Postural instability in Parkinson's disease

s41 Clinical Neurology and Neurosurgey, 94 (Suppl.) (1992) S41 -S45 0 1992 Elsevier Science Publishers B.V. All rights reserved 0303~8467/92/$0X00 CNN 00103

Postural instability in Parkinson’s disease B.R. Bloem Department of Neurology, UniversiryHospital, Leiden (The Netherlands)

Key words: Postural

instability;

Falls; Parkinson’s

disease,

posture

Summary Postural instability is one of the most disabling features of Parkinson’s disease. Many factors contribute to balance impairment of Parkinson patients, including disturbed postural reflexes and poor control of voluntary movement. Additional factors which place Parkinson patients at risk for falls are side-effects of medication (dyskinesias), the poor response of postural instability to antiparkinsonian medication, orthostatic hypotension, gait abnormalities, muscular weakness in leg muscles and superimposed age-related changes such as reduced peripheral sensation. Future studies should not only investigate means of preventing falls in unstable patients, but should also be directed towards development of new treatment. Because accumulating evidence indicates that postural instability is at least partially related to non-dopaminergic lesions, these novel therapeutic approaches should be aimed at overcoming non-dopaminergic neurotransmitter deficiencies.

Introduction

Selection and execution of postural reflexes

Postural instability constitutes one of the clinical hallmarks of Parkinson’s disease (PD). Balance impairment typically occurs in advanced stages of the disease [l], although in occasional patients it may be the presenting symptom [2]. Postural instability marks the onset of severe disability, particularly because the instability responds poorly to treatment with antiparkinsonian agents

Automatic postural responses constitute the “second line of defence” against perturbations of stance. One experimental paradigm which has been applied to investigate postural reflexes in PD involves the use of a movable

[1,3,41. Normal postural responses can grossly be subdivided into passive and active stabilizing factors. The passive visco-elastic properties of stretched muscles, tendons and ligaments constitute the “first line of defence” against a disturbance of equilibrium and contribute to postural stability in case of small postural perturbations. However, active muscular forces are of fundamental importance if larger perturbations occur. These active forces consist of automatic and voluntary postural responses which act in sequential order to reverse a displacement of the center of gravity. Abnormalities of active responses in PD will be described below (Table l), with special emphasis on our own studies of postural reflexes.

Correspondence to: Bastiaan R. Bloem, Department of Neurology, University Hospital, P.O. Box 9600,230O RC Leiden, The Netherlands. Tel.: (71) 263920; Pax: (71) 154537.

TABLE

1

PACIORS POSTURAL

WIIICII ARE INSTABILITY

TIIOUGIIT IN PD

TO

CONTRIBUTE

References which describe abnormalities of the postural PD are shown between brackets. The Arabic numerals correspond to those used in the text. Abnormal

component

of the postural

response

Selection and execution of postural reflexes 1. postural strategies (6-8) 2. reflex amplitudes (7,9-11,13,15,19) 3. reflex latencies (9,13,20) 4. anticipatoty postural reflex (13,22-24) Voluntary

postural

responses

(3,25,26)

Additional factors 1. antiparkinsonian medication (1,3,4,18,28) 2. detection of sensory information (28,29) 3. orthostatic hypotension (3,30) 4. gait abnormalities (28) 5. reduced muscular strength (31,32) 6. superimposed age-related changes (20)

TO

response in in the table

S42 TRIGGER

195 mVolt

I

CONTROL

;ri

0 mVolt

MG

470 mVolt

SL

ML

i 0 mVolt 420 mVolt

1 0 mVolt

MG

460 mVolt

0 mVolt

I

I

I

I

I

-100

0

100

200

300

Time (msec) Fig. 1. Representative traces of SL, ML and LL reflexes in response to a single “toe-up” platform rotation for a normal subject and a patient with MPTP-induced partinsonism. The indentation in the top trace indicates the trigger pulse, the falling edge of which marks the onset of platform movement. The onset of the three muscle responses is marked by an arrow. Vertical scales differ for the control and the patient. The “rati ML response is clearly increased in the MPTP patient compared to the control subject. However, note the noisy pre-stimulus background activity in both the gastrocnemius and tibialis anterior muscie of the MYFP patient. After correction for the influence of high backgroundmuscle activity, ML responses in the MFTP patient no longer differed from the control subject. MG = medial gastrocnemius; TA = tibiahs anterior.

forceplate platform upon which patients are freely standing. Sudden shifts or rotations of this platform elicit a complex set of postural reflexes in arm, leg, trunk and neck muscles. The most important postural reflexes are those which act at the ankle joint. In case of sudden toe-up rotations of the supporting platform, these ankle responses include short latency (SL) and medium latency (ML) reflexes in the stretched gastrocnemius muscle, and long latency (LL) reflexes in the shortened tibialis anterior muscle (Fig. 1). In this specific experimental paradigm, the summed force of SL and ML responses accentuates the posterior body sway which was induced by the toe-up platform movement. In contrast, the force exerted by LL responses represents the first active stabilizing factor [5]. Selection of the appropriate postural response involves activation of the correct postural strategy in which timing and amplitude of reflexes in different muscles are fine-tuned to counteract the disturbance of equilibrium. Centrally pre-programmed postural strategies are triggered and determined by the “on-line” available sensory

information and prior experience. Potential postural responses include (a) the ankle stru&y with a distal-toproximal reflex activation, sufficient to compensate for small and slow perturbations, (b) the hip strategy which is selected if larger or faster perturbations occur, or if subjects are standing on a small support surface, and (c) the stepping or stumbling strategy which is selected in case of an imminent fall [6). Depending on the environmental context (support surface conditions, nature of the postural perturbation) and the degree of prior experience, normal subjects can select a spectrum ofcombinations of the ankle and hip strategies. This “mixed ~pattern” always consists of a sequential activation of the contributingpostural strategies. 1. Abnormal postural strategies

Parkinson patients are able to detect and respond to a perturbing stimulus, but they inadvertently select inappropriate responses. Thus, in contrast to normal subjects, postural unstable Parkinson patients select a reversed,

S43 proximal-to-distal,

innervation

sequence

even

in re-

sponse to small perturbations [7]. Furthermore, Parkinson patients seem to activate different postural strategies simultaneously rather than sequentially [6]. This results in co-contraction and joint stiffness which will interfere with the rapid corrective movements needed to prevent a fall. Finally, postural strategies in PD are futed and independent of the functional demands [S]. 2.Abnomtal reflex ampliludes In Parkinson patients, amplitudes of postural destabilizing ML reflexes are increased in the toe-up rotational paradigm described above [7,9]. We have observed a similar increase of ML responses in four out of five patients with parkinsonism induced by 1-methyl-Cphenyl1,2,3,4_tetrahydropyridine or MPTP [lo] (Fig. 1). Increased ML responses are present in Parkinson patients with severe postural instability, but not in patients with mild to moderate instability [7]. This suggests that abnormal ML reflexes, as measured in a toe-up rotational paradigm, may contribute to balance impairment in daily life. We have recently noted that amplitudes of postural stabilizing LL reflexes are decreased in PD (unpublished observations). Corrective responses were also decreased in Parkinson patients who were subjected to sudden backward translations of a supporting platform [ll]. Furthermore, underscaling does not seem to be the only abnormality of LL responses in PD. Thus, normal subjects adapt the size of postural reflexes to counteract large perturbations by large compensatory LL responses and vice versa [12]. Parkinson patients have lost this ability and select a constant reflex size irrespective of the magnitude of the platform movement [13]. It thus appears that a combination of increased destabilizing responses and abnormal compensatory responses contributes to the postural instability in PD. However, it must be noted that the observed abnormalities of postural reflexes may at least partially be a consequence, rather than a cause, of the postural instability in PD [14]. Accumulating evidence indicates that some of the abnormal postural reflexes described above are related to the presence of non-dopaminergic lesions which emerge in advanced stages of PD [l]. Thus, ML responses are not enhanced in parkinsonian patients with a selective dopamine deficiency induced by chronic use of neuroleptics [15]. Furthermore, after correction for the confounding influence of high background muscle activity, ML responses are normal in patients with MPTP-induced parkinsonism and in patients with young-onset PD (unpublished observations) who both have a selective dopamine deficiency [16,17]. These results could at least partially explain why the appearance of postural instability and ab-

normal

postural

movements

in later stages of PD is par-

alleled by a poor response to dopaminergic compounds [1,3,4,18]. However, it must be realized that other abnormalities of postural reflexes are not related to non-dopaminergic lesions. For example, scaling of LL-amplitudes to small and large platform perturbations is impaired not only in PD [13], but also in patients with MPTP-induced parkinsonism [19]. We therefore postulate that lesions in both dopaminergic and non-dopaminergic circuits cause abnormal postural reflexes and contribute to balance impairment in PD. 3.Abnomzal reflex latencies Several investigators have noted that onset latencies of postural reflexes are delayed in PD [9,13,20]. It is possible that a delayed onset of postural reflexes contributes to the instability of Parkinson patients since studies in hemiplegic patients show that even minimally prolonged onset latencics of ankle reflexes may lead to destabilizing movements at the knee and hip [20]. A similar delay of LL responses is thought to contribute to the postural instability of patients with Huntington’s disease [21]. 4. Abnomlal anticipatory postural reflexes Any voluntary motor act performed during stance produces a disturbance of equilibrium which needs to be preceded by anticipatory postural adjustments. This predictive correction of body sway requires the use of prior experience, advance information or internal feedforward control. The amplitude of anticipatory postural reflexes seems to be diminished or even absent in PD, particularly in postural unstable patients [22]. This issue is not completely resolved since Diener and Dichgans [23] contended that anticipatory postural reflexes which precede rapid arm extension or a sudden raise on the tip toes are normal in PD. These conflicting results show that some components of the anticipatory response may be abnormal in PD, whereas others remain intact. However, the fundamental abnormality in PD seems to be the coordination of fragments which constitute the overall anticipatory postural reflex. Thus, according to Nashner [24] anticipatory postural reflexes are not merely lost in PD, but Parkinson patients simply select inappropriate and poorly coordinated responses. Another form of predictive control of reflex amplitudes also seems abnormal in PD. Thus, when presented with a random mix of small and large platform rotations, normal subjects assume a “worst-case-scenario” and select a large default LL response whose size equals the one needed for the large perturbation [12]. In contrast, postural unstable Parkinson patients are unable to modify the

s44

size of the LL response under these unpredictable conditions [13]. Voluntary postural responses

The compensatory visco-elastic and reflexive reactions described above are followed by a “third line of defence” which consists of voluntary corrective responses such as protective arm movements. Impaired programming and poor coordination of voluntary movement likely contribute to the instability of Parkinson patients [25]. A relation between bradykinesia and clinically rated balance impairment has been noted by several investigators [3,26]. It requires little elaboration to realize that sudden freezing of voluntary movements may cause serious impairment of balance control. Furthermore, the rigidity and intrinsic muscle stiffness which occur in PD [11,27] may superimpose upon the bradykinesia and cause a biomechanical delay of voluntary compensatory responses. Additional

factors

1. Antiparkinsonian medication

Antiparkinsonian medication often fails to correct balance impairment in PD. Furthermore, dyskinesias due to an “overshoot” effect of medication may cause involuntary body sway beyond the limits of postural stabilization and result in a fall [28]. 2. Detection of sensory information An integration of proprioceptive, visual and vestibular feedback is required to adequately maintain the body in a vertical position. Detection of sensory information via these mechanisms may be impaired in PD. Thus, reduced peripheral sensation due to an unrelated neuropathy [28], oculomotor disturbances [28] and abnormal labyrinthine function [29] may contribute to the instability of Parkinson patients. 3. Orthostatic hypotension Autonomic disturbances may lead to orthostatic hypotension and thus contribute to the instability. However, autonomic disturbances are rarely severe in true PD [30], and orthostatic hypotension did not correlate with postural instability in one study [3], suggesting that autonomic disturbances do not play a major role in the balance impairment in PD. 4. Gait abnormalities The typical shuffling gait of Parkinson patients creates a risk of falling [28]. Retro- and propulsion are obviously also a considerable risk factor for falls.

5. Reduced muscular strength in leg muscies

Reduced muscular strength in the tibialis anterior muscle may play a role in backward falls of healthy elderly subjects [20]. A similar situation may exist in PD since both isotonic [31] and isometric strength 1321seem to be reduced in Parkinson patients. 6. Superimposed age-related factors

Parkinson patients are of course not exempt from “normal” age-related changes which compromise balance control in the elderly [20]. Thus, factors such as heart disease, sensory changes in legs and dementia may contribute to balance impairment in elderly Parkinson patients. Conclusions Treatment of balance impairment in PD imposes a major challenge for clinicians. In view of the many factors involved it is unlikely that a single therapeutic approach will be able to correct all balance problems. Physical therapy and assistance in creating a safer home environment should clearly be part of the approach. Use of a cane, walker or wheelchair can sometimes help prevent falls. Finally, future research should focus on the development of new antiparkinsonian agents which are capable of restoring non-dopaminergic neurotransmitter deficiencies in PD. Ghika and colleagues [33] reported that an increase of central norepinephrine by idazoxan led to both objective and subjective improvement of postural instability in patients with progressive supranuclear palsy. Such studies provide hope that an effective treatment of postural instability in PD may become available in the near future. Acknowledgements Raymund Roos, Gert van Dijk, Dennis Beckley, Mike Remler and Gerard van der Giessen are gratefully acknowledged for their help in the preparation of this manuscript. This study was supported in part by Stichting de Drie L,ichten.

References 1 Bonnet A-M, Loria Y, Saint-Hilaire M-H, Lhermitte F, Agid Y. Neurology 1987; 37: 1539-1542. 2 Klawans HL, Topel JL. JAMA 1974; 230: 1555-1557. 3 Koller WC, Glatt S, Vetere-Overfield B, Hassanein R Chin Neuropharmacol1989; 12: 98-105. 4 Klawans HL. Movement Disorders 1986; 1: 187-192. 5 Dichgans J, Diener HC. In Struppler A, Weindl A (eds.), Clinical Aspects of Sensory Motor Integration. Berlin-Heidelberg: Springer-Verlag, 1987: 165-173. 6 Nashner LM, McCollum G. Behav Brain Sci 1985; 8: 135-172. 7 Beckley DJ, Bloem BR, van Dijk JG, Roos RAC, Remler Ml’, Electroenceph clin Neurophysiol 1991; 81: 263-268.

8 Horak FB, Nashner LM, Nutt JG. Sot Neurosci Abstr 1986; 10: 634. 9 Scholz E, Diener HC, Noth J, Friedemann H, Dichgans J, Bather M. J Neurol Neurosurg Psychiat 1987; 50: 66-70. 10 Bloem RR, Beckley DJ, Tetrud JW, et al. Movement Disorders

22 Traub MM, Rothwell JC, Marsden CD. Brain 1980; 103: 393-412. 23 Diener 1IC, Dichgans J. In Amblard B, Berthoz A, Clarac F, (eds.), Posture and Gait: Development, Adaptation and Modulation. Amsterdam: Elsevier, 1988: 229-235.

1990; 5 (Suppl. 1): 25. 11 Dietz V, Berger W, Homtmann GA. Ann Neural 1988; 24: 660-669. 12 Beckley DJ, Bloem BR, Remler MP, Roos RAC, van Dijk JG. Electroenceph clin Neurophysiol 1991; 81: 353-358. 13 Beckley DJ, Bloem BR, Remler MP. In Agnoli A (ed.), Proc Eur Conf Parkinson’s Dis Extrapyramidal Disorders. Rome: John Libbey, 1991: 507-513. 14 Bloem BR, van Dijk JG, Beckley DJ, Roos RAC, Remler MP, Bruyn GW. Med Hypoth, 1992: in press. 15 Beckley DJ, Bloem BR, Singh J, Remler MP, Wolfe S, Roos RAC. Clin Neural Neurosurg 1991; 93: 119-122. 16 Bloem BR, Irwin I, Buruma OJS, et al. J Neural Sci 1990; 97: 273-293. 17 Blin J, Dubois B, Bonnet AM, Vidailhet M, Brandabur M, Agid Y. J Neural Neurosurg Psychiat 1991; 54: 78&782. 18 Lakke JPWF, vd Burg W, Wiegman J. Clin Neurol Neurosurg 1982; 84: 227-235. 19 Bloem BR, Roos RAC, Beckley DJ, et al. Abstr 10th Int Symp Parkinson’s Dis, Tohyo, October 27-30, 1991: 2S3. 20 Horak FB, Shupert CL, Mirka A. Neurobiol Aging 1989; 10: 727738. 21 Huttunen J, Hornberg V. J Neural Neurosurg Psychiat 1990; 53: 55-62.

24 Nashner LM. In Desmedt JE (ed.), Motor Control Mechanisms in IIcalth and Disease. New York: Raven Press, 1983: 607-619. 25 Calnc DB, Larsen TA, 13urton K. In Delwaide PJ, Agnoli A, (eds.), Clinical Neurophysiology in Parkinsonism. Amsterdam: Elsevier, 1985: 133-138. 26 Zetushy WJ, Jankovic J, Pirozzolo FJ. Neurology 1985; 35: 522-526. 27 Delwaidc PJ, Gonce M. In Jankovic J, Tolosa E (eds.), Parkinson’s Disease and Movement Disorders. Baltimore-Munich: Urban & Schwarzenberg, 1988: 59-73. 28 Paulson GW, Schacer K, Ilallum B. Geriatrics 1986; 41: 59-67. 29 Reichert WII, Doolittle J, McDowell FH. Neurology1982; 32: 11331138. 30 Barbeau A. In Vinken PJ, Bruyn GW, Klawans IIL (eds.), Handbook of Clincial Neurology, vol 5 (49): extrapyramidal disorders. Amsterdam: Elsevier, 1986: 87-152. 31 Kollcr W, Kase S. Eur Neurol 1986; 25: 13G-133. 32 Yanagawa S, Shindo M, Yanagisawa N. In Streifler MB, Korczyn AD, Melamed E, Youdim MB11 (eds.), Advances in Neurology, vol 53: Parkinson’s Disease: anatomy, pathology, and therapy. New York: Raven Press, 1990: 259-269. 33 Ghika J, Tennis M, Iioffman E, Schoenfeld D, Growdon J. Ncurology 1991; 41: 986-991.