Human Movement North-Holland

37

Science 11 (1992) 37-45

Response initiation delays in Parkinson’s disease patients * George Arizona

E. Stelmach,

State lJniuersit_y. Tempe,

Normand

Teasdale

and Jim Phillips

USA

Abstract Stelmach, G.E., N. Teasdale and J. Phillips, 1992. Response disease patients. Human Movement Science 11, 37-45.

initiation

delays

in Parkinson’s

Parkinson’s disease (PD) patients, elderly, and young subJects generated force pulse at percentages of their maximum. PD patients were substantially slower in initiating an isometric force at each of the target force levels. The reaction delay was localized in the pre-motor component, suggesting that the longer response initiation time observed in PD patients is not associated with excessive delays in motor processes but rather with delays in cognitive processes that precede the response output. Although PD patients were as accurate as elderly and young subjects in producing the required force levels, they showed more irregular force-time patterns.

Baroni et al. (1984) and Berardelli et al. (1986) have shown that PD patients have difficulties producing necessary amounts of force and are inaccurate in producing an exact specification of agonist activity. For example, Hallett and Khoshbin (1980) have shown that PD patients do not deliver the correct force command to the agonist muscle in a rapid elbow flexion movement; rather than the simple cycle of agonistantagonist-agonist activity seen in normals, PD patients exhibited a series of alternating bursts. Experiments in our laboratory have also showed that PDs exhibit problems controlling forces. Stelmach et al. (1986) asked PD patients to perform an isometric force production task at a variety of target force * This research was supported by NINDS grant NS 17421. N. Teasdale is now at Universite Laval, Laboratoire de Performance Motrice Humaine, Canada. J. Phillips is now at the Dept. of Psychology, Monash University, Victoria, Australia. Requests for reprints should be sent to G.E. Stelmach, Psychobiology Section, Exercise and Sport Science Institute, Arizona State University, Tempe, AZ 85287. USA.

0167-9457/92/$05.00

0 1992 - Elsevier Science Publishers

B.V. All rights reserved

38

G. E. Stelmach et al. / Force control in Parkinson’s

disease

levels. PD patients exhibited delayed force production and irregular force-time patterns and were especially disadvantaged when producing low levels of force. Stelmach et al. (1987) found that PD patients also had difficulties producing appropriate forces in finger-tapping sequences; they showed an abnormal prolongation of the inter-tap interval in a repetitive sequence. Subsequently, Stelmach et al. (1989) have shown that PD patients produce more irregular force time curves and have slower rates of force production. These results were interpreted as evidence that PD patients have an inability to quickly produce smooth forces. The existing database suggests that PD patients have deficiencies in the initiation, production and regulation of force. The current experiment was designed to extend the current data base by probing the locus of the response initiation delays in rapid isometric force impulses. It considered whether PD patients could generate and deliver effective rapid force impulses and whether movement delays are located in pre-motor or in motor processes.

Method Subjects

Seven patients with idiopathic PD (mean elderly control subjects (mean age 67.1 years), jects (mean age 24.6 years) were examined. normal medication schedule on the day of the was timed for their end-of-dose cycle as much Apparatus

age 65.7 years), seven and seven young subPatients followed their experiment, but testing as possible.

and subject position

The apparatus consisted of a force transducer (Interface SSM-500) attached to a rigid, wall-mounted shelf. The transducer output was directed to an amplifyling circuit and then to a PDP 1 l/73 microcomputer. Muscle discharge patterns (EMG) from the biceps brachii and triceps lateralis were recorded using physiological amplifiers placed over the muscle bellies. The signals were pre-amplified at the source, full-wave rectified, band-pass limited from 40 Hz to 4 KHz, and

G. E. Stelmach et al. / Force control in Parkinson’s disease

39

filtered with a time constant of 2.5 ms. All signals were digitized at 500 Hz. The upper arm and forearm of the subject rested on the padded surface of the shelf with the elbow flexed in the horizontal plane at an angle of approximately ninety degrees. The palmar surface of the wrist made contact with a plastic plate attached to the transducer. PD patients were tested on their more affected side, and elderly and young subjects were tested on their non-dominant side. Procedure Subjects produced five maximum isometric contractions at the presentation of a visual stimulus. Average peak force was calculated from those trials. In the second phase of the experiment, subjects were required to produce a force level of either 15, 30, 45, or 60% of their own maximum. Each trial started with a ready signal (four LEDs were simultaneously activated for 1 set). A second later, the LED cueing the required force level was activated for one second. The same LED served as an imperative signal and was reactivated a second later, signalling the subjects to initiate their responses. There were eight randomly presented blocks of experimental trials (ten trials per block at the same force level). There were two blocks of trials for each target force level. Data analysis The force-time curves were smoothed with a Butterworth secondorder filter with dual pass to remove any phase lag (10 Hz cut-off frequency). The onset and the peak amplitude of each trace were determined through interactive graphics. The EMG data were smoothed using a moving window average (10 ms window) and were also analyzed using interactive graphics methods. Reaction time was decomposed into premotor and motor components by measuring the EMG onset and substracting it from the reaction time. Premotor time was defined as the interval between stimulus onset and EMG onset. Motor time was defined as the interval between EMG onset and response initiation. The results for the dependent variables were submitted to a group by

40

G. E. Stelmach et al. / Force control in Parkinson’s

target force level analysis of variance sures on the second factor.

(ANOVA)

disease

with repeated

mea-

Results

Response initiation On average, PD patients were slower than elderly and young subjects (470, 329, and 281 ms, respectively). The main effect of group was significant (F(2,18) = 8.21, p < 0.01); a comparison of means showed that PD patients were slower than both elderly and young subjects ( ps < 0.05). For the three groups, reaction time (RT) decreased with an increased target force level (380, 349, 359, and 352 ms for the 15%, 30%, 45%, and 60% target force levels, respectively). The main effect of target force was significant (F(1,18) = 4.19, p < 0.05). A comparison of means showed that the RT at the 15% target force level was longer than at the three other force levels (p < 0.05). Further, the response latenties of PD patients were on average more than twice as variable (within-subject standard deviation) as those of the elderly and young subjects (157, 72, and 50 ms, respectively). The main effect of group was significant (F(2,18) = 7,76, p -C 0.01) and a comparison of means showed that PD patients were more variable than elderly and young subjects (ps < 0.05). To further explore the pronounced lenghtening of RT observed in PD patients, RT was decomposed into pre-motor and motor components. The motor component of RT was similar for the three groups and was unaffected by the target force level (ps < 0.05). The motor RT for the 15% and 30% target force, was on average 68 ms for the PD patients, 74 ms for the elderly, and 61 ms for the young. In contrast, PD patients had substantially longer pre-motor RT (448 ms) than both elderly (253 ms) and young subjects (230 ms), where the main effect of group was significant (F(2,16) = 6.28, p < 0.01). Force characteristics A more elaborate description of the force data is available in Stelmach et al. (1989). Table 1 presents some characteristics of the produced forces. On average, the absolute maximum peak force was

67 173 99 61 43 88 72

1 2 3 4 5 6 7

86 115

130

PD Elderly

Young

MeaIlS

Peak force (N)

337

521 295

310 535 286 603 442 1044 430

15%

699 421 400

392

575 601 378 740 971 1109 524

548 513 271 624 882 1144 403

626 340

45%

and young

376

21

19 17 38

35 35

19 51 32 18 24 28 25

710 564 487 750 952 1430 580

9 35 17 11 11 19 12

30%

791 497

subjects.

55

50 52

33 89 50 30 22 40 68

45%

Percent of peak force

elderly,

60%

for PD patients,

30%

(ms)

parameters

Time-to-peak

results of force production

PD patient

Table 1 Summary

69

63 65

51 113 62 31 28 51 47

60%

176

105

28

58 200

62

65 59 18 25 18

35 87

35 111 118 29 27 24

29

30%

15%

241

12 276

72

148 132 40 23 36

57

45%

Rate of force development

306

84 290

81

200 127 41 29 36

72

60%

(N/s)

e

2

a 8

z h”

g

2 3 a S. P S b

2 R \

42

G. E. Stelmach et al. / Force control in Parkinson ‘s disease

smaller for PD patients (86 N) than for the elderly and young subjects (115 N and 130 N, respectively). However, PD patients were able to produce peak forces that approximated the different target force levels as there were no group differences in the percent of peak force produced (F(2,18) = 1.71, p > 0.05). PD patients took nearly twice as much time (657 ms) to achieve peak force as elderly (388 ms) and young subjects (376 ms). The ANOVA yielded a significant interaction of group by target force level (F(6,54) = 2.66, p < 0.05). A decomposition of the interaction into its orthogonal components showed that the slopes differed in their linear components (p -c 0.05). This implies that, across force levels, the time-to-peak force was relatively constant for the young subjects, whereas it systematically increased with an increase in the required target force for the PD patients (from 521 ms to 781 ms) and the elderly (from 295 ms to 497 ms). Due to smaller peak forces and longer time-to-peak force, PD patients also had lower average rate of force development (ratio of mean peak force to mean time-to-peak force) than elderly and young subjects.

Discussion

The results showed that PD patients had longer RTs than both elderly and young subjects. More importantly, the longer RT was almost totally accounted for by delays in the pre-motor RT (on average, 98%). Here the slowness in initiation observed in PD patients is associated with delays in pre-motor processes and not with excessive delays in execution. In an earlier study, Stelmach et al. (1986) found that a PD group had a higher RT intercept but not a steeper slope with an increase in the number of alternatives from 1 to 8. Such a result implies a delay in ‘ input’ and/or ‘output processes, as opposed to more cognitive processes (e.g. response preparation) because the former are affected by the number of choices (Rosenbaum 1980). As suggested by many authors (Marsden 1982, 1984; Sheridan et al. 1987; Stelmach et al. 1989), the general form of the motor program seems intact in PD. These results suggest that the basal ganglia may have a more critical role in the initiation and execution of a motor program or a sequence of motor programs. Alternatively, Stern and Mayeux (1986) have

G. E. Stelmach et al. / Force control VI Parkinson’s

disease

43

focused upon sensory and attentional mechanisms, suggesting that PD patients have problems developing the correct preparatory set and are more reliant upon sensory cues. Moreover, Evarts and Wise (1984) considered additional brain structures, such as preparatory set-related cells located in the supplementary motor area, that may be responsible for longer reaction times observed in PD patients. Problems with the preparatory set observed in that experiment may also partially explain the slowness in the initiation of movement and of force production in the present experiment. Stelmach et al. (1986, 1989) suggested that the finding that PD patients can produce accurate peak forces in the absence of visual feedback argues against the generality of previous assertions that PD patients are critically dependent on such feedback. For all three groups, the relationship between force and force variability was a curvilinear, negatively accelerating function. Berardelli et al. (1986) and Hallett and Khoshbin (1980) found that PD patients have been shown to be deficient in the regulation of EMG activity. However, as Stelmach et al. (1986) previously reported, the force-time curves of PD patients varied from near normal to profoundly impaired. The force-time curves of elderly and young subjects were characterized by smooth initiation and regulation throughout the force production. In contrast, the force-time curves of PD patients were characterized by irregularities near the onset and throughout the force production. The elderly and young subjects showed discrete bursts of activity in the biceps and triceps muscles. The majority of contractions were characterized by a single burst of agonist (biceps) followed by a single burst of antagonist (triceps) activity. However, the EMG patterns of PD patients were often characterized by multiple bursts of activity. Berardelli et al. (1986) have shown that PD patients can adjust the amount of EMG activity in a wrist flexion movement with the amplitude of the movement. Their research suggests that EMG is inappropriately scaled to movement amplitude and velocity such that PD subjects show reduced gain. This argues against a simple ‘limited-energy hypothesis’. In our experiment, two PD patients also showed inappropriate activation of agonist muscles. For example fig. 1 has representative force-time curves and EMG activity for the biceps and triceps muscles for these two patients. At all four target force levels, a strong antagonist burst (triceps) was almost always present prior to or near the onset of force production. It is possible that the above

G.E. Stelmach et al. / Force control in Parkinson’s

0.0

0.5

1.5

1.0 TIME

(SIX)

2.0

2.5

3.0

0.0

0.5

1 .o

disease

1.5 TIME

2.0

2.5

3.0

(SAX)

Fig. 1, Representative force-time curves and their associated biceps and triceps EMG activity for two patients with Parkinson’s disease. In panel (a), a strong burst of triceps activity was observed near the onset of the force production. The patient showed similar force-time curves and associated EMG in more than 60% of the trials. In panel (b), irregular triceps activity was seen throughout the force production in every trial.

irregularity is an inability to inhibit antagonist muscles or an improper timing of agonists, antagonists and synergists. PD patients also took nearly twice as much time to achieve peak force as the elderly and young subjects. Because an increased time-topeak force produces a decreased force variability, PD patients may have decided to trade a fast rate of force production for more accurate force production. Such changes in response strategy have been observed extensively in elderly populations. Berardelli et al. (1986) reported that PD patients were unable to move faster when cued to do so. On the other hand, Hallett and Khoshbin (1980), though they did not report any movement time data, reported that PD patients moved faster on subsequent trials when required to do so. These data suggest that a portion of the slowness observed in PD patients may be the result of an emphasis on greater spatial accuracy. It is possible that PD patients, without any awareness of doing so, develop over several years strategies that emphasize spatial accuracy: in

G. E. Stelmach et al. / Force control in Parkinson’s

disease

45

everyday activities, the incentives to be spatially accurate far exceed the incentives to move fast. The increased latencies observed in PD patients may thus be the result of an altered movement strategy to cope with the disease itself as well as dopamine impairments.

References Baroni. A., F. Benvenutti, L. Fantini, T. Pantaleo and F. Urbani, 1984. Human ballistic arm abduction movements: Effects of L-dopa treatment in Parkinson’s disease. Neurology 34, 868-876. Berardelli, A., N. Accomero, M. Argenta, G. Meca and M. Manfredi, 1986. Fast complex arm movements in Parkinson’s disease. Journal of Neurology, Neurosurgery and Psychiatry 49, 114661149. Berardelli, A., J.P.R. Dick, J.C. Rothwell, B.L. Day and C.D. Marsden, 1986. Scaling of the size of the first agonist EMG burst during rapid wrist movements in patients with Parkinson’s disease. Journal of Neurology, Neurosurgery and Psychiatry 49, 1273-1279. Evarts, E.V. and S.P. Wise, 1984. Basal ganglia output and motor control. In: Ciba Foundation Symposium 107. Functions of the basal ganglia. London: Pitman. Hallett, M. and S. Khoshbin. 1980. A physiological mechanism of bradykinesia. Brain 103, 465-480. Marsden, C.D., 1982. The mysterious motor function of the basal ganglia: The Robert Wartenberg lecture. Neurology 32, 514-539. Marsden, C.D., 1984. Function of the basal ganglia as revealed by cognitive and motor disorders in Parkinson’s disease. Canadian Journal of Neurological Sciences 11, 129-135. Rosenbaum, D.A., 1980. Human movement initiation: Specification of arm distance, and extent. Journal of Experimental Psychology: General 109, 444474. Sheridan, M.R., K.A. Flowers and J. Huller, 1987. Programming and execution of movement in Parkinson’s disease. Brain 110, 1247-1271. Stelmach, G.E., N. Teasdale, J.G. Phillips and C.J. Worringham, 1989. Force production characteristics in Parkinson’s disease. Experimental Brain Research 76, 167-172. Stelmach, G.E., C.J. Worringham and E.A. Strand, 1986. Movement preparation in Parkinson’s disease: The use of advance information. Brain 109, 1179-1194. Stelmach, G.E., C.J. Worringham and E.A. Strand, 1987. The planning and execution of movement sequences in Parkinson’s disease. International Journal of Neuroscience 35, 51-58. Stern. Y. and R. Mayeux, 1986. ‘Intellectual impairment in Parkinson’s disease’. In: M.D. Yahr and K.J. Bergmann (eds.), Advances in neurology. New York: Raven Press.