Journal of Movement ORIGINAL ARTICLE Disorders Journal of Movement Disorders Vol. 1, No. 2, October 2008 Received October 24, 2007 Accepted May 19, 2008

Gait Analysis in Patients With Parkinson’s Disease: Relationship to Clinical Features and Freezing

Seong-Beom Koh, MD, PhDa, Kun-Woo Park, MD, PhDa, Dae-Hie Lee, MD, PhDa, Se Ju Kim, MD, PhDb, Joon-Shik Yoon, MD, PhDb

aDepartment of Neurology, Korea University College of Medicine bDepartment of Physical Medicine and Rehabilitation, Korea University College of Medicine

Background: The purpose of our study was to investigate gait dynamics and kinematics in patients with Parkinson’s disease (PD) and to correlate these features with the predominant clinical features and with the presence of the freezing of gait (FOG). We measured the temporospatial and kinematic parameters of gait in 30 patients with PD (M:F=12:18, age=68.43±7.54) using a computerized video motion analysis system. Methods: W e divided the subjects into subgroups: (1) -dominant (TD) group and postural instability and gait disturbance (PIGD) group and (2) FOG group and non-FOG group. We compared the gait parameters between the subgroups. Results: The velocity and stride length were reduced significantly in the PIGD group compared to the TD group. The PIGD group showed a significantly reduced range of motion in the pelvic and lower extremity joints by kinematics. Stride time variability was significantly increased and the pelvic oblique range was significantly reduced in the freezing gait disorder group. Conclusion: Our findings suggest that there are differences in the perturbation of the -cortical circuits based on major clinical features. The reduction of the pelvic oblique range of motion may be a compensatory mechanism for postural instability and contributes to stride time variability in patients with FOG. Journal of Movement Disorders 1(2):59-64, 2008

Key Words: Gait, Parkinson’s disease, Freezing

INTRODUCTION with bradykinesia, resting tremor and rigidity. Gait disturbance and postural instability have been recognized Patients with Parkinson’s disease (PD) often present with as important motor signs of PD. However, such clinical difficulty in initiating and/or maintaining normal walking. features do not appear in all patients with PD. Patients The disturbance of routine gait activity may cause sub- with PD can be classified by major clinical features. Jankovic stantial discomfort and impairment in activities of daily et al. classified patients with PD as a tremor-dominant living. Gait requires a complex set of behaviors which (TD) group or a postural instability and gait disturbance include the simultaneous performance of locomotion while (PIGD) group.[2] About one-third of patients with PD maintaining balance. Gait disturbance is a common feature have sudden and transient disturbance in motor perfor- of movement disorders, especially in PD. Careful obser- mance, known as the ‘freezing phenomenon’.[3,4] Freezing vation and analysis of gait patterns frequently helps diagnose of gait (FOG), occurs when a patient cannot walk forward, patients with movement disorders.[1] as if the feet were stuck to the ground. It is a unique PD is primarily a disturbance of motor function typically symptom that frequently causes patients to fall. FOG is

∙ Address for correspondence: Joon-Shik Yoon MD, PhD Department of Physical Medicine & Rehabilitation, Korea University College of Medicine at Guro Hospital, 80 Guro-dong, Guro-gu, Seoul, 152-703, Korea Tel: +82-2-818-6251, Fax: +82-2-859-5422, E-mail: [email protected] www.km ds.or.kr 59 Journal of Movement Disorders Vol. 1, No. 2, 2008

a common and disabling gait problem in patients with PD, as determined by history; tremor of face, lips or chin at especially in the advanced stages.[5] FOG, in off-phase rest; tremor of all four limbs at rest; action or postural PD, responds well to levodopa suggesting a dopa tremor in both arms as determined by the investigator’s deficiency with resulting motor circuit dysfunction as the examination. A mean score for the complex of postural underlying pathophysiological mechanism; however, instability and gait difficulty (PIGD) was calculated as the levodopa is not effective for treatment of FOG in advanced mean of the following items: falling, freezing, walking stage PD or in other parkinsonian syndromes. Recently, difficulty by history as well as gait and postural instability quantitative gait analysis studies suggested that FOG is by examination. The tremor - dominant group (TD) caused by a combination of increased inability to generate included patients with a ratio of the mean tremor stride length superimposed on an unregulated walking score/mean PIGD score greater than or equal to 1.5; cadence.[6] the PIGD group included all patients with a ratio of The main purpose of this study was to investigate the less than or equal to 1.0.[2] Patients were divided into temporospatial and kinematic patterns of gait with a three a freezing group and a non-freezing group based on dimensional gait analysis system in patients according clinical history and the UPDRS motor score. to their clinical features comparing (1) the PIGD group with the TD group, and (2) the freezing group with the 2. Gait Protocols and Instrumentation non-freezing group. All gait evaluations were performed and recorded by MATERIALS AND METHODS an identical method. The standing posture was used as the baseline standard. Subjects were instructed to walk at 1. Subjects their usual pace on level ground for 10 m, turn and walk the same route back. This was repeated until identical and We recruited patients from the Movement Disorders Unit complete walking information was obtained. at the Department of Neurology at Korea University We performed the gait analysis with an automatic com- Guro Hospital. PD was defined according to the Clinical puterized video motion analysis system (Oxford Metrix Diagnostic Criteria described by the United Kingdom vicon 512 motion analysis system). Based on biomechanical Parkinson’s Disease Society Brain Bank.[7] We excluded models, the pelvis and the lower extremities of each patients who reported clinically significant co-morbid subject were measured by 15 retro-reflective markers. disorders likely to affect gait including: , orthopedic These markers were attached to the anterior superior iliac disease, rheumatologic disease, cardiovascular disease spine, lateral aspect of the thighs and shin, knee jointaxis, and pulmonary disorders. Patients who could not perform lateral malleoli, heels, forefeet and the sacrum area. A the three dimensional gait analysis because of cognitive computerized system recorded the information from the impairment (Mini Mental Status Examination < 24) were retro-reflective markers automatically at a speed of 30 Hz. also excluded. The recruited patients participated in the This information was then transferred to a workstation. gait analysis at least 24 hours after taking anti-Parkinson The temporospatial parameters were reconstructed and medication, while in the “off” status. Written informed analyzed automatically at the workstation. Stride time consent was obtained from each patient for study partici- variability, the magnitude of the stride-to-stride fluctuations pation. This study was approved by the Institutional Review in the gait cycle duration, was calculated by determining Board at Korea University Guro Hospital. the standard deviation and the coefficient of variation An average global tremor score was calculated as the of each subject’s stride time.[8] mean of the following items: right and left arm tremor,

www.km ds.or.kr 60 Gait Analysis in Patients W ith Parkinson’s Disease: Relationship to Clinical Features and Freezing∣Seong-Beom Koh, et al.

3. Data analysis male : female = 4 : 5), and the PIGD group had 21 patients (69.14±8.14, male : female = 8 : 13). The freezing group Spatiotemporal and kinematic parameters were collected included six patients (68.67±10.41, male : female = 3 : 3), and analyzed for the entire stride period. The maximal and the non- freezing group had 24 patients (68.38±6.94, value of the amplitude for the motion of the pelvis, hip, male : female = 9 : 15). The demographic and clinical charac- knee and ankle were measured directly, and the range of teristics of patients are summarized in Table 1. motion of each joint was calculated. An independent t-test was used to compare each group (TD/PIGD and freezing 2. The tremor-dominant patient (TD) group and group /non-freezing group). Statistical results with a p-value the postural instability and gait disturbance- < 0.05 were considered significant. Statistical analysis was dominant patient (PIGD) group performed with Statistical Package for Social Science Software (version 10.0, SPSS, Chicago, IL) for Windows. 1) Spatiotemporal results The patients in the PIGD group had a significant RESULTS reduction in walking velocity, stride length and ratio of single support time to double support time compared to 1. Demographic Factors the TD group. However, the cadence, single support time and double support time were not significantly different A total of 30 patients completed the study. Twelve males (Table 2). and 18 females had a mean age of 68.8±7.49 years. Hoehn and Yahr stage I included 10 patients, II had 11 patients, 2) Kinematic results III had nine patients and IV had one patient. Among 30 For the PIGD group, the range of pelvic oblique, hip patients, the TD group included nine patients (66.78±6.02, extension, ankle plantar flexion and knee flexion during

Table 1. The demographic and clinical characteristics of enrolled patients

Subgroup 1 Subgroup 2 PD TD PIGD Freezing Non-Freezing (n=9) (n=21) (n=6) (n=24) Gender (male : female) 12：18 4：58：13 3：39：15 Age (years) 68.43±7.54 66.78±6.02 69.14±8.14 68.67±10.41 68.38±6.94 H-Y stage 2.07±0.83 1.56±0.53 2.29±0.53 3±1.03 1.9±0.72 UPDRS, part III 19.37±12.57 11.22±5.33 22.86±13.24 32.67±16.08 16.04±9.24 Body Weights (Kg) 60.17±9.83 63.33±10.59 58.81±9.42 54.5±3.56 61.58±10.42

Table 2. The profile of temporospatial gait analysis parameters in the tremor dominant group (TD) and the postural instability dominant group (PIGD)

TD PIGD p value Cadence (steps/min) 108.33±10.4 108.7±16.7 NS Walking Velocity (m/s) 0.97±0.18 0.76±0.25 p<0.05 Stride length (m) 1.06±0.12 0.81±0.26 p<0.05 Single support (%) 0.41±0.03 0.42±0.06 NS Double support (%) 0.27±0.05 0.31±0.08 NS S/D 1.61±0.28 1.36±0.24 p<0.05 S/D=single support time/double support time. s; seconds, m; meter, NS; non signigicant. www.km ds.or.kr 61 Journal of Movement Disorders Vol. 1, No. 2, 2008

the swing phase (SW) were significantly reduced. How- The stride time variability was significantly increased ever, the range of pelvic tilting, hip flexion and knee flexion in the freezing group. However, there were no significant during terminal stance (TS) and ankle dorsiflexion were differences in the walking velocity, walking speed, stride not significantly different (Table 3). length, the single support time and the double support time between the freezing group and the non-freezing group 3. The patient group with the freezing phenomenon (Table 4). and the patient group without the freezing phenomenon 2) Kinematic results In the freezing group, the pelvic oblique motion was 1) Spatiotemporal results significantly decreased. However, other kinematic para-

Table 3. The profile of kinematic parameters in the tremor dominant group (TD) and the postural instability dominant group (PIGD).

TD PIGD p value Pelvic-Tilt (degree) 6.51±2.39 7.33±3.54 NS Pelvic-obliquity (degree) 2.91±1.53 1.87±1.73 p<0.05 Hip-flexion (degree) 27.56±4.36 25.19±6.48 NS Hip-extension (degree) 14.2±3.78 7.81±8.71 p<0.05 Knee- TS (degree) 5.17±4.66 8.07±6.40 NS Knee- SW (degree) 52.67±5.15 46.14±8.68 p<0.05 Ankle-DF (degree) 16.89±2.03 16.24±2.72 NS Ankle-PF (degree) 49.08±3.20 6.50±2.71 p<0.05 TS; terminal stance, SW; swing, DF; dorsiflexion, PF; plantar flexion.

Table 4. The profile of temporospatial gait analysis parameters in the non-freezing group and the freezing group

Non-freezing Freezing p value Cadence (steps/min) 113.00±27.56 107.5±10.58 NS Walking Velocity (m/s) 0.65±0.36 0.87±0.20 NS Stride length(m) 0.62±0.32 0.95±0.19 p<0.05 Single support (%) 0.39±0.07 0.42±0.04 NS Double support (%) 0.33±0.08 0.29±0.06 NS S/D 1.18±0.30 1.45±0.27 NS Stride time variability 0.18±0.02 0.23±0.09 p<0.05 S/D=single support time/double support time. s; seconds, m; meter, NS;non significant.

Table 5. The profile of kinematic parameters in the non-freezing and the freezing group

Non-freezing Freezing p value Pelvic-Tilt (degree) 6.75±2.79 8.47±4.61 NS Pelvic-obliquity (degree)* 2.58±1.68 0.57±0.38 p<0.05 Hip-flexion (degree) 26.13±4.61 25.00±10.14 NS Hip-extension (degree) 11.68±5.03 10.10±3.96 NS Knee-TS (degree) 6.58±5.38 9.72±8.17 NS Knee-SW (degree) 49.88±6.44 41.00±11.45 NS Ankle-DF (degree) 16.46±2.59 16.33±2.42 NS Ankle-PF (degree) 5.64±3.36 0.53±5.05 NS SW; swing, TS; terminal stance, DF; dorsiflexion, PF; plantar flexion, NS; non significant. www.km ds.or.kr 62 Gait Analysis in Patients W ith Parkinson’s Disease: Relationship to Clinical Features and Freezing∣Seong-Beom Koh, et al.

meters (pelvic tilting, range of motion of hip and ankle) phenomenon.[18] Hausdorff et al. suggested that PD were not significantly different (Table 5). patients with FOG had a continuous gait disturbance: an inability to regulate the stride-to-stride variations in gait DISCUSSION timing. In our study, the stride time variability, which may suggest gait irregularity and instability, was increased in Patients with PD have distinctive features of decreased the freezing group.[19] These results suggest that FOG walking velocity, reduced stride length and intact control does not suddenly develop but rather is a progressive gait of cadence in their gait dynamics.[9,10] In addition, during disorder that reflects damage to the control of the stride a gait cycle, the ratio of double support time to single to stride gait fluctuation. Therefore, we could detect gait support time increases in patients with PD compared to disturbances between episodes of the freezing phenomenon, normal controls. Previous studies have reported that and during the freezing phenomenon with a computerized stimulation of the internal mediates a limited three dimensional gait analysis system. In addition, we affect on the control of stride length however, stimulation observed that the pelvic oblique movements were reduced of the has been shown to mediate in the freezing group. This was thought to be a compensatory a significant affect on stride length. Thus the role of the mechanism to overcome the freezing phenomenon and to subthalamic nucleus in the control of stride length may maintain postural stability. be substantial.[11-13] Therefore, excessive activity in the In conclusion, we detected changes in walking velocity subthalamic nucleus, which is observed in PD, may cause and stride length and decreased range of motion in the the abnormality in control of stride length. Morris et al. PIGD group. This was thought to be caused by abnormal have reported that the basal ganglia do not control stride regulation of the cerebral cortex-basal ganglia. Stride length directly, and that the motor cortical regions such timee variability was increased and the pelvic oblique range as the primary motor cortex, the premotor cortex, and of motion was decreased in the freezing group. It is thought the supplementary motor areas determine the range and that FOG is not a distinct phenomenon but rather a type order of motion depending on the environment and work of gait disorder, with the loss of the ability to control demands.[14,15] Functional imaging study has suggested changes of stride length, to compensate for such changes a difference between tremor-dominant PD and PIGD- and to maintain postural stability. dominant PD with regard to regional patterns of blood flow reduction in the brain.[16] In our study, the walking REFERENCES cadence of the PIGD group compared to the TD group 1. Chang R, Guan L, Burne JA. An automated form of video was not altered; nonetheless, a noticeable reduction of the image analysis applied to classification of movement walking velocity and stride length could be detected. With disorders. Disabil Rehabil 2000;22:97-108. these results, we would predict a difference in the pattern 2. Jankovic J, Kapadia AS. Variable expression of Parkinson’s disease: a base-line analysis of the DATATOP cohort. The of perturbation in the basal ganglia-cortical circuits, especially Parkinson Study Group. Neurology 1990;40:1529-34. in the subthalamic nucleus, between tremor-dominant PD 3. Giladi N, McDermott MP, Fahn S, Przedborski S, Jankovic J, Stern M, et al. Freezing of gait in PD: prospective and PIGD-dominant PD. assessment in the DATATOP cohort. Neurology 2001;56: In mild cases of PD, freezing of gait (FOG) is found 1712-1721. in approximately seven percent of patients. Fifty percent 4. Giladi N, Treves TA, Simon ES, Shabtai H, Orlov Y, Kandinov B, et al. Freezing of gait in patients with advanced of FOG patients have more advanced PD.[17] The Parkinson’s disease. J Neural Transm 2001;108:53-61. mechanisms underlying FOG has not been characterized 5. Garcia Ruiz PJ, Meseguer E, Del Val J, Vazquez A, Sanchez Bernardos V, Vazquez A. Motor complications in to date. Nieuwboer et al. has suggested that gait disturbance Parkinson disease: a prospective follow-up study. Clin could be detected prior to the development of the freezing Neuropharmacol 2004;27:49-52.

www.km ds.or.kr 63 Journal of Movement Disorders Vol. 1, No. 2, 2008

6. Nieuwboer A, Dom R, De Weerdt W, Desloovere K, stimulation on gait kinematics and kinetics in Parkinson’s Fieuws S, Broens-Kaucsik E. Abnormalities of the spatio- disease. Exp Brain Res. 2005 Jan;160:517-27. Epub 2004 temporal characteristics of gait at the onset of freezing in Oct 22. Parkinson’s disease. Mov Disord 2001;16:1066-1075. 13. Ferrarin M, Rizzone M, Lopiano L, Recalcati M, Pedotti 7. Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of A. Effects of subthalamic nucleus stimulation and L-dopa clinical diagnosis of idiopathic Parkinson’s disease: a clinico- in trunk kinematics of patients with Parkinson’s disease. pathological study of 100 cases. J Neurol Neurosurg Gait Posture 2004;19:164-171. Psychiatry 1992;55:181-184. 14. Morris M, Iansek R, Matyas T, Summers J. Stride length 8. Hausdorff JM, Cudkowicz ME, Firtion R, Wei JY, regulation in Parkinson’s disease. Normalization strategies Goldberger AL. Gait variability and basal ganglia disorders: and underlying mechanisms. Brain 1996;119:551-568. stride-to-stride variations of gait cycle timing in Parkinson’s 15. Morris M, Iansek R, Matyas T, Summers J. Abnormalities disease: stride-to-stride variations of gait cycle timing in in the stride length-cadence relation in parkinsonian gait. Parkinson’s disease and Huntington’s disease. Mov Disord Mov Disord 1998;13:61-69. 1998;13:428-437. 16. Mito Y, Yoshida K, Yabe I, Makino K, Tashiro K, Kikuchi 9. Murray MP, Sepic SB, Gardner GM, Downs WJ. Walking S, et al. Brain SPECT analysis by 3D-SSP and clinical patterns of men with parkinsonism. Am J Phys Med 1978; features of Parkinson’s disease. Hokkaido Igaku Zasshi 57:278-294. 2006;81:15-23. 10. O’Sullivan JD, Said CM, Dillon LC, Hoffman M, Hughes 17. Lamberti P, Armenise S, Castaldo V, de Mari M, Iliceto AJ. Gait analysis in patients with Parkinson’s disease and G, Tronci P, et al. Freezing gait in Parkinson’s disease. motor fluctuations: influence of levodopa and comparison Eur Neurol 1997;38:297-301. with other measures of motor function. Mov Disord 18. Nieuwboer A, Dom R, De Weerdt W, Desloovere K, 1998;13:900-906. Janssens L, Stijn V. Electromyographic profiles of gait prior 11. Allert N, Volkmann J, Dotse S, Hefter H, Sturm V, to onset of freezing episodes in patients with Parkinson’s Freund HJ. Effects of bilateral pallidal or subthalamic disease. Brain. 2004 Jul;127:1650-60. stimulation on gait in advanced Parkinson’s disease. Mov 19. Hausdorff JM, Schaafsma JD, Balash Y, Bartels AL, Disord 2001;16:1076-1085. Gurevich T, Giladi N. Impaired regulation of stride 12. Ferrarin M, Rizzone M, Bergamasco B, Lanotte M, variability in Parkinson’s disease subjects with freezing of Recalcati M, Pedotti A, et al. Effects of bilateral subthalamic gait. Exp Brain Res 2003;149:187-194.

www.km ds.or.kr 64