). Isolated ballistic movements are usually undershot and slower in velocity in PD subjects compared with controls (Hallett & Khoshbin, 1980). Hypokinesia is evident for single, isolated movements, but is more apparent for movement sequences. Studies show that not only are the individual movements slower, but it is particularly the case that the time to switch from one submovement to the next in sequences is prolonged (Benecke, R0thwell, Dick, Day & Marsden, 1986; Benecke, Rothwell, Dick, Day & Marsden, 1987). Motor instability is another phenomenon clinically apparent. This is characterized by incremental slowing in successive submovements performed down a sequence. The slowing of movements mainly affects sequences which are performed in an automatic fashion, and which require no attentional resources. This phenomenon is well demonstrated clinically by the inability of PD subjects to perform simultaneous motor tasks (see Talland & Schwab, 1964). In PD movements which are consciously performed 'utilizing attentional resources' are usually performed normally and in preference to movements requiting less attention. In simultaneous tasks, the more automatic movement may not be performed. Clinically Parkinsonian subjects tend to direct their attention to first one movement and then to the other in an alternating fashion. Such observations are in agreement with the anatomical and cerebral blood flow data which, as we saw above, demonstrated that the basal ganglia and the SMA co-

48 R. lansek et al. operate in the running of well learnt movement sequences (for example, Garnett, Nahmias & Fimau, 1984; Rascol et al., 1992). A number of studies have found that motor performance in PD subjects is helped by the provision of external cues. It has been suggested that PD subjects control movements in an open loop manner and use visual information to guide movements (Flowers, 1976). The dependency of PD subjects' movements on external information, particularly visual, has been confirmed by numerous studies (for example, Day, Dick & Marsden, 1984; Jones, Phillips, Bradshaw, Iansek & Bradshaw, 1992; Stem, 1983). However, the studies of Georgiou et al. (1993, 1994) are particularly relevant. Georgiou et al. (1993) examined a 10 button press sequence, delineated along two rows of 10 buttons, which PD subjects had to learn and perform. During the learning process external cues were provided by lights at the base of each button. The response board used in this task allowed measurement of inter-button preparation time, expressed as the time each button was depressed prior to movement to the next button, as well as inter-button movement time. In this way it was possible to examine the effects of external cues on both the preparation and execution of movement. Once the PD subjects had learnt the task, the cue lights were extinguished and subjects had to perform the same movement sequence without visual cues, thereby nmning the sequence from motor memory utilizing their own internal cues to move from one button to the next. Electrophysiological data has suggested that both the preparation for submovements as well as the submovements themselves should be affected, given that the influence of the basal ganglia cue is on the preparatory activity in the SMA for the upcoming movement. Georgiou et al. (1993) found that both the preparation thne and movement time became significantly prolonged when subjects had to perform the sequence without external cues. Substitute auditory cues were subsequently given contingent upon the subjects' motor performance (on button press, or on button release). These cues significantly improved motor performance, but not as dramatically as regular non-contingent auditory cues delivered by a metronome. The latter cues were given at a frequency (4.8Hz) derived from control subjects inter-button movement speed. These findings confu'med that submovement preparation and execution are intimately dependant upon an appropriate cue from the basal ganglia. However, the manner by which disturbed preparation for movement results in abnormal movement execution remains unclear. One possibility is that alternate motor

The BG and SMA in the Elaboration of Movement 49 executive mechanisms are utilized in PD since access to normal executive procedures is impaired due to the disruptions of preparatory process. Consequently the movements observed in PD possibly stem from compensatory mechanisms and may not necessarily be a direct consequence of basal ganglia malfunction. Indeed, Sheridan, Flowers and Hurrell (1987) proposed that movement end point inaccuracy may be the basic deficit in movement disturbance in PD. They suggested that PD subjects could produce single movements at the same speed as controls, but they did so at the expense of end point accuracy. Fast movements became inaccurate. However, Sheridan et al. (1987) used single movements which are less likely to be influenced by disturbed basal ganglia cues. Martin et al. (1994) addressed the same question, but they used a movement sequence which was both extemaUy and internally cued. In addition, to localize the nature of the disruption, they examined movement kinematics. They demonstrated that PD subjects could perform the whole sequence at the same speed as controls; however, submovement inaccuracy was greater and cumulative for each submovement down the sequence, confirming that end point inaccuracy may be the direct consequence of impaired preparation for the submovements. Kinematic analysis of the submovements demonstrated that PD subjects had difficulty in generating an appropriate burst of acceleration, especially when required to perform at the non-preferred fast speed, and continued to make accelerative attempts for the major part of the movement time. This left little time to control deceleration and thus resulted in end point inaccuracy. When the PD subjects performed the sequence at a slower, preferred speed, accuracy improved (Martin et al., 1994). In the absence of visual cues it was observed that the submovements cumulatively increased in size, velocity and overshoot of the target when the sequence was performed at a faster speed than preferred speed (Martin et al., 1994). A similar motor instability to hypokinesia was observed but now in the opposite direction, presumably due to the altered motor set in which the movement was performed. We have termed this phenomenon tachykinesia and likened it to festination. Unfortunately these studies used movements and movement sequences which required a high degree of attentional resources and were therefore less likely to be affected by basal ganglia malfunction. It is not easy to separate the requirement for faster speed in movement from the attentional requirements needed in attaining that speed. These movements certainly could not have been performed at a subconscious level. These studies do however illustrate that a

50 R. lansek et al. disturbed or deficient cue produces motor instability for fast movements as well as for preferred slow movements in PD. The instability at a fast speed is perhaps functionally more disabling, leading to a preference for slower responding. Two further points need to be considered. The first concerns whether the basal ganglia interaction with the SMA, at a neural level, occurs at the time when the SMA preparatory activity falls. In externally cued sequences this is indeed the time the external cue occurs. It would be expected that an internal cue should function similarly. Romo and Schultz (1992) showed abrupt termination of sustained activity in SMA neurones upon movement onset in monkeys. Similarly, Watts and Mandir (1992) showed sharp termination of SMA activity at movement onset, but a more prolonged peak of SMA activity in MPTP (Parkinsonian) monkeys. This would suggest the absence of an internal cue from the basal ganglia, resulting in the failure to terminate sustained pre- movement activity in the SMA. Cunnington et al. (in press) have examined this interaction in normal and PD subjects. They recorded the premovement potential from the scalp by computer averaging of the EEG. They used the same response board and the same 10 button-press sequence as was used in the study of Georgiou et al. (1993). Button release was used as a trigger for back averaging the EEG. Subjects were trained to perform the 10 button press sequence by use of external light cues embedded into the buttons. The whole sequence would be illuminated and every 4 seconds each successive light would extinguish, the subject would have to move the index finger as fast as possible to the next button, and then wait for the next cue to move. The same process was performed without cues, and the subjects had to time the 4 seconds internally prior to each movement. Cunnington et al. (in press) found that for externally cued movements control subjects generated a premovement potential, but PD subjects did not generate any preparatory potential. Parkinsonian subjects did generate a premovement potential for internally cued movements, though the potential started later than controls, and had both a prolonged peak and slower decline from the peak than that of control subjects. Movement time from button to button was no different in the two groups, thereby excluding slower movement as an explanation for the slower fall from peak. This would suggest prolonged SMA peak activity in PD subjects compared with controls. These findings were consistent with the hypothesis of Brotchie et al. (1991b) which suggested that the phasic activity from the GP was used to abruptly terminate set related activity in the SMA. It is this abrupt fall which appears to be responsible for an

The BG and SMA in the Elaboration of Movement 51 appropriately executed movement to follow. A slower fall, as shown in the Cunnington et al. (in press) study would explain an inappropriately executed movement. The lack of a premovement potential for externally cued movement in PD subjects suggests that an alternate cortical mechanism must be utilized other than the SMA. In keeping with electrophysiological studies (for example, Diebner et al., 1991; Mushiake et al., 1991; Weinrich & Wise, 1982) the PMA seems to be the most likely cortical area which is activated by external cues in the organization of such movements in PD subjects. The second point relates to the mechanism by which the premovement activity translates to movement execution, presumably through the MSC. The SMA has direct projections to the spinal cord as well as to the MSC and thus its influence on movement may not be necessarily via the MSC. However cerebral blood flow and electrophysiological studies do suggest that the two cortical regions work together in movement performance. Unfortunately, no data exit which explain how this interaction occurs at a neural level. We do have indirect data, however, which confirm that the SMA preparatory activity influences movement execution. The studies of Georgiou et al. (1993) demonstrated that the disturbed preparation in PD, which presumably occurred as a result of defective cue production by the basal ganglia, was accompanied by abnormal movement execution. The study of Martin et al. (1994) demonstrated that movement execution was disrupted because movement acceleration was impaired. Georgiou et al. (1993) also demonstrated that PD subjects were unable to rescale movement amplitude and movement velocity if this was required as part of the execution of a learnt movement sequence. Subjects had to perform slightly longer submovements for diagonal as compared to linear button movements. In these circumstances both the button down time and the inter-button movement time were prolonged when the movements were performed without external cues. External auditory cues, delivered in a regular non contingent manner, restored both the down time and movement time to normal. These data confirm the close relationship between movement preparation and movement execution, and demonstrate how disturbance of preparation can lead to abnormal execution of the prepared movement. However it is still not completely clear what aspects of movement execution are prepared by the SMA. Some have suggested that SMA preparatory activity determines the timing of the upcoming movement (Halsband et al., 1993; Komhuber, Deeke, Lang, Lang & Komhuber, 1989; Lang, Obrig, Lindinger, Cheyne & Deeke, 1990) and others

52 R. lansek et al. have suggested that it is concerned with programming the upcoming movement (Goldberg, 1985; Orgogozo & Larsen, 1979; Roland et al., 1980b). Cunnington et al. (in press) have looked at this question by comparing the Bereitschaftspotential (BP) in normal subjects using a ten-button-press sequence for four different conditions. The button-press sequence was either spatially or temporally predictable and unpredictable. These conditions were made possible by using a predictable pathway which was lit throughout the button press sequence and an unlit pathway whose spatial characteristics were unpredictable; the next button would light only when the previous button had been pressed for a certain time period. Similarly, a predictable time of 4 seconds was used between each button press, and an unpredictable time of between 4 and 6 seconds between button presses. In this way it was possible to vary the spatial and temporal parameters to examine which of them influenced the development of the BP. Cunnington et al. (in press) found that the BP correlated with the predictable timing of the upcoming movements and not the spatial characteristics of the movement. These findings suggested that the SMA is concerned with the correct 6rning of predictable movements, presumably releasing a packaged movement for the motor cortex to execute. The release needs to be precise for the movement to be executed normally. In PD subjects, the BP peak is prolonged and the fall from peak is slow (Cunnington et al., in press), with the resultant impact on sub movement acceleration, speed and amplitude. This is the precise temporal locus where Brotchie et al. (1991a, b) suggested that the basal ganglia cue would influence SMA preparatory activity and enable a rapid decline from peak activity. The findings from Cunnington et al. (in press) support this concept. 6. SUMMARY The evidence presented in this review suggests that the basal ganglia influence movement via the motor cortical regions - in particular the SMA and PMA. This review has concentrated on the SMA-basal ganglia interaction. This interaction is concerned with the nmning of movement sequences which are well learnt and predictable. In this regard it refers to skilled movements, such as writing, knitting, playing the piano, tennis or football. It does not refer to novel or unskilled movements. Alternative motor mechanisms are involved in these circumstances. In executing skilled movement sequences evidence has been presented that the SMA prepares for each submovement in

The BG and SMA in the Elaboration of Movement 53 the sequence. It does so by generating premovement sustained activity. The basal ganglia generate a cue which is represented at a neural level as phasic neural activity. This cue occurs towards the end of each movement, and terminates premovement sustained activity in the SMA. It is this rapid termination which releases the movement ready for execution. In this regard the SMA-basal ganglia interaction is concerned with the correct timing of submovements strung together in the sequence. The initiation of the sequence is performed by the SMA and does not involve the basal ganglia until the sequence is established. Once between two and three submovements have been performed and the sequence is established, then the basal ganglia becomes involved in providing internal motor cues. The other function of the SMA and PMA is to provide preparatory activity for whole movement sequences. The basal ganglia contribute to this preparatory activity and at a neural level this is represented by sustained premovement activity. This contribution enables the basal ganglia to be involved in the initiation of movement sequences as well as to contribute to the running of the sequences themselves. In Parkinson's disease it was hypothesized that the internal motor cue and the preparatory activity are defective. Such abnormalities manifest in the execution of movement sequences and movement initiation. The abnormal cue results in disturbed preparation for submovements, with slower fall of SMA preparatory activity and the consequent release of an abnormal submovement. The submovement is usually slower and of reduced amplitude. This deficit is typically additive down the sequence, resulting in motor instability. Sequences are usually performed in a slow motor set, as the instability in a fast motor set becomes functionally incapacitating for PD subjects. The defect in basal ganglia preparatory activity also results in defective preparation for whole movement sequences. As a result the sequences themselves cannot be released appropriately and the initiation, which is normally performed at a cortical level, is defective. However, the mechanism by which the SMA interacts with the motor cortex is not yet clear, particularly how movement preparation influences the execution of forthcoming movements. REFERENCES Alexander, G.E., & Crutcher, M.D. (1990). Functional architecture of basal ganglia circuits: Neural subitutes of parallel processing. TINS, 13, 266-271.

54 R. lansek et al. Benecke, R., Rothwell, J.C., Dick, J.P.R., Day, B.L., & Marsden, C.D. (1986). Performance of simultaneous movements in patients with Parkinson's disease. Brain, 109, 739-757. Benecke, R., Rothwell, J.C., Dick, J.P.R., Day, B.L., & Marsden, C.D. (1987). Disturbance of sequential movements in patients with Parkinson's disease. Brain, 110, 361-379. Brotchie, P., Iansek, R., & Home, M.K. (199 la). Motor function of the monkey globus pallidus: Neuronal discharge and parameters of movement. Brain, 114, 1667-1683. Brotchie, P., Iansek, R., & Home, M.K. (1991b). Motor function of the monkey globus pallidus: Cognitive aspects of movement and phasic neuronal activity. Brain, 114, 1685-1702. Brotchie, P., Iansek, R., & Home, M.K. (1991c). A neuronal network model of discharge patterns in the monkey globus pallidus. Neuroscience Letters, 131, 33-36. Calabresi, P., Mercuri, N.B., Sancesario, G., & Bemardi, G. (1993). Electrophysiology of dopamine-denervated striatal neurons. Brain, 116, 433-452. Carpenter, M.B., & Peter, P. (1972). Nigrostriatal and nigrothalamic fibres in the Rhesus monkey, Journal of Comparative Neurology, 144, 93-116. Chevalier, G., & Deniau, J.M. (1990). Disinhibition as a basic process in the expression of striatal functions. TINS, 1, 277-280. Conde, H. (1992). Organization and physiology of the substantia nigra. Experimental Brain Research, 88, 233-248. Cunnington, R., Iansek, R., Bradshaw, J.L., & Phillips, J.G. (in press). Movement- related potentials in Parkinson's disease: Presence and predictability of temporal and spatial cues. Brain. Day, B.L., Dick, J.P.R., & Marsden, C.D. (1984). Patients with Parkinson's disease can employ a predictive motor strategy. Journal of Neurology, Neurosurgery and Psychiatry, 47, 1299-1309. Deiber, M.P., Passingham, R.E., Colebatch, J.G., Friston, K.J., Nixon, P.D., & Frackowiak, R.S.J. (1991). Cortical areas and the selection of movement: a study with positron emission tomography. Experimental Brain Research, 84, 393-402. DeLong, M.R. (1971). Activity of pallidal neurons during movement. Journal of Neurophysiology, 34, 414-427.

The BG and SMA in the Elaborationof Movement 55 DeLong, M.R. (1990). Primate models of movement disorders of basal ganglia origin. Trends in Neurosciences, 13, 281-285. Flowers, K. (1976). Visual 'closed-loop' and 'open-loop' characteristics of voluntary movement in patients with Parkinsonism and intention tremor. Brain, 99, 269-310. Garnett, E.S., Nahmias, C., & Fimau, G. (1984). Central dopaminergic pathways in hemiparkinsonism examined by Positron Emission Tomography. Canadian Journal of Neurological Science, 11, 174-179. Georgiou, N., Iansek, R., Bradshaw, J.L., Phillips, J.G., Mattingley, J.B. & Bradshaw, J.A. (1993). An evaluation of the role of internal cues in the pathogenesis of Parkinsonian hypokinesia. Brain, 116, 1575-1587. Georgiou, N., Bradshaw, J., Iansek, R., Phillips, J., Mattingley, J.B., & Bradshaw, J.A. (1994). Reduction in external cues and movement sequencing in Parkinson's disease. Journal of Neurology, Neurosurgery and Psychiatry, 57, 368-370. Georgopoulos, A.P., DeLong, M.R., & Crutcher, M.D. (1983). Relation between parameters of step-tracking and movements and single cell discharge in the globus pallidus and subthalamic nucleus of the behaving monkey. Journal of Neuroscience, 3, 1586-1598. Gingrich, J.A., & Caron, M.G. (1993). Recent advances in the molecular biology of dopamine receptors. Annual Review of Neuroscience, 16, 299-321. Goldberg, G. (1985). Supplementary motor area structure and function: review and hypotheses. Behavioral and Brain Sciences, 8, 567-615. Hallett, M., & Khoshbin, S. (1980). A physiological mechanism of bradykinesia. Brain, 103, 301-314. Halsband, U., Ito, N., Tanji, J., & Freund, H.J. (1993). The role of premotor cortex and the supplementary motor area in the temporal control of movement in man. Brain, 116, 243-266. Hartmann-von Monakow, K., Akert, K., & Kunzle, H. (1978). Projections of the precentral motor cortex and other cortical areas of the frontal lobe to the subthalamic nucleus in the monkey. Experimental Brain Research, 33, 395-403. Hazrati, L.N., Parent, A., Mitchell, S., & Haber, S.N. (1990). Evidence for intemonnections between the two segments of the globus pallidus in primates: A PHA-L anterograde tracing study. Brain Research, 533, 171-175.

56 R. lansek et al. Hoover, J.E., & Strick, P.L. (1993). Multiple output channels in the basal ganglia. Science, 259, 819-821. Ikeda, A., Luders, H.O., Burgess, R.C., & Shibasaki, H. (1992). Movement-related potentials recorded from supplementary motor area and primary motor area. Brain, 115, 1017-1043. Jenkins, I.H., Ferandez, W., Playford, E.D., Lees, A.J., Frackowiak, R.S.J., & Passingham, R.E. (1992). Impaired activation of the supplementary motor area in Parkinson's disease is reversed when akinesia is treated with Apomorphine. Annals of Neurology, 32, 749-757. Jones, D.L., Phillips, J.G., Bradshaw, J.L., Iansek, R., & Bradshaw, J.A. (1992). Programming of single movements in Parkinson's disease:comparison with Huntington's disease. Journal of Clinical and Experimental Neuropsychology, 14, 762-772. Jones, E.G., Coulter, J.D., Burton, H., & Porter, R. (1977). Cells of origin and terminal distribution of corticostriatal fibres arising in the sensory motor cortex of monkeys. Journal of Comparative Neurology, 173, 53-80. Kemp, J.M., & Powell, J.S. (1970). The cortico-striatal projection in the monkey. Brain, 93, 525-546. Kim, R., Nakano, K., Jayaraman, A., & Carpenter, M.B. (1976). Projections of the globus pallidus and adjacent structures. Journal of Comparative Neurology, 169, 263-290. Kimura, M., Aosaki, T., Hu, Y., & Ishida, A. (1992). Activity of primate putamen neurones is selective to the mode of voluntary movement: Visually guided, self- initiated or memory guided. Experimental Brain Research, 89, 473-477. Komhuber, H.H., Deeke, L., Lang, W., Lang, M., & Komhuber, A. (1989). Will, volitional action, attention and cerebral potentials in man: Bereitschaftspotential, performance-related potentials, directed attention potential, EEG spectrum changes. In W.A. Hershberger. Volitional action, Advances in Psychology, 62 (pp. 107-168), Amsterdam: Elsevier. Lang, W., Obrig, H., Lindinger, G., Cheyne, D., & Deeke, L. (1990). Supplementary motor areaactivation while tapping bimanually different rhythms in musicians. Experimental Brain Research, 79, 504-514.

The BG and SMA in the Elaboration of Movement 57 Manetto, C., & Lidsky, T.I. (1989). The effects of movements on caudate sensory responses. Neuroscience Letters, 96, 295-299. Martin, K.E., Phillips, J.G., Bradshaw, J.L., & Iansek, R. (1994). Could bradykinesia in Parkinson's disease simply be compensation? Journal of Neurology, 241,439-447. Matsumura, M., Kojima, J., Gardiner, T.W., & Hikosaka, O. (1992). Visual and oculomotor functions of monkey subthalamic nucleus. Journal of Neurophysiology, 67, 1615-1632. Mink, J.W., & Thach, W.T. (1987). Preferential relation of pallidal neurons to ballistic movements. Brain Research, 417, 393-398. Mushiake, H., lnase, M., & Tanji, J. (1990). Selective coding of motor sequence in the supplementary motor area of the monkey cerebral cortex. Experimental Brain Research, 82, 208-210. Mushiake, H., Inase, M., & Tanji, J. (1991). Neuronal activity in the primate premotor, supplementary, and precentral motor cortex during visually guided and internally determined sequential movements. Journal of Neurophysiology, 66, 705-718. Orgogozo, J.M., & Larsen, B. (1979). Activation of the supplementary motor area during voluntary movement in man suggests it works as a supramotor area. Science, 206, 847-850. Parent, A., & Hazrati, L.N. (1993). Anatomical aspects of information processing in primate basal ganglia. TINS, 16, 111-116. Phillips, J.G., Bradshaw, J.L., Iansek, R., & Chiu, E. (1993). Motor functions of the basal ganglia. Psychological Research, 55, 175-181. Rascol, O., Sabatini, U., ChoUet, F., Celsis, P., Fabre, N., Senard, J.M., & Montastruc, J.L. (1992). Supplementary and primary sensory motor area activity in Parkinson's disease. Archives of Neurology, 49, 144-148. Richter, R. (1945). Degeneration of the basal ganglia in monkeys from chronic carbon disulfide poisoning. Journal of Neuropathology and Experimental Neurology, 4, 324-353. Roland, P.E., Skinhoj, N.A., Lassen, N.A., & Larsen, B. (1980a). Different cortical areas in man in organization of voluntary movements in extrapersonal space. Journal of Neurophysiology, 43, 137-150.

58 R. lansek et al. Roland, P.E., Larsen, B., Lassen, N.A., & Skinhoj, E. (1980b). Supplementary motor area and other cortical areas in organization of voluntary movements in man. Journal of Neurophysiology, 43, 118-136. Roland, P.E., Meyer, E., Shibasaki, T., Yamamoto, Y.L., & Thompson, C.J. (1982). Regional cerebral blood flow changes in cortex and basal ganglia during voluntary movements in normal human volunteers. Journal of Neurophysiology, 48, 467-480. Romo, R., & Schultz, W. (1992). Role of the primate basal ganglia and frontal cortex in the internal generation of movements. 111. Neuronal activity in the supplementary motor area. Experimental Brain Research, 91, 396-407. Sadikot, A.F., Parent, A., Smith, Y., & Bolam, J.P. (1992). Efferent connections of the centromedian and parafascicular thalamic nuclei in the squirrel monkey: A fight and electron microscope study of the thalamostriatal projection in relation to striatal heterogeneity. Journal of Comparative Neurology, 320, 228-242. Schell, G.R., & Strick, P.L. (1984). The origin of thalamic inputs to the arcuate premotor and supplementary motor areas. Journal of Neuroscience, 4, 539-560. Schultz, W., & Romo, R. (1992). Role of the primate basal ganglia and frontal cortex in the internal regulation of movements:l 1. Movement-related activity in the anterior striatum. Experimental Brain Research, 91, 385-395. Seitz, R.J., & Roland, P.E. (1992). Learning of sequential finger movements in man: A combined kinematic and positron emission tomography study. European Journal of Neuroscience, 4, 154-165. Shibasaki, H., Sadato, N., Lyshkow, H., Yonekura, Y., Honda, M., & Nagamine, T. (1993). Both primary motor cortex and supplementary motor area play an important role in complex finger movement. Brain, 116, 1387-1398. Sheridan, M.R., Flowers, K.A., & HurreU, J. (1987). Programming and execution of movement in Parkinson's disease. Brain, 110, 1247-1271. Stem, Y., Mayeux, R., Rosen, J., & Ilson, J. (1983). Perceptual motor dysfunction in Parkinson's disease: a deficit in sequential and predictive voluntary movement. Journal of Neurology, Neurosurgery and Psychiatry, 46, 145-151. Talland, G.A., & Schwab, R.S. (1964). Performance with multiple sets in Parkinson's disease. Neuropsychologica, 2, 45-53. Tanji, J., & Kurata, K. (1982). Comparison of movement-related activity in two cortical motor areas of primates. Journal of Neurophysiology, 48, 633-653.

The BG and SMA in the Elaboration of Movement 59 Tanji, J., & Kurata, K. (1985). Contrasting neuronal activity in supplementary and precentral motor cortex of monkeys. 1. Responses to instructions determining motor responses to forthcoming signals of different modalities. Journal of Neurophysiology, 53, 129-141. Tokuno, H., Kimura, M., & Tanji, J. (1992). Pallidal inputs to thalamocortical neurons projecting to the supplementary motor area: an anterograde and retrograde double labeling study in the macaque monkey. Experimental Brain Research, 90, 635-638. Watts, R.L., & Mandir, A.S. (1992). The role of motor cortex in the pathophysiology of voluntary movement deficits associated with Parkinsonism. Neurologic Clinics, 10, 451-469. Weinrich, M., & Wise, S.P. (1982). The premotor cortex of the monkey. Journal of Neuroscience, 2, 1392-1345. Weissendanger, R., & Weisendanger, M. (1985). The thalamic connections with medial area 6 (supplementary motor cortex) in the monkey. Experimental Brain Research, 59, 91-104. Wichmann, T., & DeLong, M.R. (1993). Pathophysiology of Parkinsonian motor abnormalities. Advances in Neurology, 60, 53-61. Wilson (1912). Progressive lenticular degeneration: A familial nervous disease associated with cirrhosis of the fiver. Brain, 34, 295-309. Yeterian, E.H., & Van Hoesen, G.W. (1978). Cortico-striate projections in the Rheus monkey: the organisation of certain cortico-caudate connection. Brain Research, 139, 43-63. Yoshida, S., Nambu, A., & Jinnai, K., (1993). The distribution of the globus pallidus neurons with input from vartious cortical areas in the monkeys. Brain Research, 611, 170-174.

Motor Control and Sensory Motor Integration: Issues and Directions 61 D.J. Gleneross and J.P. Piek (Editors) 9 1995 Elsevier Science B.V. All fights reserved. Chapter 4 MOTOR CONTROL CONSIDERATIONS FOR THE REHABILITATION OF GAIT IN PARKINSON'S DISEASE M.E. Morris Geriatric Research Unit, Kingston Centre, Warrigal Rd, Cheltenham, 3192, Australia. Schools of Physiotherapy and Behavioural Health Sciences, La Trobe University, Bundoora, 3083, Australia. R. lansek, Geriatric Research Unit, Kingston Centre, Warrigal Rd, Cheltenham, 3192, Australia. JJ. Summers, Department of Psychology, University of Southern Queensland, Toowoomba, 4350, Australia. T.A. Matyas School of Behavioural Health Sciences, La Trobe University, Bundoora, 3083, Australia. Recent advances in the study of Parkinson's disease (PD) suggest that two primary disorders affect movement performance: akinesia and hypokinesia. Patients frequently experience difficulty walking because they cannot initiate the stepping response, or because they cannot maintain a large stride. Whereas traditional methods of movement rehabilitation focussed on the treatment of rigidity, tremor and generalised muscle weakness, current methods place greater emphasis on the management of hypokinesia and akinesia, which are considered to have greater impact on functional movements such as walking. Based on this approach, together with the findings of recent research on the motor control of gait and the motor functions of the basal ganglia, we propose five principles for movement rehabilitation in PD: (1) the ability to generate normal movement is not lost in PD, the problem is one of activation of the intact neural networks; (2) more normal movement can be elicited using external visual auditory or proprioceptive cues; (3) by consciously focussing attention on the criterion movement pattern more normal movement can be generated; (4) movement performance will deteriorate when secondary cognitive or motor tasks are performed; (5) movement disorders are more pronounced for long or complex movement sequences. We describe the theoretical basis for these principles and discuss how they can be incorporated into movement strategies to normalize the walking pattern in PD.

62 M.E. Morris et al. 1. INTRODUCTION One of the great challenges that confronts specialists in movement rehabilitation is to determine how people with Parkinson's disease (PD) can be assisted to walk with greater independence and a more normal walking pattern. Parkinson's disease affects up to one in every 100 people over the age of 65 years (Schoenberg, 1988) and of those afflicted, at least 80% develop gait disorders as the disease progresses. Because gait disturbance is associated with an increased risk of falls (Koller, Glatt, Vetere-Overfield & Hassanein, 1989) and is one of the primary reasons why patients can eventually require hospitalization, the rehabilitation of walking is a primary focus of health care services for PD. In this chapter we outline the major gait disorders in PD and review the literature which investigates the efficacy of traditional rehabilitation methods for this problem. The review reveals only limited success for the traditional approaches, prompting an analysis of underlying reasons, which indicates that the limited success was partly due to an inadequate knowledge of the pathophysiology of PD. There is now considerable evidence that PD is due to basal ganglia (BG) disturbance (Bemheimer, Birkmayer, Homykiewicz, JeUinger & Setelberger, 1973; Biller & Brazis, 1990; De Caballos, Ferrandez, Jenner & Marsden, 1993; Marsden, 1990). Moreover the BG are fundamentally implicated in the control of normal gait (Blin, Ferrandez, Pailhouse & Serratrice, 1991; Blin, Ferrandez & Sermtrice, 1990; Brooks, Salmon, Mathias, Quinn et al., 1990; Martin, 1967; Stem et al., 1983). The role of the BG in movement control is therefore considered in detail. Two major roles are elaborated, a contribution to movement preparation and another to internal cueing of submovements in well learned motor sequences such as walking. These theoretical concepts are then used to generate several hypotheses for therapeutic intervention which are presented in the final stages of the chapter. 2. GAIT DISORDERS IN PARKINSON'S DISEASE The two most frequently observed gait disorders in PD are hypokinesia, or reduced amplitude and speed of walking, and akinesia, which refers to difficulty in gait initiation. Clinically, the person with gait hypokinesia is seen to walk with short, shuffling steps, reduced movement of the head, arms and trunk and reduced angular displacement of the lower limb joints (Blin et al., 1990; Blin et al., 1991; Knutsson & Martensson, 1971; Murray, Sepic, Gardner & Downs, 1978). In contrast to the haphazard, uncoordinated walking pattern of a person with cerebellar ataxia, the person with PD has poverty of movement. The basic kinematic patterns for walking appear to be preserved (Knutsson,

Motor Control Considerations in Parkinson's Disease 63 1972; Stem, Franklyn, Imms & Prestidge, 1983). However the amplitude and in some cases the timing of movement is reduced, which in turn leads to a low velocity (hypokinetic) gait. Gait akinesia is characterized by difficulty in activating the stepping mechanism and in some cases there can be a complete absence of movement when the person tries to step forward. Gait akinesia can also be associated with \"freezing\" episodes (motor blocks) during walking and turning. Studies on akinetic freezing by Stem, Lander & Lees (1980) and Giladi, McMahon, Przedborski, Flaster et al. (1992) indicate that motor blocks are context dependent. The person might, for example, experience akinetic freezing when attempting to walk through a narrow doorway or when navigating their way around bedroom furniture. Walking in a crowded shopping centre can become extremely difficult due to recurrent freezing episodes even though the ability to walk outside in large open spaces is retained. The tendency to freeze appears to be exacerbated when there is a need to quickly assimilate complex visual information during the act of walking. Stepping on to a moving escalator or stepping aside to avoid colliding with a dog, small child or shopping trolley are examples of this difficulty. Yet paradoxically people with PD also report that visual cues provide a powerful means by which freezing can be overcome (Quintyn & Cross, 1986). For example, the person who freezes when walking through a particular doorway can accomplish this task easily when white markers are placed on the floor at the appropriate footstep distance (Mc Goon, 1990). In a similar way, an upturned walking stick can be used as a visual cue to trigger stepping when the person is walking outside or in the community (Dunne, Hankey & Edis, 1987). The stick is turned upside down and attention is directed toward stepping over the handle, which quickly elicits a stepping response. Alternatively, when visual cues are unavailable mental imagery can be of assistance. Mc Goon (1990), for example, reported that start hesitancy could be overcome by visualizing stepping over a log. These anecdotal reports highlight the importance of cognitive and visual mechanisms in the control of locomotion in PD. Additional gait disorders can include dyskinesia (over activity which manifests as wriggling or writhing movements), dystonia (excessive muscle activity in a specific muscle group such as the invertor muscles of the foot which leads to abnormal alignment of the body segment), rigidity and tremor. The tremor is not a major problem because it is a resting tremor which disappears on movement (Hallett, 1993). As described in James Parkinson's (1817) essay on the \"Shaking Palsy\" a proportion of patients also exhibit gait festination, whereby the foot steps become progressively shorter and faster. However recent reports indicate that festination only occurs in approximately 16% patients and at the final stages of disease progression (Calne, Larsen & Burton, 1985).

64 M.E. Morris et al. The short stepped hypokinetic walking pattern is much more characteristic of the \"typical\" Parkinsonian gait. 3. TRADITIONAL GAIT REHABILITATION A primary aim of movement rehabilitation in PD is to augment the benefits of pharmacological therapy by teaching patients and their carers alternative strategies for moving with greater ease, speed, independence and skill (Banks & Caird, 1989). Pharmacological therapy, and in particular levodopa therapy, can initially have dramatic effects on movement performance. However the majority of patients suffer motor fluctuations subsequent to the first five years of levodopa (Marsden, Parkes & Quinn, 1982) and disorders of some gait parameters such as the walking cadence (steps per minute) appear to be resistant to medication (Blin et al., 1990). Movement rehabilitation strategies can be particularly beneficial at the end of dose, when akinesia, hypokinesia and dyskinesia are heightened, or at the stage of disease progression when medication has ceased to have a beneficial effect on movement control. Although there is now considerable evidence that the key gait disorders in PD are hypokinesia and akinesia (Hallett, 1993), traditional physical therapy focussed on the treatment of rigidity, tremor and secondary musculoskeletal disorders associated with disuse. The earliest methods incorporated massage, relaxation and soft tissue mobilization to help reduce rigidity and stiffness (Ball, 1967; Doshay, 1962). Another popular method, known as proprioceptive neuromuscular facilitation (PNF) included a technique which aimed to reduce trunk rigidity by facilitating trunk rotation in supine (Knott, 1957; Minnigh, 1971). It was assumed that after the therapist reduced the rigidity of the trunk muscles in supine a more normal walking pattern would be achieved. The commonly held belief was that\" ...to increase movement there must be a decrease in rigidity\" (Umphred, 1988, p.563). Yet another method was directed towards reduction of tremor of the limbs in supine by graduated weight bearing and resistance (Umphred, 1988, p.565). It was presumed that reduction of tremor achieved with resisted mat exercises would lead to more coordinated walking and limb movements. Traditional physical therapy for PD also placed a great deal of emphasis on group exercise therapy with the aim of maintaining muscle strength, range of joint motion and cardiovascular fitness (refer to Bilowit, 1956; Davis, 1977; Franklyn, 1983; Hurwitz, 1989; Homberg, 1993; Szekely, Kosanovich & Sheppard, 1982; Wroe & Greer, 1973). Yet there is no evidence that Parkinsonian patients suffer problems in these areas any more than elderly people in general. Although both fit healthy elderly people and elderly people with PD might show some general benefits from regular fitness routines, there is

Motor Control Considerations in Parkinson's Disease 65 no clear data to suggest that general exercises have an effect on the specific movement disorders associated with PD. There are a number of additional reasons why the traditional approach is now difficult to justify. Firstly, patients very rarely report rigidity and tremor as problems In many cases rigidity goes unnoticed until the clinician draws attention to the \"stiffness\" in the trunk and limbs. Likewise the tremor exhibited in PD is a resting tremor that disappears during movement and patients do not perceive it to be the major limiting factor in ambulation. The second reason why these methods are difficult to justify is they assume that achievement of normal movement performance in one context, such as lying supine performing an exercise with a therapist, will carry over to a different context, such as walking independently without the therapist. The motor skills literature indicates that transfer of training is enhanced when the desired movement is practised in the context in which it will eventually be used (Newell, 1991). There is little evidence to suggest that practice of trunk rotation movements or weight bearing activities in supine would carry over to the comparatively unrelated task of walking. The third difficulty with these facilitatory techniques is that they have not been substantiated with controlled research. Although clinical reports have overwhelmingly emphasized the beneficial effects of facilitatory methods (eg. Ball, 1957; Knott, 1957; Umphred, 1988) there has not been a trial which demonstrates that they are effective for PD, despite their use for nearly 30 years. The controlled trials of physical therapy for PD that have been conducted have generally yielded disappointing results. Each of these studies have evaluated the effects of group exercise programs. In one study Gibberd, Page, Spencer, Kinnear & Hawksworth (1981) found that treatment that aimed to improve range of movement, trunk rotation, balance and walking utilizing facilitatory techniques derived from PNF, Bobath and Peto methods was not helpful for the movement disorders associated with PD. Similarly, Pedersen, Oberg, Insulander & Vretman (1990) found no measurable improvement in gait, muscle strength or function from a 12 week group exercise program. Patients did, however, report an impression that physical therapy was beneficial. Only three studies (Banks & Caird, 1989; Formisano, Pratesi, Modarelli, Bonifati & Meco, 1992; Palmer, Mortimer, Webster, Bistevins & Dickinson, 1986) have claimed positive effects for group exercise therapy. In one trial, Palmer, Mortimer, Webster, et al. (1986) compared the effects of two different 12 week programs for PD. The effects of a stretch exercise program led by a physical therapist was compared with upper body karate techniques taught by a rehabilitation nursing student with a black belt in karate. Patients in both groups showed increased strength and coordination and decreased tremor although there was a decline on tests requiring timed movements of the limbs and

66 M.E. Morris et al. trunk. The finding that such different treatment approaches yielded similar results provides some suggestion that non-specific treatment effects such as enhanced motivation and attention might have been responsible for the changes observed. Because a no-treatment control group or sham treatment group was not included in the trial, this possibility cannot be discounted. The same criticism applies to the study conducted by Banks and Caird (1989) who claimed that a daily exercise program provided for two weeks was effective in improving walking, turning over in bed, sitting up and rising from a chair. Likewise, the finding of Formisano et al. (1992) that a three month program of passive mobilization, mat exercises, finger exercises and wide-based walking was effective in improving movement speed, tremor and rigidity was clouded by the failure to include controls for maturational, historical and motivational variables. Collectively, these studies do not provide convincing evidence that facilitatory techniques, routine exercises and group classes are effective in ameliorating the specific movement disorders associated with PD. These considerations raise the question of why facilitatory techniques and general fitness classes were originally advocated for the treatment of PD. The likely answer seems that these methods were in use in the 1950s and 1960s for the treatment of stroke and poliomyelitis, at a time when rehabilitation therapists were branching into other areas of neurological practice. It appears that they were adopted for the treatment of PD when little was known about the motor functions of the basal ganglia (BG) and when clinical practice evolved from empirical observations rather than consideration of theories of how movement is controlled. Given the recent advances in understanding how the brain controls movement and how movement control is disrupted in BG disease, it now seems timely to reconsider methods of movement rehabilitation for PD. In the following sections we will illustrate how a scientific framework for rehabilitation practice can be derived from contemporary knowledge of the motor functions of the BG. 4. MOTOR FUNCTIONS OF THE BASAL GANGLIA Current concepts of the role of the BG in movement control are based on several lines of research. These include studies on the anatomical relationships of the BG with the motor areas of the cortex (Parent & Hazrati, 1993; Alexander & Crutcher, 1990; Schell & Strick, 1984); cerebral blood flow studies in people trained to perform sequential finger movements (Roland, Meyer, Shibasaki, Yamamoto & Thompson, 1982; Sietz & Roland, 1992); single cell recordings in primates (Brotchie, Iansek & Home, 1991 a, b; Tanji & Kurara, 1985; Tokuno, Kimura & Tanji, 1992); and investigations of movement performance in PD (Georgiou, Iansek, Bradshaw, Phillips, Mattingley & Bradshaw, 1993; Georgiou, Bradshaw, Iansek, Phillips, Mattingley & Bradshaw, 1994; Martin,

Motor Control Considerations in Parkinson's Disease 67 Phillips, Iansek, & Bradshaw, 1994) and Huntington's Chorea (Bradshaw, Phillips, Dennis, Mattingley et al., 1992). The main theme to emerge from this literature is that the BG participates in a motor circuit (sensorimotor cortex-BG-thalamus, supplementary motor area/premotor area-sensorimotor cortex) that enables movement sequences to be initiated and executed with normal speed, amplitude and force. Anatomical studies have demonstrated that the BG consist of a number of subcortical nuclei located deep within the brain. These nuclei are the putamen, caudate nucleus, internal and external segments of the globus pallidus, substantia nigra and subthalamic nucleus (Figure 1). The putamen and globus pallidus constitute the motor portion of the BG (Schell & Strick, 1994) and the major input to these areas is from the sensorimotor area of the cortex. Dopamine is one of the main neurotransmitters used by the nigrostriatal pathway which innervates the putamen and its depletion is associated with akinesia and hypokinesia. In PD there is a progressive reduction of neurones in the substantia nigra of the brainstem which are responsible for the secretion of dopamine and as a consequence movements become progressively slower and more difficult to initiate (for further detail refer to the preceding chapter by Iansek, Bradshaw, Phillips, Cunnington & Morris this volume). CEREBRAL CORTEX SMA PMC MC 1 ! BASAL GANGLIA1 1 GLOBUS~ lCAUDATE ~0 To spinal SMA Supplementary Motor Area PMC Premotor Cortex cord MC Motor Cortex VL VenterolateraTlhalamus STN SubthalmicNucleus SN SubstantiaNigra Intralaminar Nucleus Figure 1. Majorprojectionsof the BasalGanglia

68 M.E. Morris et al. The major output projections of the BG are the supplementary motor area (SMA) and premotor cortex (PMC) and it has been argued that the BG influence the running of movement sequences via their interaction with these two areas (Brotchie et al., 1991 a; Brotchie et al., 1991 b; Marsden, 1990). Cerebral blood flow studies provide some indication that the SMA plays a role in the preparation for and the execution of movement sequences that are predictable and well learned (Roland, Larsen, Lassen & Skinhoj, 1980). Positron emission tomography (PET) studies further suggest that the putamen and globus pallidus interact with the SMA in running predictable movement sequences (Roland et al., 1980). The interaction between the BG and SMA only occurs when the motor skill is well established. During the early stages of learning a motor sequence the cortical regions are predominantly activated (Sietz & Roland, 1992). However once the sequence is well learned and the movement is automatic, the BG are activated preferentially to the cerebral cortex (Sietz & Roland, 1992). Single cell recordings in behaving primates have also demonstrated a close interaction between the BG and SMA in the execution of movement sequences. Tanji and Kurata (1985) showed that the SMA neurones have sustained discharge prior to the onset of movement. The BG neurones demonstrate brief bursts of discharge throughout the movement and towards the end of each component movement in a sequence (Brotchie et al., 1991a). These brief bursts of activity in the BG (which could be considered to be phasic cues) are only generated if movements are predictable and easy to perform. When movements are unpredictable, novel or require a great deal of attention, little or no phasic activity is generated within the BG. Brotchie and colleagues (1991b) suggested that the phasic activity generated within the BG during sequential movements is timed to switch off the premovement activity in the SMA that is being generated in anticipation of the next submovement in the sequence. This abrupt termination of activity in the SMA enables the next submovement in the sequence to be executed and at the same time a new cue is generated within the BG to turn off the SMA preparatory activity for the following submovement. Once this process has been initiated the entire sequence can be run automatically. Collectively these findings suggest two likely ~nctions of the BG. One function appears to be the provision of internal motor cues for the running of well learned, predictable movement sequences that require little attention to be executed (eg, walking, speaking, swallowing, f'mger tapping, knitting, writing). Second, the BG may contribute to the pre-movement preparatory activity of the SMA. This preparatory activity is known as \"motor set\" and enables the entire movement sequence to be maintained in a state of readiness for action (Robertson & Flowers, 1990). It would be expected that a deficiency of internal motor cues would lead to a difficulty in the switching process between one submovement and the next within a sequence. At a neuronal level this

Motor Control Considerations in Parkinson's Disease 69 provides a potential explanation for the clinical observations of hypokinesia (reduced movement). An absence of the contribution of the BG to premovement activity would lead to difficulty in the maintenance of preparedness for the entire sequence which could lead to both akinesia (absence of movement) and hypokinesia. The alternative viewpoint is that hypokinesia and akinesia in PD arise from increased inhibition of the thalamus by the globus pallidus (refer to Alexander & Crutcher, 1990). In addition to the BG-thalamocortical motor circuit, two projection systems occur within the BG circuitry. There is a \"direct\" pathway which projects from the putamen to the motor sections of the internal segment of the globus pallidus and substantia nigra pars reticulata. There is also an \"indirect\" pathway which arises from putamen and influences the BG indirectly through its connections with the external segment of the globus pallidus and subthalamic nucleus (De Long, 1990). The direct pathway normally provides positive feedback (facilitation) to the SMA, PMA and primary motor cortex whereas the indirect pathway inhibits these areas (Alexander & Crutcher, 1990). Alexander and Crutcher (1990) hypothesized that shifts in the balance between the activity of the direct and indirect pathways and the subsequent changes in activity of the internal globus paUidus and substantia nigra lead to the slowness of movement observed in Parkinsonian hypokinesia and the fast, involuntary, ballistic movements exhibited in Huntington's chorea. They proposed th, t increased conduction in the indirect pathway leads to hypokinesia by increasing the pallidothalamic inhibition whereas decreased conduction in the direct pathway leads to hyperkinesia by reducing pallidothalamic inhibition. This interpretation is not necessarily contradictory to the view put forward by Brotchie et al. (1991 a, b) and Marsden (1990). However, it neglects to provide an explanation for the interaction of the BG and SMA in movement performance as revealed by cerebral blood flow studies and single cell recordings. Furthermore, as pointed out by Marsden and Obeso (1994), the hypothesis that the motor loops within the striato-pallidal circuits inhibit unwanted movement via their thalamic connections is at odds with the results of stereotaxic surgery, which show that pallidal and thalamic lesions do not worsen hypokinesia and akinesia in PD, even though they reduce tremor and rigidity. Investigations on movement disorders in PD further reinforce the suggestion that the BG play a role in the control of serially ordered, long or complex movements that are well learned (refer to Phillips, Bradshaw, Iansek & Chiu, 1993 for a review). Ballistic movements are also compromised in PD because the initial burst of agonist activity is under scaled and the person requires additional muscle contractions to generate the required force level (Hallett & Koshbin, 1980). However the movements that are most disrupted axe complex sequential actions that are performed automatically. The difficulty in performing these movements becomes most apparent when the person attends to a

70 M. E. Morris et al. secondary task. A classic experiment by TaUand and Schwab (1964) required patients to perform two separate motor tasks (drawing a triangle and squeezing an ergometer bulb) in isolation and then simultaneously. Each task could be performed normally in isolation. However when they were performed concurrently the task relegated to \"subconscious\" control showed marked deterioration in speed and accuracy. In a similar series of experiments Benecke, RothweU, Dick, Day & Marsden (1986; 1987) demonstrated that isolated movements were not substantially slower in PD whereas the same movements performed as part of a movement sequence or simultaneous motor task were slower than normal. Clinical studies have also shown that motor performance in PD is enhanced when patients are provided with visual, auditory or proprioceptive cues. Presumably external cues assist patients to move with greater speed and coordination because they substitute for the disordered phasic cue from the gl0bus pallidus by allowing alternative control mechanisms to be utilized. Recent research on premovement potentials of the brain (Cunnington, Iansek, Bradshaw & Phillips, in press) suggests that in PD control mechanisms switch from SMA to premotor cortex (PMC) mediated control when conscious attention strategies and external cues are available. In contrast to the SMA, which appears to be cohcerned with internally guided movements, it has been argued that the PMC is more concerned with the external regulation of movement (refer to the chapter by Iansek et al., this volume). To summarize, converging lines of evidence support the view that the BG have two important motor functions. The first is to provide the internal cues that trigger submovement execution within a movement sequence. The second is to enhance the preparatory activity of the SMA to establish \"motor set\" for the entire movement sequence. When these functions fail the person experiences movement slowness and difficulty in initiating action. To compensate, the Parkinsonian patient is seen to rely heavily on visual, somatosensory and proprioceptive feedback to control movement. Furthermore, by consciously attending to movements whilst they are performed rather than nmning the movements automatically, the person with PD can move more quickly and easily. 5. GAIT CONTROL AND PARKINSON'S DISEASE The question of how human walking is controlled has been the subject of considerable debate amongst motor control theorists. Although a great deal has been written on the subject of gait control, most of this work has been concerned with animal models and, to a lesser extent, locomotion in the intact human central nervous system (CNS). In movement rehabilitation, however, the aim is to teach people with CNS disorders

Motor Control Considerations in Parkinson's Disease 71 strategies to achieve a more independent, coordinated walking pattern. The paucity of literature on the motor control of gait in PD makes this a difficult task and currently clinicians are largely in the position of having to derive treatment principles from what is known about gait control in intact systems. In teaching the person with PD strategies for walking more normally it is useful to consider the fundamental requirements for successful locomotion. First, the person needs to be able to generate a basic rhythmical stepping pattern (Patla, 1991a; Nutt, Marsden & Thompson, 1993). Second, sufficient force needs to be generated in the lower limb musculature with appropriate timing to propel the person in the desired direction (Patla, 1991a; Nutt, Marsden & Thompson, 1993). This necessitates the conversion of potential energy of the body into kinetic energy to permit the body to fall ahead of the supportive extremity (Knutsson, 1972; Winter, 1992) as well as re- conversion of kinetic energy back into potential energy on weight acceptance (Inman, Ralston & Todd, 1981; Winter, 1992). Activation of the locomotor muscles also serves to prevent collapse of the legs during the stance phase of gait (Winter, 1980) and to regulate the trajectory of the foot to enable ground clearance (Winter, 1992). The third requirement for successful locomotion is the ability to maintain balance and equilibrium in order to retain an upright posture (Nashner, 1980, 1982; Cappozzo, 1981; Patla, 1991a). Finally, the person needs to have the ability to adapt balance and locomotion according to changing environmental demands, the goals of the task and their intentions. In the following sections we briefly summarize some of the motor control mechanisms related to each of these requirements and use this information as a basis for understanding gait disorders in PD. 5.1 Generation of Rhythmical Stepping Several lines of evidence suggest that the basic pattern of movement for repetitive, rhythmical behaviors such as walking is produced by specialized groups of neurones found in the CNS, known as central pattern generators (CPGs) (refer to Selverston, 1980 for a review). Central pattern generators were originally described as discrete ensembles of neurones located in the spinal cord that could generate a rhythmic motor pattern without the need for peripheral sensory input (Wilson, 1961). Investigations by Shik, Orlovskii and Severin (1968) and Grillner (1975) then went on to show that the basic motor pattern could also be produced without complex input from supraspinal regions. They used a preparation known as \"fictive locomotion\" in experimental cats to demonstrate this point. In the fictive preparation the spinal cord was transected to isolate it from supraspinal inputs and the dorsal columns were severed to prevent sensory feedback. The cats were then placed on a moving treadmill and the brainstem locomotor regions in the mesencephalon or the spinal cord were electrically stimulated. In these

72 M.E. Morris et al. conditions rhythmical, alternating leg movements which resembled locomotion were observed. A further f'mding was that when the cats were held on a split belt treadmill with the belts moving at two different speeds the resultant walking pace was the average of the two speeds (Forssberg, GriUner, Halbertsma & Rossignol, 1980). These results indicated that spinal cord regions could generate locomotor patterns without the need for sensory feedback or supraspinal input to specify the direction of movement or how much each muscle should contract. Therefore in the earliest conceptualization of the CPG it was thought that simple, non specific input was able to drive complex, patterned output from spinal cord generators (Patla 1991 a; Patla, 1991 b). The CPGs were considered to be made up of motor neurones, intemeurones or both types of neurones (Selverston, 1980). Although this basic model of the CPG may well apply for the regulation of rhythmical movements in amphibians, reptiles, birds and sub-primates (see Baev & Shimansky, 1992), it is clear that in humans complex supraspinal input modifies the timing of the locomotor rhythm and the activation profiles of specific muscles (Patla, 1991a; Patla, 1991b). Moreover it has been demonstrated that somatosensory input (Swinnen, Massion & Heuer, 1994) and reflex responses (Duysens & Tax, 1994) contribute to the control of rhythmical movements such as walking, running and hopping in humans. Therefore, the most recent conceptua!izafions of the CPG highlight that peripheral feedback and input from supraspinal structures such as the cortex, mesencephalic and pontine motor regions, the BG and the cerebellum interact with the spinal ensembles to produce coordinated walking that is specific to the goals of the task and the environment in which movement occurs. At a clinical level it is clear that brain regions such as the BG, cortex and cerebellum must play a role in locomotor control because damage to these areas leads to classic gait disorders such as the wide based ataxic gait associated with cerebellar disease (Nutt et al., 1993), knee hypemxtension in stroke (Morris, Matyas, Bach & Goldie, 1992) and hypokinetic walking exhibited in PD (Murray et al., 1978; Nutt et al., 1993). In PD, gait hypokinesia is directly related to a problem in regulating the length of the stride (as will be further discussed in the following section). Of interest, when visual cues are provided which specify the appropriate stride length, the walking cadence, velocity and stride length closely approximate normal values (Morris, Iansek, Matyas & Summers, 1994 a, b). Therefore people with PD retain the ability to elicit a normal stepping pattern when they are given the appropriate input. This finding suggests that the spinal locomotor networks remain intact although their activation requires appropriate input from the motor regions of the BG. When the mediating influence of the BG is lost the person can still walk, however the quality of the gait pattern is significantly compromised.

Motor Control Considerations in Parkinson's Disease 73 It should also be mentioned that although animal studies support the existence of CPGs, the ability of the spinal cord to produce fundamental stepping rhythm has never been conclusively demonstrated in humans. There are notable differences between human and animal gait. The primary difference is that human locomotion is bipedal therefore there is greater need for supraspinal input to ensure that the center of mass is controlled over a very small base of support. In contrast, in quadrupeds there are usually three limbs in contact with the ground which provides for greater stability. The bipedal gait pattern also imposes greater demands for propulsion and force control. Whereas the spinal cord networks may well be able to generate the basic pattern of rhythmical output, the need for careful regulation of force in order to produce concentric or eccentric muscle contractions with appropriate timing is possibly mediated by supraspinal influences (refer to Patla, 1991 a). The capability of humans to free the upper limbs for complex manipulative tasks whilst ambulating and their ability to change the gait pattern at will also highlights the role played by supraspinal influences in the final expression of the gait pattern. 5.2 Regulation of Muscle Force and Timing During Gait For the person to be able to transfer potential energy to kinetic energy and back again during walking, muscles need to be activated with appropriate levels of force at the optimal time in the gait cycle. As we have already outlined, some theorists have argued that muscle activation patterns are controlled by CPGs located in the spinal cord which are strongly influenced by brainstem and cortical input. However not all of the control of human locomotion is mediated by the central nervous system (CNS). As highlighted by the dynamical systems approach (eg. refer to Kugler & Turvey, 1987), the periodicities evident in cyclical, rhythmical movements such as walking are regulated by a range of factors such as the physical properties of muscle, including its stiffness, length and contractility; the constraints of the anatomical configuration of the skeleton; and the effects of gravity, inertia and ground reaction forces on the center of mass of the body and the center of mass of each leg. Dynamical accounts of locomotion also incorporate the concept of CPGs (eg. Collins & Stewart, 1992; Collins & Stewart, 1993) although they are seen as coupled nonlinear oscillators which can be assembled to specify the core locomotor pattern for a given task rather than immutable spinal generators which contain all of the muscle commands for movement. From this point of view it has been further argued that the phasing of muscle activity during locomotion emerges as a consequence of the natural oscillatory frequencies of the limbs as the person interacts with the environment for a given task (eg., refer to Whithall & Clark, 1994). Many aspects of rhythmical movement have been modelled according to coupled oscillatory systems. One of the best examples of an oscillator is the pendulum.

74 M.E. Morris et al. The characteristic feature of a pendulum is that its trajectory through space (ie. the phase path) is highly reproducible. Once an impulse has been delivered to initiate the movement of the pendulum it will swing with a very consistent phase path, with ever decreasing amplitude until it eventually settles. The second characteristic is that if there is a perturbation to the motion of the pendulum or a change in initial conditions, it will soon return to the same phase path. The tendency of the pendulum to return to its original behavior is known as \"structural stability\". Another characteristic of oscillators is that if they are paired, the frequency of movement of one tends to entrain to the rhythm of the other. This tendency to synchronize in phase with each other is a feature of limit-cycle oscillators and can be observed across many levels of biological systems. The rhythmical movements of the lower limbs during locomotion can be considered in terms of coupled oscillatory systems. More specifically, each leg can be modelled dynamically as a limit cycle attractor, and the coordination between the two legs can be modelled as a system of non-linear limit cycle oscillators (Whitall & Clark, 1994). Once force pulses have been delivered at two adjacent phases of the gait cycle known as A4 (calf push off) and H3 (hip flexor pull off) (Winter, 1992), the leg swings like a pendulum through space until its motion is arrested by strong contraction from the hamstring and quadriceps, which serve to decelerate the leg. When the leg is briefly perturbed, for example when the person catches the foot on an obstacle, its motion quickly returns to the same phase path, or preferred behaviour (refer to Kay, Saltsman & Kelso, 1991). Furthermore when the velocity of movement is steadily increased there eventually comes a critical point when a transition is made from walking to running gait. This \"phase transition\" from one mode of behavior to another is also characteristic of limit cycle oscillators and lends support to the concept that biological motion appears to be self organized (Kugler & Turvey, 1987). Because much of the walking pattern can be controlled by the natural tendency of the legs to act as oscillators, the number of degrees of freedom to be controlled by higher order executive structures is reduced, although these higher levels can take over when required. This literature is important from a clinical perspective for a number of reasons. First, it suggests that clinicians may not need to teach patients to consciously focus on controlling the entire movement trajectory of the lower limb during walking. Simply focussing on generating force at the critical loci in the gait cycle and allowing the leg motion to naturally respond with a pendular like motion might prove to be enough for generating a more normal gait. The second implication is that patients might be able to enhance their walking pattern by capitalizing on the natural predisposition toward entrainment in complex systems. The extent to which the motion of the lower limbs during gait can be entrained to cyclical movements of the arms or the rhythmicities of speech in PD is a question that warrants further investigation by clinicians and theorists

Motor Control Considerations in Parkinson's Disease 75 alike. Finally, if locomotion is governed by constraints imposed by the CNS, musculoskeletal system, environment and task, then clinicians might be able to identify and manipulate key constraints in order to shift the system toward a more normal gait. In this context the skill of the clinician is to identify the critical control parameters and then to deliberately change these parameters in order to enhance behavior (refer to Clark, 1994). Potential control parameters that appear to influence gait hypokinesia in PD include stride length and velocity. The environment and task also afford a range of visual and auditory stimuli that can constrain movement, as will be discussed in Section 6. 5.3 Balance and Postural Control In addition to the ability to generate a rhythmical locomotor pattern, the capacity to maintain balance and postural control is an essential requirement for normal walking. This is illustrated by studies on the emergence of locomotion in human infants which have shown that although ten month old babies cannot walk they can generate a basic locomotor pattern when the propulsive and balance requirements for walking are eliminated, for example by placing the legs and trunk in water or by holding the child on a treadmill ('I'helen 1986; Thelen, Ulrich & Miles, 1987). Similarly, Bussel, Robi- Bramby, Azouvi, Birab et al. (1988) demonstrated that patients with spinal cord lesions were able to generate a basic rhythmical stepping pattern when they were placed on a treadmill whilst their body weight was relieved by a harness. Balance disturbance is a characteristic feature of PD and some, such as Knutsson (1972), have argued that gait hypokinesia is an adaptation to allow the person to balance more easily and to diminish the requirement for rapid alterations of muscle activity during walking. Therefore the limitations in range and speed of movement in Parkinsonian gait may not reflect the ultimate limits of motor control. It is well recognized that people with PD have balance disorders and are particularly vulnerable to unexpected perturbations to their center of mass. There is an increased risk of falls in PD (Koller et al., 1988) and the majority of falls occur during ambulatory tasks (Smithson, Morris & Iansek, 1994). People with PD have been reported to spend considerably longer in the double limb support phase of gait (Murray et al., 1978), presumably to decrease the demands of balancing on one leg. Nevertheless, as yet there is no experimental evidence that gait hypokinesia is simply a compensation for decreased balance and postural control.

140 140 120- 120 I00 100 \"~ 80 o 60 \"~ 80 40 o 60 z 40 CS 140 CS Preferred Gait Walk Faster 140_ 120 [ ~ CS Visual Cues 120 .............................. |00. . .~ 80 !00 .......... 80- 0 60\" 2; ~. 40 2; ~. 40- 20 0 20 0 VC S Metronome And Walk Faster Figure 2. Means and 95% confidence intervals for gait velocity, stride length and cadenc permission of Oxford University Press, from Morris et al, (1994). The pathogenesis of gait h

r4o V = Velocity C = Cadence 12o S = StrideLength O0 \"•80. ~; 60. 40. S Vc Metronome 14o 120 ............. I00 ............. _'71- \"-'1\"- j -~ ~o ~6o ~. 40 20 90 1 S VCS Visual Cues And Walk Faster ce in Parkinson's disease, expressed as a percentage of normal. (Reprinted by hypokinesis in Parkinson's disease, Brain).

Motor Control Considerations in Parkinson's Disease 77 Recently we completed research which suggested that in mild PD gait hypokinesia is fundamentally due to a problem in regulating the stride length and that balance disturbance has less of an effect on the stepping pattern (Morris et al., 1994a). When Parkinsonian subjects were asked to walk at their preferred speed their average velocity was 47.6 m/min and their average stride length was 0.89 m. When instructed to walk at the speed of age matched controls they could usually achieve the desired velocity (on average 65.6 m/min). However this was achieved with an abnormally high cadence and a very short stride length (Figure 2). When patients were provided with white floor markers placed to externally cue the appropriate stride length, the gait velocity; stride length, cadence and period of time spent in the double limb support phase all approached normal values (Figure 2). Provision of auditory cues from a metronome helped to normalize the walking cadence yet had little effect on stride length. Together these findings suggest that a fundamental deficit in hypokinetic gait is the regulation of stride length. When this is controlled by the use of external cues a near normal gait pattern can be elicited (Morris et al., 1994a). The other notable fmding was that the period spent in the double limb support phase of gait decreased with visual cues without patients losing their balance. This appears to indicate that balance disturbance was not the primary cause of gait hypokinesia in this sample of people with relatively mild PD. 5.4 Environmental and Task Constraints on Gait The ability of humans to quickly adapt their walking pattern to changing environmental conditions and different task demands is a remarkable skill. This is particularly the case when one considers that such adjustments are usually achieved without thought or whilst the person performs a secondary task. Such is not always the case for people with PD however, who often need to attend carefully to changes in the slope, stability and compliance of the support surface or the trajectory of moving objects in the near environment during gait. When there are competing demands, such as the need to talk or watch oncoming traffic whilst walking, the gait pattern usually slows down (Mc Goon, 1990). As previously mentioned, some environmental constraints such as narrow doorways or confined spaces promote akinetic and hypokinetic gait. Yet paradoxically, other environmental stimuli, such as lines on the pavement afford a more normal walking pattern (Martin, 1967). It appears as if the person with PD is more highly dependent than normal on contextual information, and in particular visual information, to regulate locomotion. Purdon Martin (1967) was one of the first clinicians to make the astute observation that certain visual stimuli could enhance the walking pattern in PD. He noted that ...\"whim lines, one or two inches wide, and eighteen inches or so apart on a dark ground

78 M.E. Morris et al. produce a pronounced, sometimes dramatic effect enabling a patient who seemed unable to walk, or the shuffler on short steps, to step out strongly.\" (1967, p.33). Subsequent studies have confirmed that visual cues have a powerful effect on gait performance in PD (Bagley, Kelly, Tunnicliffe, Turnbull & Walker, 1991; Dunne et al., 1987; Forssberg, Johnels & Steg, 1984; Morris et al., 1994a,b; Weissenborn, 1993). The reason why visual cues enhance gait in PD is currently speculative although two possibilities can be considered. One explanation is that people with PD are more heavily reliant on external cues to control their attention and to guide their movement (Brown & Marsden, 1988). When people with PD have to rely on internal control for regulating attentional resources their performance deteriorates (Brown & Marsden, 1988). This finding is in agreement with the studies by Brotchie and colleagues (1991a, b) suggesting that the BG provide an internal cue that triggers submovements in a sequence. The phasic activity of the globus pallidus turns off sustained preparatory activity in the SMA and the abrupt decline in SMA activity triggers movement execution. If the internal cue is abnormal or absent, as in PD, there would be faulty preparation for the next movement and slowness in movement execution. It could be hypothesized that to compensate for this disorder of the internal cueing mechanism, people with PD become highly attuned to external cues to regulate movement. The second possibility is that visual cues constrain the walking pattern by providing a rich source of information that directly specifies stride length and time to contact for consecutive steps. Some investigators (eg., Lee, Lishman & Thompson, 1982; Patla, Robinson, Samways & Armstrong, 1989; Warren, Young & Lee, 1986) have suggested that locomotor control depends on the accurate pick up of information about time to contact of approaching objects and support surfaces. For example, it has been shown that in long jumping the regulation of flight time for the last three steps prior to the jump is closely associated with time to contact with the stepping board ~ et al., 1982). It has also been shown that subjects adjust their step length during walking in response to time to contact information provided by visual cues provided at different loci in the gait cycle (Patla et al., 1989). According to Lee and Young (1986) it is not necessary to perceive both distance and velocity in order to perceive time to contact. Rather a variable known as the the tau-margin can be determined directly by the pattern of input on the optical array, provided the approach velocity is constant. The tau-margin provides extrinsic timing information about a persons relationship to an object or surface. It is possible that this information could be used to assemble the appropriate muscle linkages for walking without the need for intervening central programs to compute the muscle activation patterns (refer to Turvey, 1990). It could also be hypothesized that in PD the rich flow of optical texture in the visual field generated by cues might directly

Motor Control Considerations in Parkinson's Disease 79 constrain the muscle linkages used in locomotion without the need for the BG thalamocorfical circuit to process information. 5.5 Cognitive Constraints on Locomotion The walking pattern of humans is very much influenced by cognitive factors such as the persons goals, intentions, expectations, knowledge and their attention on the movement pattern (Bonnard & Pailhouse, 1993). The fact that humans can imitate another persons walking pattern or learn to walk with a different gait, for example with short, frequent steps (Nilsson & Thorstensson, 1987) further stresses that cognition can strongly influence ambulation. In PD, cognitive input can have a powerful effect on the ability to initiate and sustain rhythmical walking. In one of his most important observations, Purdon Martin (1967) noted that: \"..though it seems at first that stepping is disordered, investigation shows that in appropriate conditions normal stepping can be elicited: the stepping mechanism itself is therefore unaffected, and what is at fault is a physiologically higher function by which stepping is initiated and controlled.\" (Martin, 1967, p. 29). The higher order functions that Martin refers to is probably the BG mechanism that allows well learned movement sequences to run automatically. As outli'ned in Section 2 Parkinsonian patients exhibit a range of gait disorders such as reduced ann swing, diminished stride length and reduced angular displacement of the lower limb joints when they walk without thinking about their gait pattern. However when patients make a concerted attempt to consciously focus their attention on walking \"normally\" many of the movement disorders can be temporarily overcome (Morris, Iansek, Matyas & Summers, submitted). This reinforces the finding of primate studies (eg. Brotchie et al., 1991a, b) that the BG play a role in the control of well learned movement sequences that are run automatically. When movement is shifted from automatic to conscious control, performance is enhanced. The finding that people with PD can use intentional processes to voluntarily change their mode of walking provides rehabilitation clinicians with an opportunity for teaching effective strategies to initiate motion or to alter an aberrant gait pattern. One interesting clinical example is that some people who are unable to initiate walking due to severe akinesia can elicit locomotion by deliberately running (Mc Goon, 1990). By phase locking into a running mode rather than a slower velocity walking gait, rhythmical leg movements can be initiated and sustained. People with PD also appear to have a remarkable capacity to imitate another person's walking pattern provided they do not have a major problem with postural control (Quintyn & Cross, 1986). Yet another clinical observation is that people with balance disturbance can maintain postural control during walking by deliberately preparing for possible perturbations before they walk. For example, by stopping at the beginning of a crowded room and preparing for potential

80 M.E. Morris et al. threats to balance, the ability to withstand a perturbation appears to be enhanced. These anecdotal reports underscore the significance of cognitive constraints on motor performance in PD. The observations just described also suggest the primary disturbance in Parkinsonian gait is not muscle weakness, abnormalities in the sequencing of muscle activation or reduced range of movement of the lower limb joints. Dietz, Berger & Horstmann (1988) confirmed this finding with electromyographic recordings which showed that patients with gait hypokinesia had normal muscle power and their ability to elicit reciprocal activation of lower limb muscles was intact. Gait velocity can also be increased to normal values when the focus of attention is on walking faster (Morris et al., 1994a b, submitted) and the amplitude of ann swing increases when attention is directed toward upper limb movement (Knutsson, 1972). These clinical observations can be interpreted in fight of the model of BG function outlined in Section 4. As indicated by the model, the two functions of the BG are to ensure adequate preparation of the motor cortex for a forthcoming movement sequence and to trigger submovement execution for well learned movement sequences. A disorder of motor set would lead to inadequate motor preparation in the SMA and motor cortex and the descending input from the motor cortex to the spinal locomotor networks would therefore be abnormal. Although the basic walking pattern would be retained, the amplitude of movement would presumably be reduced. A disorder in the phasic cue generated by the BG has the potential to slow down the execution of submovements in the sequence which would manifest as slowness in the timing of consecutive footsteps. However by consciously focussing attention on walking faster, taking a larger stride or swinging the arms, the BG-thalamocortical circuit could be bypassed, allowing more normal expression of the locomotor pattern by spinal cord mechanisms. Rather than using the BG circuitry, it is possible that conscious attention strategies utilize the frontal and prefrontal regions of the brain to prepare the motor cortex for forthcoming movements (refer to Iansek et al., this volume). 6. CURRENT STRATEGIF~ FOR GAIT REHABILITATION Because the kinetics and dynamics of walking in PD have not been adequately investigated, movement rehabilitation strategies are currently based on the therapists' observations of the major kinematic disorders exhibited by people with PD, coupled with their knowledge of basic gait mechanisms in humans. The view put forward in this chapter is that the neural control of locomotion is distributed across many levels of the central nervous system. Although there may exist locomotor networks in the spinal cord that enable the core stepping pattern to emerge, there is powerful input from brainstem, cortical, cerebellar and BG regions that modulate the final expression of the gait pattern.

Motor Control Considerations in Parkinson's Disease 81 There is also evidence that reflexes and somatosensory input play a role in the control of locomotion as do biomechanical factors such as the stiffness of muscle, the anatomical configuration of the musculoskeletal system and the effects of gravity and inertia on body movement. In addition, there are environmental and task constraints on walking. Consideration of these factors as well as the motor functions of the BG gives rise to several hypotheses which can be used to generate new strategies for gait rehabilitation. In the final section we present these hypotheses and explore the ways in which the strategies can be implemented. 1. Parkinsonian patients can elicit a normal stepping pattern, given the appropriate conditions. In this chapter we have highlighted that normal walking is dependent on the ability to (1) generate a rhythmical stepping pattern (2) maintain balance and postural control (3) adapt locomotion according to task and context. One of the most positive findings evident in the literature is that the ability to generate a normal stepping pattern is not lost in PD (Martin, 1967; Morris et al., 1994a, b, submitted). Provided that appropriate conditions avail, locomotion can be initiated and sustained. Furthermore, patients can learn strategies to generate a criterion gait pattern at will (Morris et al., submitted). Through careful assessment of each person's movement disorders the skilled clinician should be able to suggest the most effective strategies for eliciting the stepping mechanism. For some people, particularly those with gait hypokinesia, visual cues might prove effective. Focussing attention on walking with large steps might also be of benefit provided that the person has a good understanding of the criterion stride length and provided that they do not have severe cognitive impairment. On the other hand, akinetic patients might respond more favorably to auditory cueing (for example from an electronic metronome worn on the belt), entrainment to music, mental imagery or focussing on the goal of the task rather than on leg movement. People with disturbed balance may well find that rhythmical stepping can be elicited using hydrotherapy because the buoyancy of the water reduces the requirements for postural control. In a similar way, the provision of a walking frame with wheels can help those with major balance disturbance to generate stepping without fear of falling. Effective movement rehabilitation is dependent on the ability of the physical therapist to assess the responsiveness of the individual to a range of environmental and task conditions for a given medication status, and to adapt the environment and the way in which the task is performed to enable the person to move more easily.

82 M.E. Morris et al. 2. Balance disorders, hypokinesia and akinesia appear to be improved by precueing and conscious attention strategies. The full extent to which people with PD can learn new strategies for improving balance and postural control during walking remains an area for further investigation. Clinical reports do, however, suggest that precueing methods and mental rehearsal can assist in preparation for forthcoming perturbations to the center of mass (Smithson et al., 1994). With an unexpected perturbation in standing the person with PD usually falls backwards with little evidence of protective arm, ankle, hip or stepping strategies (Horak, 1992). However following instruction to prepare for a forthcoming balance disturbance, appropriate righting and equilibrium reactions can be observed. In a similar way, if the person mentally rehearses the appropriate response strategy prior to the perturbation, a more normal reaction can usually be elicited. For those with end stage PD who have major problems with balance and postural control and who cannot utilize these cognitive strategies, the provision of a walking frame increases stability during walking and probably helps to lessen the incidence of falls. The severity of hypokinesia and akinesia also appear to be ameliorated by preparing in advance for context dependent threats to locomotion. One strategy is to train the person to scan the environment and to identify potential environmental constraints that trigger motor blocks. In environments with competing demands on the visual, auditory and somatosensory systems the severity of akinesia and hypokinesia can be reduced by deliberately thinking about the action sequence prior to movement and by developing a plan of action in advance. For akinesia, more basic conscious attention strategies can be evaluated. For example, if the akinetic person thinks of stepping over a log, they can often generate a stepping response, even if cues are absent (Mc Goon, 1990). In some cases stepping can be elicited when the person focuses on the goal of movement (such as walking to the kitchen to get chocolate) rather than thinking about walking with a normal pattern of movement. A recent clinical trial by Yekutiel, Pinhasov, Shahar & Sroka (1991) demonstrated that akinetic freezing and gait hypokinesia were reduced during turning tasks when people with PD were taught to consciously think about turning in a large arc rather than making abrupt switches in direction. By drawing to conscious attention the performance of movements that are normally automatic, both the speed and pattern of movement were enhanced. Both visualization (mental imagery) and imitation techniques can be useful in this respect (Quintyn and Cross, 1986). Furthermore, we have recently demonstrated that patients can generate a normal gait pattern at will by consciously focussing on walking with a criterion stride length following an intensive period of training using physical practice, mental rehearsal and visualisation techniques (Morris et al., submitted). However optimal performance appears to depend on the person developing a very clear mental picture of the required stride size.

Motor Control Considerations in Parkinson's Disease 83 3. The ability to adapt locomotion to changing task demands is enhanced by the provision of external cues. People with PD characteristically have difficulty adapting their walking pattern across a range of contexts. Yet paradoxically, environmental constraints on action, particularly visual constraints, play an important role in the control of locomotion in PD. It is clear that some features of the environment such as narrow doorways, furniture, moving surfaces such as escalators and elevators or the multiple visual inputs in a crowded shopping center predispose people with PD to motor blocks. Yet visual information from floor markers which afford rich input on time to contact can trigger the normal stepping response (Morris et al., 1994a, b; submitted). Clinicians could usefully explore the effects of different environmental constraints on action and modify the persons home environment or way of performing the task to optimize the walking pattern. For example, rearrangement of furniture to ensure that the person does not have to negotiate their way through a cluttered room or make rapid switches in direction could be beneficial. Strips of white tape placed on the floor at the appropriate stride length can be used to trigger stepping in confined spaces and to overcome hypokinesia. Although it has previously been reported that auditory and proprioceptive cues further assist with movement (eg. Laurent & Pailhouse, 1986; Martin, 1967), to date there have been no clinical trials to evaluate the effects of these cues on PD gait. Similarly, there are clinical reports that tactile cues such as a touch on the shoulder or hand, or pressure behind the occipital area of the head can stimulate movement (see Quintern & Cross, 1986), although these reports have not been subjected to empirical confh'mation. The challenge for clinicians is to identify the most effective external cues for generating more normal walking, based on the person's specific movement disorders. 4. Gait performance is compromised when the person with Parkinson's disease performs simultaneous motor tasks. The recent literature indicates that people with PD have difficulty performing two different tasks at the same time. In particular, the walking pattern can deteriorate when the focus of attention is on a secondary task. Mc Goon (1990) noted that when attention is on talking at the same time as walking, the footsteps become progressively shorter, and in some cases the person stops walking. Similarly, the stepping pattern can become compromised when the person attempts to carry a tray of drinks or push a shopping trolley. A preliminary study by Smithson, Iansek and Morris (1994) also showed that the incidence of falls was greatest when Parkinsonian subjects performed two simultaneous tasks such as walking and turning or walking and carrying a tray. Our own research (Morris et al., Submitted) has recently demonstrated that when people with PD focus

84 M.E. Morris et al. attention on secondary cognitive tasks during gait following training with visual cues or attentional strategies, both the stride length and the velocity of walking deteriorate. However when the secondary tasks are removed, the improvements in gait achieved with rehabilitation are again observed. These findings highlight that PD patients should be instructed to perform one motor task at a time where possible. For example, they should avoid talking when walking or balancing a tray of drinks when walking, because when the focus of attention is on the secondary task the gait pattern and balance will probably deteriorate. The patient should also be informed that when a secondary task is unavoidable, the short stepped, shuffling gait pattern might re-emerge and they are probably at greater risk of falling. Therefore compensatory strategies should be adopted accordingly. 5. Parkinsonian hypokinesia is accentuated in the performance of long or complex movement sequences. Hypokinesia is accentuated when people with PD perform long or complex movement sequences. For example, in Parkinsonian handwriting the amplitude and speed of movement progressively diminish as the sequence proceeds (Phillips et al., 1993). Similarly, people with gait festination walk with footsteps that become progressively shorter as the length of the sequence is increased (Martin, 1967; Parkinson, 1817). Clinical observations suggest that when submovements such as turning or sitting down and getting up from a chair are added into the gait task, both hypokinesia and akinesia are accentuated. To counter this problem, strategies can be learned which break down gait sequences into small subcomponents (walk, then turn, then sit) and which keep the task as simple as possible. Furthermore, by consciously focussing on each subcomponent as it is performed the walking pattern can be optimized. Patients are encouraged to practise these strategies until the new way of performing the movement is acquired. The literature highlights that people with PD do have the potential to learn new movement strategies although the amount of practice needed for skill acquisition is considerably more than for age matched controls (Soliveri, Brown, Jahanshahi & Marsden, 1992; Robertson & Flowers, 1990; Roy, Saint-Cyr, Taylor & Lang, 1993). The literature also indicates that procedural memory is affected by PD, possibly because the cortico-striatal pathways are disrupted, even though declarative memory, which is predominantly under the control of the frontal cortex, remains relatively unaffected (Roy et al., 1993). Therefore in the clinical setting patients might be able to rote learn basic strategies for movement more easily than attempting to learn long or complex movement sequences. As yet the literature provides little insight into the optimal frequency and scheduling of practice for motor skill learning in PD, hence

Motor Control Considerations in Parkinson's Disease 85 therapists need to ascertain this by adopting an experimental approach to clinical intervention. 7. CONCLUSION The empirical literature reviewed in this chapter indicates that traditional methods of gait rehabilitation for PD based on the treatment of rigidity and tremor are now difficult to justify for two reasons. First, the results of clinical trials have failed to demonstrate that traditional methods are effective in overcoming gait disorders in PD. Second, examination of contemporary evidence on the motor functions of the BG indicates that the key movement disorders in PD are hypokinesia and akinesia and that rigidity and tremor have lesser impact on functional movement. From this perspective the poor results from traditional methods are not surprising. This contemporary perspective can also be utilized to generate hypotheses for new strategies in gait rehabilitation. Our analyses indicated that people with PD should be able to activate a normal stepping pattern given the appropriate conditions, and evidence was presented to support that prediction. In addition, balance disorders, hypokinesia and akinesia appear amenable to therapeutic intervention by strategies that involve cognitive preparation in the premovement period and direction of attention towards specific aspects of movement during its execution. The ability to adapt locomotion can also be enhanced by judicious use of external cues. In contrast, gait disorders are likely to be accentuated when walking is performed simultaneously with other motor or cognitive tasks. Similarly, hypokinesia and akinesia are highlighted when walking occurs as part of a long or complex movement sequence. Thus consideration of the contemporary motor control and clinical rehabilitation literature suggests that it is possible to bridge the gap between theory and practice. However, at this time, empirical validation of these plausible hypotheses remains a major challenge to clinicians and motor control theorists alike. REFERENCES Alexander, G.E., & Crutcher, M.D. (1990). Functional architecture of BG circuits: Neural substrates of parallel processing. TINS, 13, 266-271. Baev, K.V., & Shimansky, Y.P. (1992). Principles of organization of neural systems controlling automatic movements in animals. Progress in Neurobiology, 39, 45-112. Bagley, S., Kelly, B., Tunnicliffe, N., Turnbull, G., & Walker, J. (1991). The effect of visual cues on the gait of independently mobile Parkinson's disease patients. Physiotherapy, 77, 415-420.

86 M.E. Morriset al. Ball, J.M. (1967). Demonstration of the traditional approach in the treatment of a patient with Parkinsonism. American Journal of Physical Medicine, 46, 1034-1036. Banks, M.A., & Caird, F.I. (1989). Physiotherapy benefits patients with Parkinson's disease. ClinicalRehabilitation, 3, 11-16. Benecke, R., Rothwell, J.C., Dick, J.P.R., Day, B.L., & Marsden, C.D. (1986). Performance of simultaneous movements in patients with Parkinson's disease. Brain, 109, 39-757. Benecke, R., Rothwell, J.C., Dick, J.P.R., Day, B.L., & Marsden, C.D. (1987). Disturbance of sequential movements in patients with Parkinson's disease. Brain, 110, 361-379. Bernheimer, H., Birkmayer, W., Homykiewicz, K., Jellinger, K., & Setelberger, F. (1973). Brain dopamine and the syndromes of Parkinson and Huntington. Journal of Neurological Science, 20, 415-455. Biller, J., & Brazis, P.W. (1990). The localization of lesions affecting the basal ganglia. In P.W. Brazis, C.J. Masdeau & J. Biller (Eds.) Localization in clinical neurology, (2nd ed.) (pp.345-360). Boston: Little, Brown and Company. Bilowit, D.S. (1956). Establishing physical objectives in rehabilitation of patients with Parkinson's disease (gymnasium activities). Physical Therapy Review, 36, 176-178. Blin, O., Ferrandez, A.M., Pailhouse, J., & Serratrice, G. (1991). Dopa-sensitive and dopa-resistant gait parameters in Parkinson's disease. Journal of Neurological Science, 103, 51-54. Blin, O., Ferrandez, A.M., & Serratrice, G. (1990). Quantitative analysis of gait in Parkinsonian patients: Increased variability of stride length. Journal of Neurological Science, 98, 91-97. Bonnard, M., & Pailhouse, J. (1993). Intentionality in human gait control: Modifying the frequency to amplitude relationship. Journal of Experimental Psychology: Human Perception and Performance, 19, 429-443. Bradshaw, J.L., Phillips, J.G., Dennis, C., Mattingley, J.B., Andrewes, D., Chiu, E., Pierson, J.M., & Bradshaw, J.A. (1992). Initiation and execution of movement sequences in those suffering from and at-risk of developing Huntington's disease. Journal of Clinical and Experimental Neuropsychology, 14, 179-192. Brooks, D.J., Salmon, E.P., Mathias, C.J., Quinn, N., Leenders, K.L., Bannister, R., Marsden, C.D., & Frackowiak, R.S.J. (1990). The relationship between locomotor disability, autonomic dysfunction, and the integrity of the striatal dopaminergic system in patients with multiple system atrophy, pure autonomic failure, and Parkinson's disease, studies with PET. Brain, 113, 1539-1552.

Motor Control Considerationsin Parkinson'sDisease 87 Brotchie, P., Iansek, R., & Home, M.K. (1991a). Motor function of the monkey globus pallidus. 1. Neuronal discharge and parameters of movement. Brain, 114, 1667- 1683. Brotchie, P., Iansek., R, & Home, M.K. (1991b). Motor function of the monkey globus pallidus. 2. Cognitive aspects of movement and phasic neuronal activity. Brain, 114, 1685-1702. Brown, R.G., & Marsden, C.D. (1988). Internal versus external cues and the control of attention in Parkinson's disease. Brain, 111,323-345. Bussel, B., Robi-Bramby, A., Azouvi, P.H., Birab, A., Yakovleff, A., & Held, J.P., (1988). Myoclonus in a patient with spinal cord transection. Brain, 111, 1235- 1245. Calne, D.B., Larsen, T.A., & Burton, K. (1985). Posture and gait in Parkinson's disease. In P.J. Delwaide & A. Agnoli (F_As), Clinical neurophysiology in Parkinsonism (pp. 133-138). Amsterdam: Elsevier. Cappozzo, A. (1981). Analysis of the linear displacement of the heel and trunk during walking at different speeds. Journal of Biomechanics, 14, 411-425. Clark, J.E. (1994). Dynamical systems perspective on gait. In R.L. Craik & C.A. Oatis (Eds.), Gait analysis: Theory and application. (pp. 2-19). New York: CV Mosby. Collins, J.J., & Stewart, I.N. (1992). Symmetry-breaking bifurcation: A possible mechanism for 2:1 frequency-locking in animal locomotion. Journal of Mathematical Biology, 30, 827-838. Collins, J.J., & Stewart, I. (1993). Hexapodal gaits and coupled nonlinear oscillator models. Biological Cybernetics, 68, 287-298. Cunnington, R., Iansek, R., Bradshaw, J., & Phillips, J. (in press). Movement related potentials in Parkinson's disease: Presence and predictability of temporal and spatial cues. Brain. Davis, J.C. (1977). Team management of Parkinson's disease. American Journal of Occupational Therapy, 31,300-308. De Caballos, M., Ferrandez, A., Jenner, P., & Marsden, C.D. (1993). Parallel alterations in Met-enkephalin and substance P levels in medial globus pallidus in Parkinson's disease patients. Neuroscience Letters, 160, 167-170. De Long, M.R. (1990). Primate models of movement disorders of basal ganglia origin. TINS, 13,281-285. Diem, V., Berger, W., & Horstmann, G.A. (1988). Posture in Parkinson's disease: Impairment of reflexes and programming. Annals of Neurology, 24, 660-669. Doshay, L.J. (1962). Current concepts in therapy. New England Journal of Medicine, 266, 878-880.

88 M.E. Morriset al. Dunne, J.W., Hankey, G.J., & Edis, R.H.. (1987). Parkinsonism: Upturned walking stick as an aid to locomotion. Archives of Physical Medicine and Rehabilitation, 68, 380-381. Formisano, R., Pratesi, L., ModareUi, F.T., Bonifati, V., & Meco, G. (1992). Rehabilitation and Parkinson's disease. Scandinavian Journal of Rehabilitation Medicine, 24, 157-160. Forssberg, H., Johnels, B., & Steg, S. (1984). Is Parkinsonian gait caused by a regression to an immature walking pattern? Advances in Neurology, 40, 375-379. Forssberg, H., Gfillner, S., Halbertsma, J., & Rossignol, S. (1980). The locomotion of the low spinal cat. 2: Interlimb coordination. Acta Physiologica Scandinavia, 108, 283-295. Franklyn, S. (1983). The value of physiotherapy for Parkinson's disease. International Journal of Rehabilitation Research, 6, 271. Georgiou, N., Bradshaw, J., Iansek, R., Phillips, J.G., Mattingley, J.B., & Bradshaw, J.A. (1994). Reduction in external cues and movement sequencing in Parkinson's disease. Journal of Neurology, Neurosurgery and Psychiatry, 57, 368-370. Georgiou, N., Iansek, R., Bradshaw, J., Phillips, J.G., Mattingley, J.B., & Bradshaw, J.A. (1993). An evaluation of the role of internal cues in the pathogenesis of Parkinsonian hypokinesia. Brain, 116, 1575-1587. Gibberd, F.B., Page, N.G.R., Spencer, K.M., Kinnear, E., & Hawksworth, J.B. (1981). Controlled trial of physiotherapy and occupational therapy for Parkinson's disease. British Medical Journal, 282, 1196. Giladi, N., McMahon, D., Przedborski, S., Flaster, E., Guillory, S., Kostic, V., & Fahn, S. (1992). Motor blocks in Parkinson's disease. Neurology, 42,333-339. Grillner, S. (1975). Locomotion in vertebrates: Central mechanisms and reflex interaction. Physiological Review, 55, 247-304. Hallett, M. (1993). Parkinson's disease: Slowness of movement. Proceedings Motor Control (#145). American Academy of Neurology, New York (pp. 145-169). Hallett, M., & Koshbin, S. (1980). A physiological mechanism of bradykinesia. Brain, 103, 301-314. Homberg, V. (1993). Motor training in the therapy of Parkinson's disease. Neurology, 43 (supp 6), $45-$46. Hurwitz, A. (1989). The benefit of a home exercise regimen for ambulatory Parkinson's disease patients. Journal of Neuroscience Nursing, 21,180-184. Inman, V.T., Ralston, H.J., & Todd, F. (1981). Human walking. Baltimore, MD: Williams & Wilkins.

Motor Control Considerationsin Parkinson's Disease 89 Kay, B.A., Saltsman, E.L. & Kelso, J.A.S. (1991). Steady-state and perturbed rhythmical movements: A dynamical analysis. Journal of Experimental Psychology: Human Perception and Performance, 17, 183-198. Knott, M. (1957). Report of case of Parkinsonism treated with proprioceptive facilitation techniques. Physical Therapy Review, 37, 229. Knutsson, E. (1972). An analysis of Parkinsonian gait. Brain, 95, 475-486. Knutsson, E., & Martensson, A. (1971). Quantitative effects of L-Dopa on different types of movements and muscle tone in Parkinsonian patients. Scandinavian Journal of Rehabilitation Medicine, 3, 121-130. Koller, W.C., Glatt, S., Vetere-Overfield, B., & Hassanein, R. (1989). Falls and Parkinson's disease. Clinical Neuropharmocology, 12, 98-105. Kugler, P.N., & Turvey, M.T. (1987). Information, natural law and the self-assembly of rhythmic movement. Erlbaum, HiUsdale, New Jersey. Laurent, M., & Pailhouse, M. (1986). A note on modulation of gait in man: Effects of constraining stride length and frequency. Human Movement Science, 5, 333-343. Lee, D.N., & Young, D.S. (1986). Visual timing of interceptive action. In D. Ingle, M. Jeannerod, & D.N. Lee (Eds.) Brain mechanisms and spatial vision (pp.l-30). Amsterdam: Dordrecht, Markinus Nijhoff. Lee, D.N., Lishman, J.R. & Thomson, J.A. (1982). Visual regulation of gait in long jumping. Journal Experimental Psychology: Human Perception and Performance, 8, 448-459. Marsden, C.D. (1990). Neurophysiology. In G.E. Stern (Ed.), Parkinson's disease (pp.57-97). Baltimore: John Hopkins University Press. Marsden, C.D., & Obeso, J.A. (1994). The functions of the basal ganglia and the paradox of stereotaxic surgery in Parkinson's disease. Brain, 117, 877-897. Marsden, C.D., Parkes, J.D., & Quinn, N. (1982). Fluctuations of disability in Parkinson's disease - clinical aspects. In C.D. Marsden & S. Fahn (Eds.), Movement disorders (pp. 96-122). London: Butterworth Scientific. Martin, J.P. (1967). The basal ganglia and posture. London: Pitman Medical. Martin, K.E., Phillips, J.G., Iansek, R., & Bradshaw, J.L. (1994). Inaccuracy and instability of sequential movements in Parkinson's disease. Experimental Brain Research, 102, 131-140. Mc Goon, D.C. (1990). The Parkinson's Handbook. New York: WW Norton. Minnigh, S.C. (1971). The Northwestern University concept of rehabilitation through group physical therapy. Rehabilitation Literature, 32, 38,39,50. Morris, M.E., Iansek, R., Matyas, T., & Summers, J.J. (1994a). The pathogenesis of gait hypokinesia in Parkinson' s disease. Brain, 117, 1169-1181.

90 M.E. Morris et al. Morris, M.E., Iansek, R., Matyas, T., & Summers, J.J. (1994b). The ability to modulate walking cadence remains intact in Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry, 57, 1532-1534. Morris, M.E., Iansek, R., Matyas, T., & Summers, J.J. A comparison of the effects of training using visual cues and attentional strategies on the maintenance of normal stride length in Parkinson's disease. Manuscript submitted for publication. Morris, M.E., Matyas, T., Bach, T., & Goldie, P.A. (1992). Electrogoniometric feedback: Its effect on knee hyperextension in stroke. Archives of Physical Medicine and Rehabilitation, 73, 1147-1154 Murray, P., Sepic, S.B., Gardner, G.M., & Downs, W.J. (1978). Walking patterns of men with Parkinsonism. American Journal of Physical Medicine, 60, 278-294 Nashner, L. (1980). Balance adjustments of humans perturbed while walking. Journal of Neurophysiology, 44, 650-664. Nashner, L. (1982). Adaptations of human movement to altered environments. TINS, 3, 358-361. Newell, K.M. (1991). Motor skill acquisition. Annual Review of Psychology, 42, 213- 37. Nilsson, J., & Thorstensson, A. (1987). Adaptability in frequency and amplitude of leg movements during human locomotion at different speeds. Acta Physiologica Scandinavica, 129, 107-114. Nutt, J.G., Marsden, C.D. & Thompson, P.D. (1993). Human walking and higher-level gait disorders, particularly in the elderly. Neurology, 43, 268-279. Palmer, S.S., Mortimer, J.A., Webster, D.D., Bistevins, R., & Dickinson, G.L. (1986). Exercise therapy for Parkinson's disease. Archives of Physical Medicine & Rehabilitation, 67, 741-745. Parent, A., & Hazrati, L.N. (1993). Anatomical aspects of information processing in primate basal ganglia. TINS, 16, 111-116. Parkinson, J. (1817). An essay on the shaking palsy. Reprinted in Medical Classics, 2, (pp. 964-997), 1938. Patla, A. (1991a). Understanding the control of human locomotion: A Prologue. In A.E. Patla (Ed.) Adaptability of human gait: Implications for the control of locomotion (pp. 3-20). Amsterdam: North-Holland. Patla, A. (1991b). Understanding the control of human locomotion: A \"Janus\" Perspective. In A.E. Patla (Ed.) Adaptability of human gait: Implications for the control of locomotion (pp.441-452). Amsterdam: North-Holland.

Motor Control Considerationsin Parkinson'sDisease 91 Patla, A.E., Robinson, C., Samways, M., & Armstrong, C.J. (1989) Visual control of step length during overground locomotion: Task: specific modulation of the locomotor synergy. Journal of Experimental Psychology: Human Perception and Performance, 15, 603-617. Pedersen, S.W., Oberg, B., Insulander, A., & Vretman, M. (1990). Group training in Parkinsonism: Quantitative measurements of treatment. Scandinavian Journal of Rehabilitation Medicine, 22, 207-211. Phillips, J.G., Bradshaw, J.L., Iansek, R., & Chiu, E. (1993). Motor functions of the basal ganglia. Psychological Research, 55, 175-181. Quintyn, M., & Cross, E. (1986). Factors affecting the ability to initiate movement in Parkinson's disease. Physical and Occupational Therapy in Geriatrics, 6, 51-60. Robertson, C., & Flowers, K.A. (1990). Motor set in Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry, 53, 583-592. Roland, P.E., Larsen, B., Lassen, N.A., & Skinhoj, E. (1980). Supplementary motor area and other cortical areas in organization of voluntary movements in man. Journal of Neurophysiology, 43, 118-136. Roland, P.E., Meyer, E., Shibasaki, T., Yamamoto, Y.L., & Thompson, C.J. (1982). Regional cerebral blood flow changes in cortex and basal ganglia during voluntary movements in normal human volunteers. Journal of Neurophysiology, 48, 467-480. Roy, E.A., Saint-Cyr, J., Taylor, A. & Lang, A. (1993). Movement sequencing disorders in Parkinson's disease. International Journal of Neuroscience, 73, 183-194. Schell, G.R., & Strick, P.L. (1984). The origin of thalamic inputs to the arcuate premotor and supplementary motor areas. Journal of Neuroscience, 4, 539-560. Selverston, A.I. (1980) Are central pattern generators understandable? Behavioral and Brain Sciences, 3, 535-571. Shik, M.L., Orlovskii, G.N., & Severin, F.V. (1968). Locomotion of the mesencephalic cat elicited by stimulation of the pyramids. Biofizika, 13, 127-135. Shoenberg, B.S. (1988). Epidemiology of movement disorders. In C.D. Marsden & S. Fahn (Eds). Movement disorders, 2. (pp.17-32). London: Butterworths. Sietz, R.J., & Roland, P.E. (1992). Learning of sequential finger movements in man: A combined kinematic and positron emission tomography (PET) study. European Journal of Neurosciences, 4, 154-165. Smithson, F., Morris, M.E., & Iansek, R. (1994). Predisposing factors for falls in Parkinson's disease. Proceedings 29th National Conference Australian Association Gerontology, Perth, Western Australia. Soliveri, P., Brown, R.G., Jahanshahi, M. & Marsden, C.D. (1992). Effect of practice on performance of a skilled motor task in patients with Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry, 55, 454-460.

92 M.E. Morriset al. Stem, G.M., Lander, C.M., & Lees, A.J. (1980). Akinetic freezing and trick movements in Parkinson's disease. Journal of Neural Transmission, Suppl. 16, 137-141. Stem, G.M., Franklyn, S.E., Imms, F.J., & Prestidge, S.P. (1983). Quantitative assessments of gait and mobility in Parkinsons disease. Journal of Neural Transmission, Suppl. 19, 201- 214. Swinnen, S.P., Massion, J., & Heuer, H. (1994). Topics on interlimb coordination. In S. Swinnen, J. Massion, & H. Heuer (Eds.) Interlimb coordination: Neural dynamical and cognitive constraints (pp. 1-27). San Diego: Academic Press. Szekely, B.C., Kosanovich, N.N., & Sheppard, W. (1982). Adjunctive treatment in Parkinson's disease: Physical Therapy and comprehensive group therapy. Rehabilitation Literature, 43, 72-76. TaUand, G.A. & Schwab, R.S. (1964). Performance with multiple sets in Parkinson's disease. Neuropsychologia, 2, 45-53. Tanji, J., & Kurata, K. (1985). Contrasting neuronal activity in supplementary and precentral motor cortex of monkeys. 1. Response to instructions determining motor responses to forthcoming signals of different modalities. Journal of Neurophysiology, 53, 129-141. Thelen, E. (1986). Development of coordinated movement: Implications for early human development. In M.G. Wade & H.T.A. Whiting (Eds.) Motor development in children: Aspects of coordination and control (pp 107-124). Boston: Martinus Nijhoff. Thelen, E., Ulrich, B.D. & Miles, D. (1987). Bilateral coordination in human infants: Stepping on a split-belt treadmill. Journal of Experimental Psychology: Human Perception and Performance, 13, 405-410. Tokuno, H., Kimura, M., & Tanji, J. (1992). PaUidal inputs to thalamocortical neurons projecting to the supplementary motor area: An anterograde and retrograde double labelling study in the macaque monkey. Experimental Brain Research, 90, 635-638. Turvey, M.T. (1990). Coordination. American Psychologist, 45, 938-953. Umphred, D.A. (1985). Neurological Rehabilitation. St Louise: Mosby. Warren, W.H., Young, D.S. & Lee, D.N. (1986). Visual control of step length during running over irregular terrain. Journal of Experimental Psychology: Human Perception and Performance, 12, 259-266. Weissenborn, S. (1993). The effect of using a two-step verbal cue to a visual target above eye level on the Parkinsonian gait: A case study. Physiotherapy, 79, 26-31. WhitaU, J., & Clark, J.E. (1994). The development of bipedal interlimb coordination. In S.P. Swinnen, J. Massion & H. Heuer (Ed). lnterlimb coordination: Neural dynamical and cognitive constraints (pp. 395-412). San Diego: Academic Press.

Motor Control Considerations in Parkinson's Disease 93 Wilson, D.M. (1961). The central nervous control of flight in a locust. Journal of Experimental Biology, 38, 471-479. Winter, D.A. (1980). Overall principal of lower limb support during stance phase of gait. Journal of Biomechanics, 13, 923-927. Winter, D.A. (1992). The biomechanics and motor control of human gait: Normal, pathological and elderly. Ontario, Canada: University of Waterloo Press. Wroe, M., & Greer, M. (1973). Parkinson's disease and physical therapy managemenL Physical Therapy, 53, 849-855. Yekutiel, M.P., Pinhasov, A., Shahar, G., & Sroka, H. (1991). A clinical trial of the re- education of movement in patients with Parkinson's disease. Clinical Rehabilitation, 5, 207-214.

Motor Control and Sensory Motor Integration: Issues and Directions 97 D.J. Glencross and J.P. Piek (Editors) 9 1995 Elsevier Science B.V. All rights reserved. Chapter 5 ADAPTIVE OPTIMAL CONTROL OF HUMAN TRACKING Peter D Neilson, Megan D Neilson, & Nicholas J O'Dwyer Cerebral Palsy Research Unit, Institute of Neurological Sciences, The Prince Henry Hospital and School of Electrical Engineering University of New South Wales The motor behaviour of subjects performing visual tracking tasks is quantified by identifying the mathematical relationship between the visual information presented to the eye and the resulting motor response generated at the hand. It has long been known that this relationship is equivocal and that no unique mathematical model exists to describe the behaviour of the human operator. In whatfollows we develop the hypothesis that tracking behaviour is variable because the central nervous system (CNS) functions as an adaptive optimal controller of muscles, biomechanics and external systems. It automatically tunes its input-output relationship to compensate for the dynamics of the system being controlled and to compensate for inherent time delays by predicting future values of the input signals. We explore the proposal that the CNS plans motor responses to achieve goals using a minimum of input muscular energy and that it can trade tracking accuracy against demand for input energy by altering the speed of the response. Hypotheses about information processing performed by the CNS during visual tracking are presented in the form of a computer simulation. Distributed parallel processing circuitry is employed in the simulator to construct adaptive digital filters which operate independently and in parallel These digital filters mimic the behaviour of hypothesized neural adaptive filters within the CNS. Indeed in general descriptions of the simulator can be taken as hypotheses about the structure and function of neural circuitry and about the information processing performed by the CNS during control of movement. As with any scientific theory, the hypotheses are tested experimentally by comparing the behaviour of the simulator with that of human subjects performing the same task. A summary of key findings from a number of studies of human tracking behaviour carried out at our laboratory is presented and many of the findings are compared with the behaviour of the simulator.

98 P.D. NeUson, M.D. NeUson and N.J. O'Dv~er 1. I N T R O D U C T I O N Among the many experimental paradigms used in the study of human movement, tracking holds a distinguished and historical place. The understanding of tracking performance became crucially important in a variety of applications in World War II and it was there that the first attempts were made to specify a subject's behaviour in the language of systems control theory. The potential for tracking to shed light on sensory-motor mechanisms was recognized in those early times by none other than Kenneth Craik writing circa 1943: [Tracking] is a form of coordinated sensory-motor reaction capable of all degrees of complexity, blending at the one end, in the simplest reactions, into something very near a conditioned reflex, and at the other into the application of the most complicated skills and habits, in which anticipation, prediction, grasping a problem, and calculation of the future may be involved. It can be tackled in a form in which physiological and also engineering (servo-motor) terminology is appropriate and suggestive. Thus the datum to which the operator responds is usually described as a 'misalignment' between target and graticule, to which he makes a control movement; as Hick has pointed out, the misalignment acts as a stimulus, evoking a response from the operator...The study of the response to serially presented stimuli may throw some light on the nature and duration of the 'central delay' in reaction times and has interesting analogies with refractory periods in rhythmic reflexes; while the ability of the operator to compensate for the limited 'response rate' of this central mechanism by the appreciation of patterns of groups of stimuli as wholes, and the formulation of complex motor responses or unitary response,patterns to deal with them suggests an interesting field for investigation of sensory and motor integration... (Craik, 1943/1966, pp.47-48) Despite this early systems orientation of Craik and others to tracking, many later skills oriented psychological studies abandoned the determination of transfer characteristics of the human operator in favour of simpler measures of performance such as overall error measures. Poulton (1974) and Hammerton (1981) both reflect disillusionment with the approach on the basis that there is no unique description of the human operator, that models can be overfitted and assumptions of the methods are often not justified. Moray (1981) takes a different view along with other engineering oriented contributors to the psychological literature such as Pew (1974), Jagacinski (1977) and Wickens and Gopher (1977).