- Email: [email protected]
Heel to toe motion characteristics in Parkinson patients during free walking
Clinical Biomechanics 16 (2001) 806±812
www.elsevier.com/locate/clinbiomech
Heel to toe motion characteristics in Parkinson patients during free walking Stefan Kimmeskamp *, Ewald M. Hennig Biomechanics Laboratory, University of Essen, Henri-Dunant-Str 65, Essen 45145, Germany Received 15 September 2000; accepted 23 July 2001
Abstract Objective. Plantar pressures of Parkinson patients in a mild or moderate stage of the disease were analyzed in order to determine characteristics of the heel to toe motion of the foot in Parkinson patients during free walking. Design. Pressure sensitive insoles were used to quantify the in-shoe pressure distribution for 24 patients with Parkinson's disease and for 24 age-matched healthy adults. Peak plantar pressures, relative loads and the variability of relative loads were analyzed for 10 dierent anatomical foot areas. Inferential statistics and regression analyses were performed to compare subject groups and to relate pressure data to a clinical score (Webster). Background. It is well-known that Parkinson patients experience disorders of balance, posture and gait. Recent studies revealed that Parkinson patients show abnormalities in foot strike during walking. However, only little is known about the heel to toe motion of the foot in Parkinson patients in compensating for instability during walking. Results. Parkinson patients show signi®cant changes in foot loading behavior. Parkinson patients have a reduced impact at heel strike. This mechanism was found to be related to the severity of the disease. Furthermore, Parkinson patients show a trend towards higher relative loads in the forefoot regions combined with a load shift towards medial foot areas. These mechanisms are highly stereotypical. Conclusion. Parkinson patients, even in a mild or moderate stage of the disease, show signi®cant changes in heel to toe motion of the foot during free walking. The characteristics of Parkinsonian gait are probably caused by adaptive mechanisms of the patients to avoid unsteadiness during walking. Relevance The results of this study revealed that Parkinson patients have characteristic heel to toe motion pattern. The determination of the foot-loading pattern using plantar pressure measurement may be used as a valuable tool for diagnostic, treatment and rehabilitation purposes. Furthermore, the strategy of the forefoot control in gait of Parkinson patients should be considered in Parkinsonian research. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Parkinson's disease; Gait analysis; Plantar pressure; Heel to toe motion; Postural stability; Variability; Balance
1. Introduction Parkinson's disease (PD) is a chronic and progressive disease of the brain. PD is caused by a de®ciency in the neurotransmitter dopamine in the region of the basal ganglia of the brain. This disorder of the basal ganglia evokes characteristic changes in motor activity. Patients with PD, notwithstanding the etiology of the disease, show combined clinical symptoms of rigidity, bradyki-
*
Corresponding author. E-mail address: [email protected] (S. Kimmeskamp).
nesia or even akinesia, as well as resting tremor. They also experience disorders of posture, balance, and gait, which often result in accidental falls [1,2]. Most studies on gait in PD were undertaken to quantify kinematic characteristics of the Parkinsonian gait in order to investigate the underlying mechanism of PD [3±8]. Typical features of the Parkinsonian gait are small shuing steps and a general poverty and slowness of movement, called hypokinesia. Compared to healthy elderly subjects, patients with PD demonstrate reduced stride length and walking velocity during free ambulation. Double support duration and cadence rate are increased [3±8]. However, Blin [9] suggested that the basic gait patterns are preserved in PD.
0268-0033/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 1 ) 0 0 0 6 9 - 9
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
Blin [9] and Hausdor [6] investigated the variability of the Parkinsonian gait and found higher variability values for spatial temporal gait parameter including stride length and double support duration. It seems that PD patients have diculties combining single steps to an accurate sequence of movement. This can be a result of disturbed electromyographic activity of the lower leg muscles [10,11], caused by diculties in coordinating submovements or due to a lack of the basal ganglia in providing accurate internal stimulus [4,12]. An increased variability of gait, especially of double support duration, is assumed to be related to an impaired balance control in PD [13]. PD patients show signi®cant abnormalities in foot strike [14±18]. Therefore, they face a higher risk of falling. In normal gait, the foot strikes the ground with the heel ®rst, followed by the mid- and forefoot. This procedure is called heel to toe walking or heel to toe motion. It was shown that PD patients in more advanced stages of the disease tend to strike the ground simultaneously with the heel and the forefoot resulting in a ¯at foot strike [16]. Sometimes even the forefoot strikes the ground ®rst (toe-heel-walking). In addition PD patients have a reduced foot lifting during the swing phase [5], which produces a smaller clearance between the toes and the ground. Whereas the kinematics of gait in PD is frequently documented only little is known about the kinetic features of the Parkinsonian gait. The vertical ground reaction force (GRF) component in normal gait has typically two peaks separated by a force minimum in the middle part of the GRF curve. The ®rst peak occurs with heel strike while the second peak is caused by the push-o from the ground. Force reduction in the middle part of the force curve becomes more pronounced with higher walking speeds. This is a consequence of GRF reduction by knee ¯exion of the supporting leg and lift of the center of mass (CoM) by the swinging leg. It was found that the shape of the vertical GRF signal is abnormal in PD [15,17]. Using force plate measurements reduced forces were found for heel contact and the push-o phase [15]. However, the shape of the vertical GRF signal, including two marked peaks, was similar for PD patients with moderate gait disorder as compared to elderly controls. PD patients with the typical shuing gait show only one narrow peak in vertical GRF signal [17]. Nieuwboer et al. [14] investigated the heel to toe motion of the foot in PD patients using pressure sensitive insoles. They found that PD patients demonstrate reduced forces at the heel and forefoot and increased load under the midfoot. Furthermore, the pathway of the center of force was found to start more towards the midfoot in PD patients. Additionally, the end of the pathway tends to be located less far forward for PD patients. Therefore, it was concluded that PD patients in later stages of the disease have a reduced heel to toe motion of the foot and that
807
the load shift towards the midfoot could be due to a compensation for possible loss of postural control. However, only little is known about heel to toe motion in Parkinson patients in a mild or moderate stage of the disease. Furthermore, the role of dierent heel to toe motion strategies in compensating for instability during walking is not yet fully understood. Therefore, the aim of this study was to investigate the heel to toe motion characteristics in Parkinson's patients during walking, especially the aspects of anterior±posterior and medial± lateral control. 2. Methods 2.1. Subjects Twenty-four patients (17 male and 7 female) with a diagnosis of idiopathic Parkinsonism (PD) participated in this study. The mean age of the PD group was 63.75 (SD, 10.05) years and the mean weight was 78.67 (SD, 13.61) kg. All patients lived independently in their own apartments and could ambulate freely without any help or walking aids. A neurologist examined the patients in order to classify the severity of the disease according to the Webster score [8]. The Webster score evaluates 10 single categories. These categories are bradykinesia of hands, rigidity, posture, upper extremity swing, gait, tremor, facies, seborrhea, speech and self-care. Each category is tested separately and scored from 0 for no impairment to 3 for severest impairment. Adding the scores of each category results in the total Webster score. A total Webster score of 30 indicates the severest state of Parkinson disease. The Webster scores for the patients ranged from 3 to 17 and the group mean score was 9.96 (SD, 3.36). The PD group measurements were performed in the on-phase approximately 1 h after individual medication of each patient. During this period best medication status can be expected. No ``freezing'' eects or festination were observed. PD patients, showing other movement disorders in addition to the Parkinsonian movement impairment, were excluded from this study. 10 male and 14 female age- and weightmatched healthy adults belonged to the control group. These subjects had a mean age of 66.06 (SD, 9.16) and a mean weight of 75.5 (SD, 14.51). No subjects with foot deformities participated in the study. 2.2. Instrumentation and procedure The Pedar insole system (NOVEL GmbH, Munich, Germany) was used to determine the in-shoe pressure distribution during normal gait of all participants. The system is based on ¯exible insoles with 99 capacitive transducers, which were sampled at 50 Hz. The transducer signals were sent to an electronic processing unit
808
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
(<600 grams), which was fastened by a belt to the subjects' waists. The electronic unit was connected via a 10-m long cable to the computer. Five dierent sizes of insoles (European sizes; 36-37, 38±39, 40±41, 42±43 and 44±45) were used to detect plantar foot pressures. Measuring equipment and cable were arranged to minimize interference with the subjects' gait. The demographic data assessment, including a neurological examination of the patients, was performed prior to the gait measurements. Following instrumentation, the subjects were asked to walk a distance of 11 m at their preferred speed to obtain undisturbed individual gait pattern. All subjects used identical socks and the same kind of footwear, a simple canvas shoe without any support structures. 2.3. Data analysis Walking speed, cadence and the plantar pressure data for the right and left foot were collected throughout the 11-m walk. As PD patients have problems initiating and terminating gait, the ®rst and the last steps were excluded from the analysis. Only steps in the middle of the walk, representing ``individual normal walking'', were used for further evaluation. Parkinson's symptoms may occur on one side of the body. To determine the characteristic features of the Parkinsonian gait, it is important to look at the steps of the left and the right foot separately. Therefore, the data of ®ve single steps of the left and the right foot were processed for each participant, averaged and statistically analyzed. To determine dierences in foot loading between the PD patients and the control group, a mask of 10 different anatomical areas was de®ned with the NovelWin software and applied to all insole sizes (Fig. 1). From the pressure data the relative loads (RL) were calculated for each anatomical area separately. The
relative load analysis gives more detailed information about the characteristics of the heel to toe motion of the foot than the peak pressures, because the time domain of loading is considered. Furthermore, the relative load analysis largely compensates for individual dierences in subjects' body weight and ground contact times. Therefore, it is an appropriate method to compare walking pattern of patients and healthy subjects, even if they exhibit slight dierences in walking speeds [19]. To calculate the relative loads, the force time integral (FTI) of each anatomical area was divided by the sum of all force time integral values and multiplied by 100. Consequently, the relative loads were given in a percent value. In addition, the gait line of each single step was determined. The gait line represents the movement of the center of pressure (CoP) under the foot during the heel to toe motion. To describe a more medial or lateral orientation of the gait line a lateral-index (LI) was de®ned. The area between the gait line and the lateral border of the insole, as well as the length of the gait line were used to determine the LI. The LI was calculated by dividing the cm2 -value of the area by the squared value of the length of the gait line. A greater LI value indicates a more medially oriented path of the gait line. Furthermore, the intra-subject variability of all parameters was determined by calculating each subject's coecient of variability for the ®ve single steps. Simple regression analyses were used to analyze the relationship between plantar peak pressures and the severity of the disease (expressed by the Webster score). Regressions were undertaken on peak pressures in the heel region, the mid- and forefoot areas and for the hallux. The statistical analyses were performed with a statistics software package (StatView II). A simple analysis of variance (A N O V A ) was performed to determine whether signi®cant dierences exist between the Parkinson group and the control group. The Schee` test was used for post hoc testing. For multiple tests a Bonferroni correction was done. Signi®cance levels were chosen at 6 0:05 and 6 0:01, respectively. 3. Results
Fig. 1. Foot areas.
Table 1 summarizes the demographic and spatialtemporal data of the PD patients and the control group. Although not signi®cant, the walking speed and cadence rate were slightly lower for the PD patients. Probably as a consequence of the reduced walking speed, the PD group also shows a slightly higher mean ground contact time. An analysis of the GRF curves revealed that all controls and 20 patients exhibited pronounced double peak GRF curves, whereas four patients showed single peak GRF force curves.
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
809
Table 1 Demographic features and spatial-temporal data for PD patients and the control group [means; (SD)]
n Age Weight (kg) Walking speed (m/s) Cadence (step/min) Foot contact time (ms) Webster score
Parkinson patients
Control group
P
24 63.75 (10.05) 78.67 (13.61) 0.97 (0.33) 104.21(18.22) 724.28 (101.38) 9.96 (3.36)
24 66.08 (9.16) 75.5 (14.51) 1.1 (0.27) 109.69 (15.62) 689.42 (115.63) ±
± n.s. n.s. n.s. n.s. n.s. ±
3.1. Peak pressures The single values of peak pressures, relative loads and intra-individual variability of the relative loads were illustrated in Table 2. The plantar pressure analysis revealed similar pressure distribution pattern for PD patients and the control group. However, Parkinson patients have signi®cantly P 0:031 reduced peak pressures in the lateral heel region compared to the control group. In the medial heel region peak pressures ranged from 217.75 kPa for the patients to 242.05 kPa for the control group. Although this dierence in medial heel was not signi®cant P 0:08 a general trend towards lower peak pressures for PD patients in the heel region can be observed.
icant P 0:035. For the metatarsal heads and the toe regions no signi®cant dierences in RL were found for patients and the control group. Nevertheless, it seems that PD patients show a general load shift from the heel region towards the mid- and forefoot regions compared to the control group (Fig. 2). Unexpected results were found for the analysis of the intra-individual variability of the RL distribution. Whereas the variability in the heel region was slightly higher for patients, the variability in all mid- and forefoot areas was reduced for the PD group (Fig. 3). The
3.2. Relative loads and relative load variability The RL for the heel region are comparable to the results of the peak pressure distribution. While the RL for the lateral heel are signi®cantly lower for the PD group P < 0:01 the load reduction for the patients in M2 (medial heel) is only moderate. On the other hand, patients have higher RL in the midfoot areas. However, only the dierence in the medial midfoot area is signif-
Fig. 2. Relative loads for patients and controls. Single values are summarized in Table 2, (P < 0:05; P < 0:01).
Table 2 Plantar peak pressures, relative loads and variability (CoV) of relative loads [means; (SD)]
*
Masks (anatomical area)
Peak pressures (kPa)
Relative loads (%)
CoV relative load (%)
n 24
PD patients
Control
PD patients
Control
PD patients
Control
M1 (medial heel) M2 (lateral heel) M3 (medial midfoot) M4 (lateral midfoot) M5 (1. metatarsal head) M6 (2. metatarsal head) M7 (3.-5. metatarsal heads) M8 (hallux) M9 (2. toe) M10 (lateral toes)
217.75 (78.3) 216.29 (73.8) 80.69 (26.5) 132.12 (54.8) 217.23 (75.4) 223.02 (67.4) 218.09 (75.4) 205.59 (97.0) 123.58 (65.4) 101.84 (49.5)
242.05 (57.7) 245.83 (57.4) 75.83 (28.9) 142.40 (62.1) 202.74 (53.4) 244.89 (69.6) 242.74 (61.2) 228.61 (89.2) 132.60 (43.3) 114.79 (37.9)
16.81 (7.9) 15.72 (5.3) 3.39 (2.8) 12.43 (6.4) 11.97 (5.4) 10.76 (3.3) 17.78 (5.5) 5.37 (3.2) 2.81 (2.0) 2.98 (2.3)
17.58 (3.2) 18.96 (2.7) 2.41 (1.6) 11.58 (3.8) 10.48 (3.0) 10.49 (2.2) 18.56 (3.2) 4.6 (2.1) 2.54 (1.0) 2.81 (1.6)
18.81 16.77 25.23 27.71 25.58 12.34 16.51 24.18 20.14 25.76
16.78 14.96 33.32 29.34 31.23 12.88 19.23 29.64 21.32 27.46
P 6 0:05. P 6 0:01.
**
(10.8) (9.7) (15.0) (15.7) (10.6) (6.8) (9.1) (12.1) (11.7) (13.7)
(16.8) (8.5) (18.0) (12.2) (15.3) (6.7) (9.1) (13.6) (12.0) (14.2)
810
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
(Webster score). A higher Webster score characterizes an advanced severity of the disease. A strong negative relationship r 0:58 between the peak heel pressures and the Webster score was found P 0:002. A decrease of heel loading occurs with an increase of the severity of the disease (Fig. 4). 4. Discussion Fig. 3. Variability of the relative loads for patients and controls. Single values are summarized in Table 2, P < 0:05; P < 0:01.
reduction of RL for the patients was revealed to be signi®cant in the medial midfoot and forefoot regions (M3, P 0:019; M5, P 0:038; M8, P 0:033). 3.3. Center of pressure Considering the length of the gait line, only small, non-signi®cant dierences were observed for PD patients (184.4 mm) and the control group (179.6 mm) indicating similar heel to toe motion pattern of the foot in anterior±posterior direction for the PD patients compared to the control group. However, the LI analysis revealed a dierent result. Whereas the mean LI for the PD group was 0.185, a LI value of 0.175 was calculated for the control group. The greater LI value for the PD group indicates a more medially oriented path of the gait line for the patients as compared to the control group P 0:048. 3.4. Relationship between clinical rating and gait parameter A simple regression analysis was performed to relate the kinetic gait parameters to the clinical rating score
Fig. 4. Regression curve of signi®cant correlation P < 0:01 between the Webster score and the heel±peak±pressures.
PD patients show slight, non-signi®cant reductions in walking speed and cadence, while total foot contact time was slightly increased (Table 1). These ®ndings are consistent with results of other studies investigating free ambulation in PD [4±6,8,9,14]. However, none of these dierences were signi®cant, suggesting a similar performance of the PD patients and the control group. A likely explanation for the similar performance may be the low severity of the disease in the participating PD patients. All patients were estimated to have only mild or moderate Parkinsonism [mean Webster score 9.96 (3.36)], living independently and managing life on their own. In order to analyze the kinetic characteristics of the heel to toe motion of the foot in PD, a mask of 10 different foot areas was applied, and pressure data were calculated for each area separately. The plantar pressure patterns result from passive and active foot functions during the heel to toe motion. Whereas the peak pressures under the heel and midfoot regions are more passively in¯uenced by accepting body weight at heel strike and midstance, the pressures at the forefoot are more likely to be in¯uenced by additional superimposing activity of the foot muscles. In our study the peak pressure distribution was similar for PD patients and the control group. However, in the heel region peak pressures are reduced for PD patients (M2; P 0:03). The reduction in heel peak pressures is an indication of a less pronounced heel strike in PD, which is similar to that reported previously [14,16]. Furthermore, this unloading of the heel is strongly related to the severity of the disease. A linear regression analysis revealed that a decrease of peak pressures in the heel region occurred with an increase of the severity of the disease indicated by a higher Webster score (Fig. 4). To analyze dierent heel to toe motion strategies during walking, the relative load analysis is the appropriate method, since this procedure largely compensates for dierences in ground contact times. This analysis suggests that PD patients in our study prefer a more forward oriented foot-loading pattern towards the midfoot and forefoot (Fig. 3). Nieuwboer et al. [14] found signi®cantly higher midfoot loading for PD patients compared to elderly controls. They suggested a more equal spread of force, and concluded that this characteristic feature might be explained by a compen-
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
satory mechanism for possible loss of postural control. However, as discussed above, the loads under the heel and midfoot region are more passively induced rather than actively in¯uenced by muscular activity. Compared to limited correction possibilities of the heel and midfoot, the ¯exible forefoot structure allows much more muscular control to compensate for instability. Therefore, it can be speculated that PD patients use a strategy of forefoot balance control during walking and that a higher midfoot loading appears as a consequence of this balance strategy. In addition, the body is in an unstable position during walking, when the heel of one leg strikes the ground and the toes of the other leg are still in contact with the ¯oor (beginning of double support). Thus, PD patients may avoid such an unstable position by shifting the load towards the forefoot and producing ¯at foot strikes with a premature contact of the forefoot with the ground. This eect could re¯ect an adaptive mechanism of PD patients to compensate for instability by generating a stable forefoot placement. A stable forefoot placement enables PD patients to gain additional safety by active muscular control. Beside a load transfer towards the forefoot a general medial load shift during the heel to toe motion was observed for the PD patients. The gait line in PD patients proceeds signi®cantly more medially as compared to elderly controls. This is demonstrated by a higher lateral index for PD patients, suggesting that the tendency to walk more on the medial site of the foot is a characteristic feature of Parkinsonian gait. The medial oriented heel to toe motion of the foot in PD and the tendency towards higher forefoot loading during gait in Parkinson patients as it was found in this study, probably re¯ect a compensatory mechanism for postural instability during walking. The variability analysis revealed that for PD patients the variability of the relative loads in the mid- and forefoot regions was reduced compared to elderly controls (Fig. 4). This was an unexpected ®nding as an increased variability of spatial temporal gait parameters is considered to be an expression of Parkinsonian movement disorder [6,9] and seems to be related to falls [13,20]. However, our results show, that PD patients in their ``on-phase'' generate highly stereotypical mid- and forefoot loading pattern, especially in the medial midand forefoot foot areas (M3; M5; M8). Cavanagh et al. [21] investigated the variability of foot loading in diabetic patients with neuropathy and healthy controls. It was found that the medial forefoot areas (Metatarsal head I, Hallux) exhibit the highest variability in plantar loading compared to other foot structures. The authors concluded that the variability in loading re¯ects the use of the medial column of the foot during walking and that step to step variations in acceleration and small changes in direction during gait are partly controlled by the medial ray of the foot. As the authors found no
811
dierences in loading variability between neuropathic patients with insensate feet and controls, they further concluded that this ``micro-control'' of gait is presumably centrally mediated. The results of our study con®rm the conclusions of Cavanagh et al. The variability of the relative load in PD patients is signi®cantly reduced in the medial foot areas (M3; M5; M8). Therefore, in Parkinson patients the micro-control of gait performed by the medial ray of the foot is characterized by stereotypical loading pattern, which is probably caused by basal ganglia dysfunction. This pattern may results in a reduced adjustability of the micro-control mechanism to compensate for external and internal changes during gait and is therefore maybe one risk factor for falls in Parkinson patients. 5. Conclusion The results of this study indicate that PD patients, even at a mild or moderate stage of the disease, show signi®cant changes in the heel to toe motion of the foot. PD patients tend to have a less pronounced heel strike. This mechanism was found to be related to the severity of the disease. Furthermore, a tendency towards higher forefoot loading and a medial load shift were determined for the PD patients. These characteristics of Parkinsonian gait are probably caused by an adaptive mechanism of PD patients to avoid unsteadiness during walking. This mechanism is stereotypical and therefore may less adjustable to compensate for external and internal changes during gait. The determination of foot loading patterns using plantar pressure measurement may be used as a valuable tool for diagnostic, treatment and rehabilitation purposes. The strategy of the forefoot control in PD patients should be considered in PD research and rehabilitation of Parkinsonian gait.
Acknowledgements This study was supported by: Alfried Krupp Krankenhaus ± Neurologie; Essen, Germany Gertrudis-Klinik; Biskirchen, Germany Klinik Ambrock ± Neurologie; Hagen, Germany.
References [1] Aita JF. Why patients with Parkinson's disease fall. Jama 1982;247:515±6. [2] Koller WC, Glatt S, Vetere-Over®eld B, Hassanein R. Falls and Parkinson's disease. Clin Neuropharmacol 1989;12:98±105. [3] Morris M, Iansek R, Matyas T, Summers J. Abnormalities in the stride length-cadence relation in parkinsonian gait. Mov Disord 1998;13:61±9.
812
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
[4] Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson's disease. Normalization strategies and underlying mechanisms. Brain 1996;119:551±68. [5] Murray MP, Sepic SB, Gardner GM, Downs WJ. Walking patterns of men with parkinsonism. Am J Phys Med 1978;57:278± 94. [6] Hausdor JM, Cudkowicz ME, Firtion R, Wei JY, Goldberger AL. Gait variability and basal ganglia disorders: stride-to-stride variations of gait cycle timing in Parkinson's disease and Huntington's disease. Mov Disord 1998;13:428±37 [published erratum appears in Mov Disord 1998 Jul;13:757]. [7] Zijlstra W, Rutgers AW, Van Weerden TW. Voluntary and involuntary adaptation of gait in parkinson's disease. Gait Posture 1998;7:53±63. [8] Vieregge P, Stolze H, Klein C, Heberlein I. Gait quantitation in Parkinson's disease ± locomotor disability and correlation to clinical rating scales. J Neural Transm 1997;104:237±48. [9] Blin O, Ferrandez AM, Serratrice G. Quantitative analysis of gait in parkinson patients: increased variability of stride length. J Neurol Sci 1990;98:91±7. [10] Dietz V, Zijlstra W, Prokop T, Berger W. Leg muscle activation during gait in parkinson's disease: adaptation and interlimb coordination. Electroencephalogr Clin Neurophysiol 1995;97:408±15. [11] Miller RA, Thaut MH, McIntosh GC, Rice RR. Components of EMG symmetry and variability in parkinsonian and healthy elderly gait. Electroencephalogr Clin Neurophysiol 1996;101:1±7. [12] Albin RL, Young AB, Penney JB. The functional anatomy of disorders of the basal ganglia. Trends Neurosci 1995;18:63±4.
[13] Gabell A, Nayak US. The eect of age on variability in gait. J Gerontol 1984;39:662±6. [14] Nieuwboer A, De Weerdt W, Dom R, Peeraer L, Lesare E, Hilde F, et al. Plantar force distribution in Parkinsonian gait: a comparison between patients and age-matched control subjects. Scand J Rehabil Med 1999;31:185±92. [15] Koozekanani SH, Balmaseda Jr. MT, Fatehi MT, Lowney ED. Ground reaction forces during ambulation in parkinsonism: pilot study. Arch Phys Med Rehabil 1987;68:28±30. [16] Hughes JR, Bowes SG, Leeman AL, O'Neill CJ, Deshmukh AA, Nicholson PW, et al. Parkinsonian abnormality of foot strike: a phenomenon of ageing and/or one responsive to levodopa therapy?. Br J Clin Pharmacol 1990;29:179±86. [17] Ueno E, Yanagisawa N, Takami M. Gait disorders in parkinsonism a study with ¯oor reaction forces and EMG. Adv Neurol 1993;60:414±8. [18] Diehl RR, Schneider U, Konietzko M, Hennerici M. Quantitative analysis of gait in neurological gait disorders. In: Woollacott MJ, Horak FB, editors. Posture and gait control mechanisms. Portland: University of Oregon Books; 1992. p. 47±50. [19] Hennig EM, Staats A, Rosenbaum D. Plantar pressure distribution pattern of young school children in comparison to adults. Foot Ankle 1994;15:35±40. [20] Maki BE. Gait changes in older adults: predictors of falls or indicators of fear. J Am Geriatr Soc 1997;45:313±20. [21] Cavanagh PR, Perry JE, Ulbrecht JS, Derr JA, Pammer SE. Neuropathic diabetic patients do not have reduced variability of plantar loading during gait. Gait Posture 1998;7:191±9.
www.elsevier.com/locate/clinbiomech
Heel to toe motion characteristics in Parkinson patients during free walking Stefan Kimmeskamp *, Ewald M. Hennig Biomechanics Laboratory, University of Essen, Henri-Dunant-Str 65, Essen 45145, Germany Received 15 September 2000; accepted 23 July 2001
Abstract Objective. Plantar pressures of Parkinson patients in a mild or moderate stage of the disease were analyzed in order to determine characteristics of the heel to toe motion of the foot in Parkinson patients during free walking. Design. Pressure sensitive insoles were used to quantify the in-shoe pressure distribution for 24 patients with Parkinson's disease and for 24 age-matched healthy adults. Peak plantar pressures, relative loads and the variability of relative loads were analyzed for 10 dierent anatomical foot areas. Inferential statistics and regression analyses were performed to compare subject groups and to relate pressure data to a clinical score (Webster). Background. It is well-known that Parkinson patients experience disorders of balance, posture and gait. Recent studies revealed that Parkinson patients show abnormalities in foot strike during walking. However, only little is known about the heel to toe motion of the foot in Parkinson patients in compensating for instability during walking. Results. Parkinson patients show signi®cant changes in foot loading behavior. Parkinson patients have a reduced impact at heel strike. This mechanism was found to be related to the severity of the disease. Furthermore, Parkinson patients show a trend towards higher relative loads in the forefoot regions combined with a load shift towards medial foot areas. These mechanisms are highly stereotypical. Conclusion. Parkinson patients, even in a mild or moderate stage of the disease, show signi®cant changes in heel to toe motion of the foot during free walking. The characteristics of Parkinsonian gait are probably caused by adaptive mechanisms of the patients to avoid unsteadiness during walking. Relevance The results of this study revealed that Parkinson patients have characteristic heel to toe motion pattern. The determination of the foot-loading pattern using plantar pressure measurement may be used as a valuable tool for diagnostic, treatment and rehabilitation purposes. Furthermore, the strategy of the forefoot control in gait of Parkinson patients should be considered in Parkinsonian research. Ó 2001 Elsevier Science Ltd. All rights reserved. Keywords: Parkinson's disease; Gait analysis; Plantar pressure; Heel to toe motion; Postural stability; Variability; Balance
1. Introduction Parkinson's disease (PD) is a chronic and progressive disease of the brain. PD is caused by a de®ciency in the neurotransmitter dopamine in the region of the basal ganglia of the brain. This disorder of the basal ganglia evokes characteristic changes in motor activity. Patients with PD, notwithstanding the etiology of the disease, show combined clinical symptoms of rigidity, bradyki-
*
Corresponding author. E-mail address: [email protected] (S. Kimmeskamp).
nesia or even akinesia, as well as resting tremor. They also experience disorders of posture, balance, and gait, which often result in accidental falls [1,2]. Most studies on gait in PD were undertaken to quantify kinematic characteristics of the Parkinsonian gait in order to investigate the underlying mechanism of PD [3±8]. Typical features of the Parkinsonian gait are small shuing steps and a general poverty and slowness of movement, called hypokinesia. Compared to healthy elderly subjects, patients with PD demonstrate reduced stride length and walking velocity during free ambulation. Double support duration and cadence rate are increased [3±8]. However, Blin [9] suggested that the basic gait patterns are preserved in PD.
0268-0033/01/$ - see front matter Ó 2001 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 8 - 0 0 3 3 ( 0 1 ) 0 0 0 6 9 - 9
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
Blin [9] and Hausdor [6] investigated the variability of the Parkinsonian gait and found higher variability values for spatial temporal gait parameter including stride length and double support duration. It seems that PD patients have diculties combining single steps to an accurate sequence of movement. This can be a result of disturbed electromyographic activity of the lower leg muscles [10,11], caused by diculties in coordinating submovements or due to a lack of the basal ganglia in providing accurate internal stimulus [4,12]. An increased variability of gait, especially of double support duration, is assumed to be related to an impaired balance control in PD [13]. PD patients show signi®cant abnormalities in foot strike [14±18]. Therefore, they face a higher risk of falling. In normal gait, the foot strikes the ground with the heel ®rst, followed by the mid- and forefoot. This procedure is called heel to toe walking or heel to toe motion. It was shown that PD patients in more advanced stages of the disease tend to strike the ground simultaneously with the heel and the forefoot resulting in a ¯at foot strike [16]. Sometimes even the forefoot strikes the ground ®rst (toe-heel-walking). In addition PD patients have a reduced foot lifting during the swing phase [5], which produces a smaller clearance between the toes and the ground. Whereas the kinematics of gait in PD is frequently documented only little is known about the kinetic features of the Parkinsonian gait. The vertical ground reaction force (GRF) component in normal gait has typically two peaks separated by a force minimum in the middle part of the GRF curve. The ®rst peak occurs with heel strike while the second peak is caused by the push-o from the ground. Force reduction in the middle part of the force curve becomes more pronounced with higher walking speeds. This is a consequence of GRF reduction by knee ¯exion of the supporting leg and lift of the center of mass (CoM) by the swinging leg. It was found that the shape of the vertical GRF signal is abnormal in PD [15,17]. Using force plate measurements reduced forces were found for heel contact and the push-o phase [15]. However, the shape of the vertical GRF signal, including two marked peaks, was similar for PD patients with moderate gait disorder as compared to elderly controls. PD patients with the typical shuing gait show only one narrow peak in vertical GRF signal [17]. Nieuwboer et al. [14] investigated the heel to toe motion of the foot in PD patients using pressure sensitive insoles. They found that PD patients demonstrate reduced forces at the heel and forefoot and increased load under the midfoot. Furthermore, the pathway of the center of force was found to start more towards the midfoot in PD patients. Additionally, the end of the pathway tends to be located less far forward for PD patients. Therefore, it was concluded that PD patients in later stages of the disease have a reduced heel to toe motion of the foot and that
807
the load shift towards the midfoot could be due to a compensation for possible loss of postural control. However, only little is known about heel to toe motion in Parkinson patients in a mild or moderate stage of the disease. Furthermore, the role of dierent heel to toe motion strategies in compensating for instability during walking is not yet fully understood. Therefore, the aim of this study was to investigate the heel to toe motion characteristics in Parkinson's patients during walking, especially the aspects of anterior±posterior and medial± lateral control. 2. Methods 2.1. Subjects Twenty-four patients (17 male and 7 female) with a diagnosis of idiopathic Parkinsonism (PD) participated in this study. The mean age of the PD group was 63.75 (SD, 10.05) years and the mean weight was 78.67 (SD, 13.61) kg. All patients lived independently in their own apartments and could ambulate freely without any help or walking aids. A neurologist examined the patients in order to classify the severity of the disease according to the Webster score [8]. The Webster score evaluates 10 single categories. These categories are bradykinesia of hands, rigidity, posture, upper extremity swing, gait, tremor, facies, seborrhea, speech and self-care. Each category is tested separately and scored from 0 for no impairment to 3 for severest impairment. Adding the scores of each category results in the total Webster score. A total Webster score of 30 indicates the severest state of Parkinson disease. The Webster scores for the patients ranged from 3 to 17 and the group mean score was 9.96 (SD, 3.36). The PD group measurements were performed in the on-phase approximately 1 h after individual medication of each patient. During this period best medication status can be expected. No ``freezing'' eects or festination were observed. PD patients, showing other movement disorders in addition to the Parkinsonian movement impairment, were excluded from this study. 10 male and 14 female age- and weightmatched healthy adults belonged to the control group. These subjects had a mean age of 66.06 (SD, 9.16) and a mean weight of 75.5 (SD, 14.51). No subjects with foot deformities participated in the study. 2.2. Instrumentation and procedure The Pedar insole system (NOVEL GmbH, Munich, Germany) was used to determine the in-shoe pressure distribution during normal gait of all participants. The system is based on ¯exible insoles with 99 capacitive transducers, which were sampled at 50 Hz. The transducer signals were sent to an electronic processing unit
808
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
(<600 grams), which was fastened by a belt to the subjects' waists. The electronic unit was connected via a 10-m long cable to the computer. Five dierent sizes of insoles (European sizes; 36-37, 38±39, 40±41, 42±43 and 44±45) were used to detect plantar foot pressures. Measuring equipment and cable were arranged to minimize interference with the subjects' gait. The demographic data assessment, including a neurological examination of the patients, was performed prior to the gait measurements. Following instrumentation, the subjects were asked to walk a distance of 11 m at their preferred speed to obtain undisturbed individual gait pattern. All subjects used identical socks and the same kind of footwear, a simple canvas shoe without any support structures. 2.3. Data analysis Walking speed, cadence and the plantar pressure data for the right and left foot were collected throughout the 11-m walk. As PD patients have problems initiating and terminating gait, the ®rst and the last steps were excluded from the analysis. Only steps in the middle of the walk, representing ``individual normal walking'', were used for further evaluation. Parkinson's symptoms may occur on one side of the body. To determine the characteristic features of the Parkinsonian gait, it is important to look at the steps of the left and the right foot separately. Therefore, the data of ®ve single steps of the left and the right foot were processed for each participant, averaged and statistically analyzed. To determine dierences in foot loading between the PD patients and the control group, a mask of 10 different anatomical areas was de®ned with the NovelWin software and applied to all insole sizes (Fig. 1). From the pressure data the relative loads (RL) were calculated for each anatomical area separately. The
relative load analysis gives more detailed information about the characteristics of the heel to toe motion of the foot than the peak pressures, because the time domain of loading is considered. Furthermore, the relative load analysis largely compensates for individual dierences in subjects' body weight and ground contact times. Therefore, it is an appropriate method to compare walking pattern of patients and healthy subjects, even if they exhibit slight dierences in walking speeds [19]. To calculate the relative loads, the force time integral (FTI) of each anatomical area was divided by the sum of all force time integral values and multiplied by 100. Consequently, the relative loads were given in a percent value. In addition, the gait line of each single step was determined. The gait line represents the movement of the center of pressure (CoP) under the foot during the heel to toe motion. To describe a more medial or lateral orientation of the gait line a lateral-index (LI) was de®ned. The area between the gait line and the lateral border of the insole, as well as the length of the gait line were used to determine the LI. The LI was calculated by dividing the cm2 -value of the area by the squared value of the length of the gait line. A greater LI value indicates a more medially oriented path of the gait line. Furthermore, the intra-subject variability of all parameters was determined by calculating each subject's coecient of variability for the ®ve single steps. Simple regression analyses were used to analyze the relationship between plantar peak pressures and the severity of the disease (expressed by the Webster score). Regressions were undertaken on peak pressures in the heel region, the mid- and forefoot areas and for the hallux. The statistical analyses were performed with a statistics software package (StatView II). A simple analysis of variance (A N O V A ) was performed to determine whether signi®cant dierences exist between the Parkinson group and the control group. The Schee` test was used for post hoc testing. For multiple tests a Bonferroni correction was done. Signi®cance levels were chosen at 6 0:05 and 6 0:01, respectively. 3. Results
Fig. 1. Foot areas.
Table 1 summarizes the demographic and spatialtemporal data of the PD patients and the control group. Although not signi®cant, the walking speed and cadence rate were slightly lower for the PD patients. Probably as a consequence of the reduced walking speed, the PD group also shows a slightly higher mean ground contact time. An analysis of the GRF curves revealed that all controls and 20 patients exhibited pronounced double peak GRF curves, whereas four patients showed single peak GRF force curves.
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
809
Table 1 Demographic features and spatial-temporal data for PD patients and the control group [means; (SD)]
n Age Weight (kg) Walking speed (m/s) Cadence (step/min) Foot contact time (ms) Webster score
Parkinson patients
Control group
P
24 63.75 (10.05) 78.67 (13.61) 0.97 (0.33) 104.21(18.22) 724.28 (101.38) 9.96 (3.36)
24 66.08 (9.16) 75.5 (14.51) 1.1 (0.27) 109.69 (15.62) 689.42 (115.63) ±
± n.s. n.s. n.s. n.s. n.s. ±
3.1. Peak pressures The single values of peak pressures, relative loads and intra-individual variability of the relative loads were illustrated in Table 2. The plantar pressure analysis revealed similar pressure distribution pattern for PD patients and the control group. However, Parkinson patients have signi®cantly P 0:031 reduced peak pressures in the lateral heel region compared to the control group. In the medial heel region peak pressures ranged from 217.75 kPa for the patients to 242.05 kPa for the control group. Although this dierence in medial heel was not signi®cant P 0:08 a general trend towards lower peak pressures for PD patients in the heel region can be observed.
icant P 0:035. For the metatarsal heads and the toe regions no signi®cant dierences in RL were found for patients and the control group. Nevertheless, it seems that PD patients show a general load shift from the heel region towards the mid- and forefoot regions compared to the control group (Fig. 2). Unexpected results were found for the analysis of the intra-individual variability of the RL distribution. Whereas the variability in the heel region was slightly higher for patients, the variability in all mid- and forefoot areas was reduced for the PD group (Fig. 3). The
3.2. Relative loads and relative load variability The RL for the heel region are comparable to the results of the peak pressure distribution. While the RL for the lateral heel are signi®cantly lower for the PD group P < 0:01 the load reduction for the patients in M2 (medial heel) is only moderate. On the other hand, patients have higher RL in the midfoot areas. However, only the dierence in the medial midfoot area is signif-
Fig. 2. Relative loads for patients and controls. Single values are summarized in Table 2, (P < 0:05; P < 0:01).
Table 2 Plantar peak pressures, relative loads and variability (CoV) of relative loads [means; (SD)]
*
Masks (anatomical area)
Peak pressures (kPa)
Relative loads (%)
CoV relative load (%)
n 24
PD patients
Control
PD patients
Control
PD patients
Control
M1 (medial heel) M2 (lateral heel) M3 (medial midfoot) M4 (lateral midfoot) M5 (1. metatarsal head) M6 (2. metatarsal head) M7 (3.-5. metatarsal heads) M8 (hallux) M9 (2. toe) M10 (lateral toes)
217.75 (78.3) 216.29 (73.8) 80.69 (26.5) 132.12 (54.8) 217.23 (75.4) 223.02 (67.4) 218.09 (75.4) 205.59 (97.0) 123.58 (65.4) 101.84 (49.5)
242.05 (57.7) 245.83 (57.4) 75.83 (28.9) 142.40 (62.1) 202.74 (53.4) 244.89 (69.6) 242.74 (61.2) 228.61 (89.2) 132.60 (43.3) 114.79 (37.9)
16.81 (7.9) 15.72 (5.3) 3.39 (2.8) 12.43 (6.4) 11.97 (5.4) 10.76 (3.3) 17.78 (5.5) 5.37 (3.2) 2.81 (2.0) 2.98 (2.3)
17.58 (3.2) 18.96 (2.7) 2.41 (1.6) 11.58 (3.8) 10.48 (3.0) 10.49 (2.2) 18.56 (3.2) 4.6 (2.1) 2.54 (1.0) 2.81 (1.6)
18.81 16.77 25.23 27.71 25.58 12.34 16.51 24.18 20.14 25.76
16.78 14.96 33.32 29.34 31.23 12.88 19.23 29.64 21.32 27.46
P 6 0:05. P 6 0:01.
**
(10.8) (9.7) (15.0) (15.7) (10.6) (6.8) (9.1) (12.1) (11.7) (13.7)
(16.8) (8.5) (18.0) (12.2) (15.3) (6.7) (9.1) (13.6) (12.0) (14.2)
810
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
(Webster score). A higher Webster score characterizes an advanced severity of the disease. A strong negative relationship r 0:58 between the peak heel pressures and the Webster score was found P 0:002. A decrease of heel loading occurs with an increase of the severity of the disease (Fig. 4). 4. Discussion Fig. 3. Variability of the relative loads for patients and controls. Single values are summarized in Table 2, P < 0:05; P < 0:01.
reduction of RL for the patients was revealed to be signi®cant in the medial midfoot and forefoot regions (M3, P 0:019; M5, P 0:038; M8, P 0:033). 3.3. Center of pressure Considering the length of the gait line, only small, non-signi®cant dierences were observed for PD patients (184.4 mm) and the control group (179.6 mm) indicating similar heel to toe motion pattern of the foot in anterior±posterior direction for the PD patients compared to the control group. However, the LI analysis revealed a dierent result. Whereas the mean LI for the PD group was 0.185, a LI value of 0.175 was calculated for the control group. The greater LI value for the PD group indicates a more medially oriented path of the gait line for the patients as compared to the control group P 0:048. 3.4. Relationship between clinical rating and gait parameter A simple regression analysis was performed to relate the kinetic gait parameters to the clinical rating score
Fig. 4. Regression curve of signi®cant correlation P < 0:01 between the Webster score and the heel±peak±pressures.
PD patients show slight, non-signi®cant reductions in walking speed and cadence, while total foot contact time was slightly increased (Table 1). These ®ndings are consistent with results of other studies investigating free ambulation in PD [4±6,8,9,14]. However, none of these dierences were signi®cant, suggesting a similar performance of the PD patients and the control group. A likely explanation for the similar performance may be the low severity of the disease in the participating PD patients. All patients were estimated to have only mild or moderate Parkinsonism [mean Webster score 9.96 (3.36)], living independently and managing life on their own. In order to analyze the kinetic characteristics of the heel to toe motion of the foot in PD, a mask of 10 different foot areas was applied, and pressure data were calculated for each area separately. The plantar pressure patterns result from passive and active foot functions during the heel to toe motion. Whereas the peak pressures under the heel and midfoot regions are more passively in¯uenced by accepting body weight at heel strike and midstance, the pressures at the forefoot are more likely to be in¯uenced by additional superimposing activity of the foot muscles. In our study the peak pressure distribution was similar for PD patients and the control group. However, in the heel region peak pressures are reduced for PD patients (M2; P 0:03). The reduction in heel peak pressures is an indication of a less pronounced heel strike in PD, which is similar to that reported previously [14,16]. Furthermore, this unloading of the heel is strongly related to the severity of the disease. A linear regression analysis revealed that a decrease of peak pressures in the heel region occurred with an increase of the severity of the disease indicated by a higher Webster score (Fig. 4). To analyze dierent heel to toe motion strategies during walking, the relative load analysis is the appropriate method, since this procedure largely compensates for dierences in ground contact times. This analysis suggests that PD patients in our study prefer a more forward oriented foot-loading pattern towards the midfoot and forefoot (Fig. 3). Nieuwboer et al. [14] found signi®cantly higher midfoot loading for PD patients compared to elderly controls. They suggested a more equal spread of force, and concluded that this characteristic feature might be explained by a compen-
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
satory mechanism for possible loss of postural control. However, as discussed above, the loads under the heel and midfoot region are more passively induced rather than actively in¯uenced by muscular activity. Compared to limited correction possibilities of the heel and midfoot, the ¯exible forefoot structure allows much more muscular control to compensate for instability. Therefore, it can be speculated that PD patients use a strategy of forefoot balance control during walking and that a higher midfoot loading appears as a consequence of this balance strategy. In addition, the body is in an unstable position during walking, when the heel of one leg strikes the ground and the toes of the other leg are still in contact with the ¯oor (beginning of double support). Thus, PD patients may avoid such an unstable position by shifting the load towards the forefoot and producing ¯at foot strikes with a premature contact of the forefoot with the ground. This eect could re¯ect an adaptive mechanism of PD patients to compensate for instability by generating a stable forefoot placement. A stable forefoot placement enables PD patients to gain additional safety by active muscular control. Beside a load transfer towards the forefoot a general medial load shift during the heel to toe motion was observed for the PD patients. The gait line in PD patients proceeds signi®cantly more medially as compared to elderly controls. This is demonstrated by a higher lateral index for PD patients, suggesting that the tendency to walk more on the medial site of the foot is a characteristic feature of Parkinsonian gait. The medial oriented heel to toe motion of the foot in PD and the tendency towards higher forefoot loading during gait in Parkinson patients as it was found in this study, probably re¯ect a compensatory mechanism for postural instability during walking. The variability analysis revealed that for PD patients the variability of the relative loads in the mid- and forefoot regions was reduced compared to elderly controls (Fig. 4). This was an unexpected ®nding as an increased variability of spatial temporal gait parameters is considered to be an expression of Parkinsonian movement disorder [6,9] and seems to be related to falls [13,20]. However, our results show, that PD patients in their ``on-phase'' generate highly stereotypical mid- and forefoot loading pattern, especially in the medial midand forefoot foot areas (M3; M5; M8). Cavanagh et al. [21] investigated the variability of foot loading in diabetic patients with neuropathy and healthy controls. It was found that the medial forefoot areas (Metatarsal head I, Hallux) exhibit the highest variability in plantar loading compared to other foot structures. The authors concluded that the variability in loading re¯ects the use of the medial column of the foot during walking and that step to step variations in acceleration and small changes in direction during gait are partly controlled by the medial ray of the foot. As the authors found no
811
dierences in loading variability between neuropathic patients with insensate feet and controls, they further concluded that this ``micro-control'' of gait is presumably centrally mediated. The results of our study con®rm the conclusions of Cavanagh et al. The variability of the relative load in PD patients is signi®cantly reduced in the medial foot areas (M3; M5; M8). Therefore, in Parkinson patients the micro-control of gait performed by the medial ray of the foot is characterized by stereotypical loading pattern, which is probably caused by basal ganglia dysfunction. This pattern may results in a reduced adjustability of the micro-control mechanism to compensate for external and internal changes during gait and is therefore maybe one risk factor for falls in Parkinson patients. 5. Conclusion The results of this study indicate that PD patients, even at a mild or moderate stage of the disease, show signi®cant changes in the heel to toe motion of the foot. PD patients tend to have a less pronounced heel strike. This mechanism was found to be related to the severity of the disease. Furthermore, a tendency towards higher forefoot loading and a medial load shift were determined for the PD patients. These characteristics of Parkinsonian gait are probably caused by an adaptive mechanism of PD patients to avoid unsteadiness during walking. This mechanism is stereotypical and therefore may less adjustable to compensate for external and internal changes during gait. The determination of foot loading patterns using plantar pressure measurement may be used as a valuable tool for diagnostic, treatment and rehabilitation purposes. The strategy of the forefoot control in PD patients should be considered in PD research and rehabilitation of Parkinsonian gait.
Acknowledgements This study was supported by: Alfried Krupp Krankenhaus ± Neurologie; Essen, Germany Gertrudis-Klinik; Biskirchen, Germany Klinik Ambrock ± Neurologie; Hagen, Germany.
References [1] Aita JF. Why patients with Parkinson's disease fall. Jama 1982;247:515±6. [2] Koller WC, Glatt S, Vetere-Over®eld B, Hassanein R. Falls and Parkinson's disease. Clin Neuropharmacol 1989;12:98±105. [3] Morris M, Iansek R, Matyas T, Summers J. Abnormalities in the stride length-cadence relation in parkinsonian gait. Mov Disord 1998;13:61±9.
812
S. Kimmeskamp, E.M. Hennig / Clinical Biomechanics 16 (2001) 806±812
[4] Morris ME, Iansek R, Matyas TA, Summers JJ. Stride length regulation in Parkinson's disease. Normalization strategies and underlying mechanisms. Brain 1996;119:551±68. [5] Murray MP, Sepic SB, Gardner GM, Downs WJ. Walking patterns of men with parkinsonism. Am J Phys Med 1978;57:278± 94. [6] Hausdor JM, Cudkowicz ME, Firtion R, Wei JY, Goldberger AL. Gait variability and basal ganglia disorders: stride-to-stride variations of gait cycle timing in Parkinson's disease and Huntington's disease. Mov Disord 1998;13:428±37 [published erratum appears in Mov Disord 1998 Jul;13:757]. [7] Zijlstra W, Rutgers AW, Van Weerden TW. Voluntary and involuntary adaptation of gait in parkinson's disease. Gait Posture 1998;7:53±63. [8] Vieregge P, Stolze H, Klein C, Heberlein I. Gait quantitation in Parkinson's disease ± locomotor disability and correlation to clinical rating scales. J Neural Transm 1997;104:237±48. [9] Blin O, Ferrandez AM, Serratrice G. Quantitative analysis of gait in parkinson patients: increased variability of stride length. J Neurol Sci 1990;98:91±7. [10] Dietz V, Zijlstra W, Prokop T, Berger W. Leg muscle activation during gait in parkinson's disease: adaptation and interlimb coordination. Electroencephalogr Clin Neurophysiol 1995;97:408±15. [11] Miller RA, Thaut MH, McIntosh GC, Rice RR. Components of EMG symmetry and variability in parkinsonian and healthy elderly gait. Electroencephalogr Clin Neurophysiol 1996;101:1±7. [12] Albin RL, Young AB, Penney JB. The functional anatomy of disorders of the basal ganglia. Trends Neurosci 1995;18:63±4.
[13] Gabell A, Nayak US. The eect of age on variability in gait. J Gerontol 1984;39:662±6. [14] Nieuwboer A, De Weerdt W, Dom R, Peeraer L, Lesare E, Hilde F, et al. Plantar force distribution in Parkinsonian gait: a comparison between patients and age-matched control subjects. Scand J Rehabil Med 1999;31:185±92. [15] Koozekanani SH, Balmaseda Jr. MT, Fatehi MT, Lowney ED. Ground reaction forces during ambulation in parkinsonism: pilot study. Arch Phys Med Rehabil 1987;68:28±30. [16] Hughes JR, Bowes SG, Leeman AL, O'Neill CJ, Deshmukh AA, Nicholson PW, et al. Parkinsonian abnormality of foot strike: a phenomenon of ageing and/or one responsive to levodopa therapy?. Br J Clin Pharmacol 1990;29:179±86. [17] Ueno E, Yanagisawa N, Takami M. Gait disorders in parkinsonism a study with ¯oor reaction forces and EMG. Adv Neurol 1993;60:414±8. [18] Diehl RR, Schneider U, Konietzko M, Hennerici M. Quantitative analysis of gait in neurological gait disorders. In: Woollacott MJ, Horak FB, editors. Posture and gait control mechanisms. Portland: University of Oregon Books; 1992. p. 47±50. [19] Hennig EM, Staats A, Rosenbaum D. Plantar pressure distribution pattern of young school children in comparison to adults. Foot Ankle 1994;15:35±40. [20] Maki BE. Gait changes in older adults: predictors of falls or indicators of fear. J Am Geriatr Soc 1997;45:313±20. [21] Cavanagh PR, Perry JE, Ulbrecht JS, Derr JA, Pammer SE. Neuropathic diabetic patients do not have reduced variability of plantar loading during gait. Gait Posture 1998;7:191±9.