MORE THAN 85% OF STROKE survivors can eventually

562 ORIGINAL ARTICLE The Relation Between Ankle Impairments and Gait Velocity and Symmetry in People With Stroke Pei-Yi Lin, MS, PT, Yea-Ru Yang, PhD, PT, Shih-Jung Cheng, MD, Ray-Yau Wang, PhD, PT ABSTRACT. Lin P-Y, Yang Y-R, Cheng S-J, Wang R-Y. The relation between ankle impairments and gait velocity and symmetry in people with stroke. Arch Phys Med Rehabil 2006; 87:562-8. Objective: To identify the most important factor among the ankle impairments on gait velocity and symmetry in stroke patients. Design: Cross-sectional, descriptive analysis of convenience sample. Setting: Patients from outpatient rehabilitation and neurovascular neurology departments in medical centers and municipal hospitals in Taiwan. Participants: Sixty-eight subjects with hemiparesis poststroke with the ability to walk independently. Interventions: Not applicable. Main Outcome Measures: Maximal isometric strength of plantarflexors and dorsiflexors were examined by a handheld dynamometer. Spasticity index, slope magnitudes between electromyographic activities, and muscle lengthening velocity of gastrocnemius during lengthening period of stance phases were measured to represent the dynamic spasticity. Passive stiffness of pantarflexors was indicated by degrees of dorsiflexion range that were less than normative values. Position error was measured by the degree of proprioceptive deficits of ankle joint by evaluating the joint position sense. Gait velocity, symmetry, and other gait parameters were measured by the GAITRite system. Results: Regression analyses revealed that the dorsiflexors strength was the most important factor for gait velocity and temporal symmetry (R 2.30 for gait velocity, P.001; R 2.36 for temporal asymmetry, P.001). Dynamic spasticity was the most important determinant for gait spatial symmetry (R 2.53, P.001). Conclusions: Gait velocity and temporal asymmetry are mainly affected by the dorsiflexors strength, whereas dynamic spasticity of plantarflexors influenced the degree of spatial gait asymmetry in our patients who were able to walk outdoors. Treatment aiming to improve different aspects of gait performance should emphasize on different ankle impairments. Key Words: Ankle; Gait disorders, neurologic; Rehabilitation; Stroke. 2006 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation From the Institute and Faculty of Physical Therapy, National Yang-Ming University, Taipei, Taiwan (Lin, Yang, Wang); and Department of Neurology, Mackay Memorial Hospital, Taipei, Taiwan (Cheng). Supported by the National Research Institute (grant no. NHRI-EX94-9413EI) and National Science Council (grant no. NSC93-2321-B-010-010), Republic of China. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. Reprint requests to Ray-Yau Wang, PhD, PT, 155, Sec 2, Li Nong St, Shih-Pai, Taipei, Taiwan, e-mail: rywang@ym.edu.tw. 0003-9993/06/8704-10301$32.00/0 doi:10.1016/j.apmr.2005.12.042 MORE THAN 85% OF STROKE survivors can eventually walk with or without assistance. 1,2 The common features of walking after stroke include decreased gait velocity and asymmetrical gait pattern. 3-6 Achieving normal gait patterns and speed are usually the ultimate goals of gait training. Although the causes of gait deviations may vary from patient to patient, several studies have examined the possible impairments affecting gait velocity during the last 2 decades. 7-12 Muscle weakness, abnormal muscle activities, inadequate muscle coactivation, sensory and visual deficits, noncontractile soft-tissue tightness, and disruption in central generation of programmed muscle activation were suggested to be the impairment factors that result in decreased gait velocity. 7-12 However, ankle impairment on gait performance has not received its deserved attention for stroke patients. Adequate ankle control during gait is important for normal gait pattern. During the push-off phase, the plantarflexors generate a large part of energy required to move the limbs forward. 13 Insufficient plantarflexors power during gait leading to decreased gait velocity has been widely observed and documented for patients with stroke. 13-16 Lamontagne et al 17 reported that half of their subjects with stroke had reduced dorsiflexion during swing phase of the affected side compared with control values. Inadequate dorsiflexion control during gait can be caused by many ankle impairments, such as weakness of dorsiflexors, spasticity of plantarflexors, passive stiffness of the plantarflexors, or ankle joint pathology. 18-21 The role of dorsiflexors strength for gait velocity has not been clearly verified. 18,19,22 Spastic gait has been widely reported in patients with stroke, but the correlations between spasticity and gait velocity have not been firmly established because of the measuring of the static spasticity instead of the dynamic spasticity. 23,24 Lamontagne et al 24 first measured spasticity of plantarflexors of subjects with hemiparesis during walking. This method for measuring dynamic spasticity has proved its significance on gait velocity. Compared with the nonaffected side or normative control, ankle passive stiffness of the affected side has been noted. 25-27 However, the resistive torque at ankle joint measured as passive stiffness of plantarflexors did not correlate significantly with gait velocity in stroke subjects. 25 Also, the influence of sensory deficits on gait is still controversial. It has been reported that the tactile and proprioception impairments of the affected leg influence the walking velocity. 12,23 But insignificant correlations between the Fugl-Meyer Assessment sensory score and gait velocity have been noted in other studies. 10,28 The importance of ankle control during gait has been emphasized. However, the ankle impairments of stroke patients on gait performance are still undetermined. Moreover, there is a lack of in-depth investigation on ankle impairments that affect gait symmetry. The purpose of this study was to identify the relation between ankle impairments, including strength of plantarflexors and dorsiflexors, dynamic spasticity of plantarflexors, passive stiffness of plantarflexors, ankle joint position sense, and gait velocity and symmetry in people with stroke.

ANKLE IMPAIRMENTS ON GAIT, Lin 563 METHODS Participants Subjects with hemiparesis resulting from stroke were recruited from rehabilitation and neurovascular neurology departments in medical centers and local hospitals in Taipei, Taiwan. The inclusion criteria were (1) a unilateral cerebral lesion confirmed by computed tomography or magnetic resonance imaging, (2) the ability to walk 10m independently without any gait aids, (3) the ability to plantarflex and dorsiflex the affected ankle actively in the supine position, and (4) the ability to follow verbal commands. Subjects were excluded if they had (1) unstable medical conditions; (2) histories of other neurologic or orthopedic problems known to affect gait performance; (3) brainstem, cerebellar, or subcortical lesion; (4) ankle dorsiflexion passive range of motion (PROM) of less than 0 ; (5) ankle joint pain; or (6) hemineglect. The experimental procedures were explained to all included subjects, who provided informed consent. The study protocol was approved by the Institutional Review Board of Taipei Veterans General Hospital. Protocol Sixty-eight subjects completed all experimental procedures. The general data including subjects age, sex, poststroke duration, and the side of paresis were collected from medical charts. Other general information for data analyses, including body weight for normalizing muscle strength, was recorded. We assessed gait performance, ankle muscle strength, dynamic spasticity, passive stiffness of plantarflexors, and ankle joint position sense in a random order. Measurements Gait performance. We used the GAITRite system a to measure gait velocity, symmetry, and other gait parameters. The standard GAITRite walkway contains 6 sensor pads with 13,824 sensors encapsulated in a roll-up carpet with an active area of 3.66m long and 0.61m wide. As the subject walks through the walkway, each footfall as a function of time transfers to a computer for analyzing the temporal and spatial gait parameters. Concurrent validity and reliability have been well established. 29,30 Subjects in the present study were asked to walk with their comfortable speed without gait aids through a 10-m hallway for 3 times. All 3 trials of walking were averaged. The GAITRite walkway was placed in the middle of the 10-m hallway to eliminate the effect of acceleration and deceleration. Gait velocity and other gait parameters, including cadence, stride length, step width, step length, cycle time, single-leg support time, and double-leg support time, were included as gait performance of our subjects. Temporal and spatial asymmetries were also calculated by the following formula. The greater value of the ratio, the greater degree of the asymmetry. Temporal asymmetry (single-support asymmetry ratio): single-support time (affected) 1 single-support time (unaffected) Spatial asymmetry (step-length asymmetry ratio): step length (affected) 1 step length (unaffected) Muscle Strength of Plantarflexors and Dorsiflexors The Power Track Il dynamometer b was used to measure the maximal isometric muscle strength of ankle plantarflexors and dorsiflexors. The testing position of our subjects for plantarflexor strength was supine with hips and knees flexed at 90 supported by a wooden block. The ankle joint was positioned in a neutral position. Stabilization was provided by straps around the pelvic region. The handheld dynamometer was held perpendicular to the foot at the metatarsal level. The testing position of subjects for dorsiflexor strength was supine with hips and knees extended. Straps around the waist and knee were used for stabilization, and the dynamometer sensor was placed over the tarsal bones for measuring dorsiflexor muscle strength. The intrarater reliability was.997 for plantarflexors and.993 for dorsiflexors in healthy adults in our pilot study. In response to a make-a-contraction instruction, each subject first performed a submaximal familiarization trial and then maximal isometric contraction for 5 seconds, 3 trials with a 15-second rest in between. After a 1-minute rest, measurement of the muscle group on the other side was conducted. The order of measurements was the unaffected side first and then the affected side to ensure the patients knew the correct muscle group to contract. The order for muscle groups tested was randomized. The average value of isometric maximal voluntary contraction (in kilograms) was then normalized by body weight for data analysis. Dynamic Spasticity of the Plantarflexors During Gait The spasticity index, measured by means of dynamic spasticity of plantarflexor muscle during gait, was defined as the value of the electromyography-lengthening slope during the lengthening period of stance phases. Because not all gait cycles were accompanied by a positive slope, we chose only the positive values to investigate the maximal influence of dynamic spasticity on gait deviations. Its use has been recently validated in subjects with hemiparesis by Lamontagne et al. 24 The electromyographic activity of medial gastrocnemius was recorded during gait by using an 11-mm Ag-AgCl electrode. c The skin was rubbed with alcohol to reduce impedance. An electrode was placed on the motor point of the muscles, which was located on 5 finger-width distal to the popliteal crease and 2cm medial to midline. 31 The 2 electronic goniometers d were placed separately on the knee and ankle joints of the affected side to measure the joint displacements during walking. For measuring ankle joint displacement, the alignment of 1 endblock was arranged paralleled with the line of head of the fifth metatarsal and lateral malleolus, and the other endblock paralleled with the line of head of fibula and lateral malleolus. 32 The 1 endblock of another goniometer for measuring knee-joint displacement was paralleled with the line of greater trochanter and lateral epicondyle of the femur. The other endblock was paralleled with the line of lateral malleolus and lateral epicondyle. 32 Two footswitches were placed under the heel and toe of the affected side to mark the stance and swing phases of the gait cycle. All signals were recorded by using AcqKnowledge software with BIOPAC MP150WMW System c (common mode rejection ratio, 104dB; analog-to-digital converter signal/noise ratio, 86dB; digital analog resolution, 16 bits; gain, 2000) with a sampling rate of 1000Hz and stored for offline analysis. The electromyographic activity was processed following Lamontagne s procedures. 24 The electromyographic activity of the medial gastrocnemius for lengthening period at stance phase was normalized to its maximal value of stance phase. The muscle-lengthening velocity of gastrocnemius was calculated by using the model developed by Winter and Scott. 33 Muscle activity was then plotted with muscle-lengthening velocity specific to each lengthening period of stance phase in gait cycle. The slope of these plots in each gait cycle was then

564 ANKLE IMPAIRMENTS ON GAIT, Lin as possible. The average differences in degrees between the unaffected and affected ankle position were represented as position error for data analysis. Statistical Analysis The SPSS e was used for statistical analysis. Stepwise regression analyses, performed by the forward stepping method, were used to identify the most important impairments for gait velocity and temporal and spatial asymmetry. The relations between gait parameters and impairment variables including muscle strength of plantarflexors and dorsiflexors, spasticity index, passive stiffness of plantarflexors, and position error were examined by Pearson correlation coefficients. Significance was set at less than.05. Fig 1. The spasticity index. Each point representing electromyographic activity (EMG) of the medial gastrocnemius (MG) as a function of lengthening velocity specific to each lengthening period of stance phase in gait cycle. The slope of these plots of each gait cycle was then calculated as spasticity index. calculated as spasticity index, and a positive value indicated that the muscle activation increased with lengthening velocity suggesting the presence of dynamic spasticity during gait (fig 1). The data for dynamic spasticity index were obtained at the same time as the data for the GAITRite were obtained. Passive Stiffness of Plantarflexor Muscles We used the electronic goniometer to measure PROM of dorsiflexion. Passive stiffness was defined as the differences between 20 (normative range for healthy adults) and our subjects dorsiflexion range. 34,35 For this measure, each subject lay in the supine position with his/her knee extended, with straps around the waist and knee for stabilization. PROM of ankle dorsiflexion was performed by the evaluator and recorded by the electronic goniometer. An end feel must be felt to document the maximal range. The intrarater reliability of this procedure was.908. Joint Position Sense of Ankle Joints Joint position sense was used to evaluate proprioceptive sensation because of its high test-retest reliability compared with kinesthesia or other methods. 36 The electronic goniometer was attached to each ankle joint for measuring degrees of position error. Joint position sense was tested 6 times by positioning the subject s affected ankle randomly, and the subject s unaffected ankle was moved to match these positions RESULTS From March 2003 to November 2004, there were 96 patients referred from rehabilitation and neurologic departments. Eightythree patients met the inclusion criteria, and 68 subjects participated in the study. Demographic data of the subjects are described in table 1. There were 52 men and 16 women who had a mean age of 61.69 13.97 years old and an average weight of 63.16 8.97kg. Forty-two subjects had right hemiparesis and 26 were left hemiparesis, with an average poststroke duration of 3.91 5.87 years. The means and standard deviations (SDs) of gait parameters for our subjects are given in table 2. Average walking velocity was 64.58 31.71cm/s (range, 4.40 148.60cm/s). According to the functional walking category, the walking ability of most of our subjects was classified as the least-limited community or community level, indicating that they can go to a local store without assistance or to a crowded shopping center with supervision only. 37 The temporal asymmetry was.23.21, whereas the spatial asymmetry was.20.32. The maximal muscle strength of the plantarflexors and dorsiflexors of the affected and unaffected sides are shown in table 3. The mean of plantarflexors strength of unaffected and affected side were 50.04% 16.63% and 37.16% 19.13% of body weight, respectively. The average values of dorsiflexors strength were 34.57% 9.84% in the unaffected side and 22.32% 13.85% in the affected side. Muscle strength of the affected side was statistically less than that of the unaffected side for both dorsiflexors and plantarflexors (both muscle groups, P.000). We also noted that the muscle strength of the affected side was relatively weaker in dorsiflexors than that in plantarflexors compared with their unaffected side strength (dorsiflexors,.64; plantarflexors,.74; P.002). Our subjects demonstrated positive spasticity indexes 44.35% of the time. Only positive values were used and averaged to indicate the degrees of spasticity. The mean spasticity index was 8.56% 6.72%/l s 1 in our subjects. The magnitude of passive stiffness on the affected side was 5.48 4.72 and 4.52 4.86 of the unaf- Table 1: Subjects Characteristics (N 68) Characteristics Values Range Mean age SD (y) 61.69 13.97 31 82 Mean weight SD (kg) 63.16 8.97 43 87 Mean time poststroke SD (y) 3.91 5.87 0.02 30.78 Sex (M/F) 52/16 NA Affected side (R/L) 42/26 NA Abbreviations: F, female; L, left; M, male; NA, not applicable; R, right; SD, standard deviation.

ANKLE IMPAIRMENTS ON GAIT, Lin 565 Table 2: Gait Parameters Gait Parameters (N 68) Mean SD Range P* Velocity (cm/s) 64.58 31.71 4.40 148.60 Cadence (steps/min) 92.30 23.93 20.40 159.55 Stride length (cm) 82.67 29.48 25.96 157.39 Step width (cm) 13.88 4.95 2.58 29.76 Step length (cm).231 Unaffected 40.15 15.31 9.35 78.33 Affected 41.25 15.64 4.45 76.82 Cycle time (s) 1.48 0.74 0.90 5.83 Single-leg support time (s).000 Unaffected 0.55 0.28 0.31 2.24 Affected 0.41 0.17 0.18 1.64 Double-leg support time (s) 0.58 0.67 0.02 4.78 Temporal asymmetry 0.23 0.21 0.00 0.80 Spatial asymmetry 0.20 0.32 0.00 2.13 *P values for affected and unaffected side comparison. fected side. The mean position error was 7.24 4.62 (see table 3). Stepwise regression analyses revealed that the muscle strength of dorsiflexors of the affected side was the primary determinant for gait velocity, accounting for 30% (R 2 )ofthe variance. In addition, the spasticity index and position error of the ankle joint were also determinants, and these 3 determinants explained up to 50% of the variances for gait velocity (table 4). For temporal asymmetry, the muscle strength of dorsiflexors of the affected side was also the most important determinant, accounting for 38% of the variance. Once the position error was added into the regression model, the explained variances reached 51%. For spatial asymmetry, spasticity index was the most important determinant, accounting for 53% of the variances (see table 4). The relations between ankle impairments and gait performance are described in table 5. Plantarflexor strength of the affected side correlated positively with velocity (r.58, P.01), cadence (r.26, P.05), stride length (r.53, P.01), and step length (affected side, r.50; unaffected side, r.55; P.01) and correlated negatively with step width (r.30, P.05), single-leg support time of unaffected side (r.25, P.05), and temporal (r.33, P.01) and spatial asymmetry (r.28, P.05). The dorsiflexors strength correlated positively and significantly with velocity (r.67, P.01), cadence (r.46, P.01), stride length (r.57, P.01), and step length (affected side, r.53; unaffected side, r.58; P.01) and correlated negatively with step width (r.26, P.05), cycle time (r.39, P.01), single-leg support time of unaffected side (r.32, P.01), double-leg support time (r.34, P.01), and temporal asymmetry (r.60, P.01). The spasticity index correlated significantly with most of the gait parameters. The highest correlations were with spatial asymmetry (r.62, P.01). It also correlated significantly with velocity (r.46, P.01), stride length (r.57, P.01), step length (affected side, r.44; unaffected side, r.64; P.01), cycle time (r.38, P.01), single-leg support time of affected (r.32, P.05) and unaffected side (r.46, P.01), double-leg support time (r.56, P.01), and temporal asymmetry (r.36, P.01). There was no significant correlation between passive stiffness and gait parameters. Position error correlated significantly with velocity (r.27, P.05), cadence (r.21, P.05), stride length (r.28, P.05), step width (r.36, P.05), step length of unaffected side (r.25, P.05), and temporal asymmetry (r.38, P.01) (see table 5). DISCUSSION We investigated the influence of ankle impairments on gait performance. Subjects in our study walked significantly slower and more asymmetrically than healthy adults. 4,38 Dorsiflexors strength of the affected side was the primary determinant for gait velocity and temporal asymmetry. Together with spasticity index and position error, 50% variances of walking velocity can be explained. Furthermore, dorsiflexors strength of the affected side and position error accounted for 51% of variances of temporal asymmetry. However, the dynamic spasticity of plantarflexors determined the spatial asymmetry, accounting for 53% of variances. According to our study, ankle impairments can explain up to half of the variances of gait velocity and asymmetry, indicating that the ankle impairments affected gait performance significantly. Knowledge of specific gait problems possibly affected by different ankle impairments should help the clinicians to program effective training. Compared with the age-matched healthy adults, our subjects with stroke walked with less cadence, shorter stride length, wider step width, longer cycle time, longer single-leg support time of the unaffected side, and longer double-leg support time. Gait deficits in our subjects were consistent with previous findings. 5,6,39 Because our subjects can be classified at the least-limited community level or the community level according to functional walking category, the present results are applicable to subjects with stroke who can walk outdoors. 37 Strength of the Dorsiflexors The dorsiflexor strength of the affected side was the primary determinant for gait velocity and explained 30% of the variances. The weakness of dorsiflexors caused inadequate dorsiflexion control during gait, which was the important factor affecting gait velocity. For foot clearance, insufficient dorsiflexion increased the swing time of the affected leg. 17 The weakness of dorsiflexors may have led to insufficient eccentric contraction at midstance, which resulted in reduced loading ability of the affected leg and increased double-leg support time for longer preparation to the next single-leg support of the affected leg. 22 The increased swing time of the affected side and double-leg support time because of weakness of dorsiflexors resulted in slow gait velocity. Although weakness in other muscles may also contribute to the increases in swing time and Table 3: Muscle Strength of the Plantarflexor and Dorsiflexor and Passive Stiffness of Both Sides, Spasticity Index, and Position Error of Subjects Variables Mean SD Range P* Plantarflexors strength (%).000 Unaffected 50.04 16.63 16.6 91.2 Affected 37.16 19.13 11.8 92.5 Ratio 0.74 NA Dorsiflexors strength (%).000 Unaffected 34.57 9.84 14.49 67.46 Affected 22.32 13.85 2.2 49.8 Ratio 0.64 NA Spasticity index (%/l s 1 ) 8.56 6.72 0.49 35.55 Passive stiffness (deg).041 Unaffected 4.52 4.86 0.00 15.01 Affected 5.48 4.72 0.00 17.82 Position error (deg) 7.24 4.62 1.15 23.71 *P values are for affected and unaffected side comparison. The ratios were the muscle strength of affected side divided by that of unaffected side.

566 ANKLE IMPAIRMENTS ON GAIT, Lin Table 4: Stepwise Regression Analysis for Gait Velocity and Asymmetry Dependent Parameters Gait velocity Temporal asymmetry Independent Parameters R 2 F P Dorsiflexor strength.41.30 19.89.000 Spasticity index.37.45 18.20.001 Position error.25.50 14.73.001 Dorsiflexor strength.62.38 32.93.000 Position error.37.51 27.56.000 Spatial asymmetry Spasticity index.77.53 52.29.000 NOTE. Data for plantarflexor and dorsiflexor strength are from the affected side. reduction in speed, the correlations between dorsiflexor strength and gait parameters in our results supported these explanations. The dorsiflexor strength of the affected side was the primary determinant for temporal asymmetry. Our results showed that the temporal asymmetry is not caused by the decreased singleleg support time of the affected side but by the increased single-leg support time of the unaffected side in our subjects. Therefore, the inability to dorsiflex effectively because of weakness of dorsiflexors caused the increased swing time, which was important in determining temporal symmetry. Only a few studies have mentioned the contribution of dorsiflexors strength on gait. Kim and Eng 14 identified the strength of the plantarflexor, not the dorsiflexor, as the most important factor in determining the gait velocity. Besides the methodologic differences in the way strength testing was performed, inconsistent findings may result from the degrees of weakness in these 2 muscles. Dorsiflexor strength was relatively weaker than plantarflexor strength in our subjects. On the contrary, Kim and Eng reported that the magnitude of weakness of plantarflexors was greater than that of the dorsiflexors in their subjects. Therefore, we suspect that if the dorsiflexor strength is relatively weak, then gait performance is influenced significantly by the dorsiflexors rather than by the plantarflexors. However, there was a moderate correlation between plantarflexor weakness and velocity, which may also suggest that ankle plantarflexor strength is important for gait speed. The fact that dorsiflexor strength played an important role in determining gait velocity in our subjects may also be attributable to their better walking abilities than patients in past studies. According to Kim and Eng s study, the average walking velocity (.45m/s) in their subjects was relatively slow. The increased plantarflexors strength may improve the ability for body propulsion forward during walking. However, in patients with faster walking velocity, as our subjects showed, the propulsion ability may not be as important in determining walk velocity compared with subjects with slower walking velocity. In addition, 20% of our subjects wore ankle-foot orthoses for their regular walking, indicating that they may not have a chance to practice proper dorsiflexor activation during walking. Also, the standard gait training programs have emphasized hip flexor and plantarflexor strengthening, which may overlook the dorsiflexor strength training. 13 Our results indicate the dorsiflexor strengthening program should also be emphasized for stroke subjects who already can walk outdoors to achieve a faster walk velocity or a more symmetrical pattern. Dynamic Spasticity of the Plantarflexors The moderate correlation between spasticity index and gait parameters found in our study indicates its importance on gait. The plantarflexor spasticity was elicited by stretching gastrocnemius during weight transfer on the affected leg and caused difficulty in moving the center of gravity forward for the next step. The limited ability of weight transferred to the affected leg also resulted in a short step length of the unaffected leg. The high correlations between dynamic spasticity and step length of the unaffected side (r.64) and stride length (r.57) support this hypothesis. The shorter step length of the unaffected side resulted from the dynamic spasticity of the plantarflexors during stance phase caused by spatial asymmetry. Therefore, it was also the most important factor in determining spatial asymmetry. Although the dynamic spasticity was measured only in stance phase, it is not surprising that induced dynamic spasticity in stance phase may also affect the swing phase. The increased plantarflexor spasticity during stance phase may have caused an excessive plantarflexion at preswing phase, which Table 5: Pearson Correlation Coefficients (r) Between Gait Parameters and Muscle Strength, Spasticity Index, Passive Stiffness, and Position Error Gait Parameters Plantarflexor Strength Dorsiflexor Strength Spasticity Index Passive Stiffness Position Error Velocity (cm/s).58.67.46.16.27* Cadence (steps/min).26*.46.02.05.21* Stride length (cm).53.57.57.17.28* Step width (cm).30*.26*.23.02.36* Step length (cm) Affected side.50.53.44.15.25 Unaffected side.55.58.64.20.25* Cycle time (s).18.39.38.24.15 Single-leg support time (s) Affected side.06.00.32*.08.08 Unaffected side.25*.32.46.22.05 Double-leg support time (s).15.34.56.22.16 Temporal asymmetry.33.60.36.06.38 Spatial asymmetry.28*.20.62.17.10 NOTE. Data for plantarflexor and dorsiflexor strength are from the affected side. *P.05. P.01.

ANKLE IMPAIRMENTS ON GAIT, Lin 567 resulted in difficulty in foot clearance and therefore increased the swing time. The increased swing time of the affected side and the decreased step length of the unaffected side were the major causes for the decreased velocity. It is thus suggested that the role of dynamic spasticity of plantarflexors on gait velocity and symmetry should not be underestimated. Joint Position Sense of Ankle Joints The position error of our subjects was significantly impaired compared with normative data. 40 The uncertain foot position during walking may cause small or altered step length and therefore affected gait velocity. The relations between position error and step length of the unaffected side in our study may support this hypothesis. It is also the second determinant for temporal asymmetry. Impaired joint position sense, and hence the hesitatance on foot landing, resulted in a longer swing time of the affected side. This also affected temporal symmetry. Passive Stiffness of the Plantarflexors Passive stiffness was not the determinant for velocity and symmetry in our study. This is because passive stiffness of our subjects may not be impaired enough to make gait deficits. According to previous reports, an adequate ankle range of motion for normal gait is 10 to 15 of dorsiflexion for tibia to move over the foot. 41,42 The average PROMs of dorsiflexion in our subjects were 15.39 on the affected side and 17.56 on the unaffected side, which were within the functional range for gait. However, the clinical measurement used in our study for passive stiffness may not be sensitive enough, although it has been reported to correlate moderately with other mechanical measurement tools. 43 As suggested previously, 27 a multifactorial analysis in a large sample of patients would be beneficial for a better understanding of passive stiffness on gait. In our study, the ankle impairments were analyzed to identify their roles on gait performance in stroke patients. Further studies are suggested to evaluate the improvement of gait performance by training of ankle dorsiflexors, normalizing dynamic plantarflexor spasticity, or improving proprioceptive sense to confirm our findings. Our study, however, was limited to a single joint and the influence of other joints on the gait was not considered. CONCLUSIONS Our results show that the ankle impairments affected gait performance significantly in stroke patients who have relatively good walking ability. Different ankle impairments affected different aspects of gait performance. Dorsiflexor strength of the affected side was the most important factor in determining gait velocity and temporal symmetry. Dynamic spasticity of the plantarflexors was the only determining factor for spatial symmetry. Awareness of specific gait problems possibly affected by different ankle impairments should help clinicians to program effective training. References 1. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Study. Arch Phys Med Rehabil 1995;76:27-32. 2. Wade DT, Wood VA, Heller A, Maggs J, Langton Hewer R. Walking after stroke. Measurement and recovery over the first 3 months. Scand J Rehabil Med 1987;19:25-30. 3. Dewar ME, Judge G. Temporal asymmetry as a gait quality indicator. Med Biol Eng Comput 1980;18:689-93. 4. Titianova EB, Tarkka IM. Asymmetry in walking performance and postural sway in patients with chronic unilateral cerebral infarction. J Rehabil Res Dev 1995;32:236-44. 5. Goldie PA, Matyas TA, Evans OM. Gait after stroke: initial deficit and changes in temporal patterns for each gait phase. Arch Phys Med Rehabil 2001;82:1057-65. 6. Ozgirgin N, Bolukbasi N, Beyazova M, Orkun S. Kinematic gait analysis in hemiplegic patients. Scand J Rehabil Med 1993;25: 51-5. 7. Berger W, Horstmann G, Dietz V. Tension development and muscle activation in the leg during gait in spastic hemiparesis: independence of muscle hypertonia and exaggerated stretch reflexes. J Neurol Neurosurg Psychiatry 1984;47:1029-33. 8. Bohannon RW, Larkin PA, Smith MB, Horton MG. Relationship between static muscle strength deficits and spasticity in stroke patients with hemiparesis. Phys Ther 1987;67:1068-71. 9. Bohannon RW, Horton MG, Wikholm JB. Importance of four variables of walking to patients with stroke. Int J Rehabil Res 1991;14:246-50. 10. Nadeau S, Arsenault AB, Gravel D, Bourbonnais D. Analysis of the clinical factors determining natural and maximal gait speeds in adults with a stroke. Am J Phys Med Rehabil 1999;78:123-30. 11. Knutsson E, Richards C. Different types of disturbed motor control in gait of hemiparetic patients. Brain 1979;102:405-30. 12. Keenan MA, Perry J, Jordan C. Factors affecting balance and ambulation following stroke. Clin Orthop Relat Res 1984;Jan- Feb(182):165-71. 13. Nadeau S, Gravel D, Arsenault AB, Bourbonnais D. Plantarflexor weakness as a limiting factor of gait speed in stroke subjects and the compensating role of hip flexors. Clin Biomech (Bristol, Avon) 1999;14:125-35. 14. Kim CM, Eng JJ. The relationship of lower-extremity muscle torque to locomotor performance in people with stroke. Phys Ther 2003;83:49-57. 15. Olney SJ, Griffin MP, McBride ID. Temporal, kinematic, and kinetic variables related to gait speed in subjects with hemiplegia: a regression approach. Phys Ther 1994;74:872-85. 16. Olney SJ, Griffin MP, Monga TN, McBride ID. Work and power in gait of stroke patients. Arch Phys Med Rehabil 1991;72:309-14. 17. Lamontagne A, Malouin F, Richards CL, Dumas F. Mechanisms of disturbed motor control in ankle weakness during gait after stroke. Gait Posture 2002;15:244-55. 18. Dietz V, Quintern J, Berger W. Electrophysiological studies of gait in spasticity and rigidity. Evidence that altered mechanical properties of muscle contribute to hypertonia. Brain 1981;104: 431-49. 19. Richards CL, Malouin F, Dumas F, Wood-Dauphinee S. The relationship of gait speed to clinical measures of function and muscle activations during recovery post-stroke. In: Proceeding of the Second North American Congress on Biomechanics; 1992 Aug 24-28; Chicago (IL). p 299-302. 20. Yang JF, Fung J, Edamura M, Blunt R, Stein RB, Barbeau H. H-reflex modulation during walking in spastic paretic subjects. Can J Neurol Sci 1991;18:443-52. 21. Crenna P. Spasticity and spastic gait in children with cerebral palsy. Neurosci Biobehav Rev 1998;22:571-8. 22. Whittle MW. Gait analysis: an introduction. Oxford: Butterworth- Heinemann; 2003. 23. Hsu AL, Tang PF, Jan MH. Analysis of impairments influencing gait velocity and asymmetry of hemiplegic patients after mild to moderate stroke. Arch Phys Med Rehabil 2003;84:1185-93. 24. Lamontagne A, Malouin F, Richards CL. Locomotor-specific measure of spasticity of plantarflexor muscles after stroke. Arch Phys Med Rehabil 2001;82:1696-704. 25. Lamontagne A, Malouin F, Richards CL. Contribution of passive stiffness to ankle plantarflexor moment during gait after stroke. Arch Phys Med Rehabil 2000;81:351-8. 26. Thilmann AF, Fellows SJ, Ross HF. Biomechanical changes at the ankle joint after stroke. J Neurol Neurosurg Psychiatry 1991;54:134-9.

568 ANKLE IMPAIRMENTS ON GAIT, Lin 27. Malouin F, Bonneau C, Pichard L, Corriveau D. Non-reflex mediated changes in plantarflexor muscles early after stroke. Scand J Rehabil Med 1997;29:147-53. 28. Dettmann MA, Linder MT, Sepic SB. Relationships among walking performance, postural stability, and functional assessments of the hemiplegic patient. Am J Phys Med 1987;66:77-90. 29. McDonough AL, Batavia M, Chen FC, Kwon S, Ziai J. The validity and reliability of the GAITRite system s measurements: a preliminary evaluation. Arch Phys Med Rehabil 2001;82:419-25. 30. Bilney B, Morris M, Webster K. Concurrent related validity of the GAITRite walkway system for quantification of the spatial and temporal parameters of gait. Gait Posture 2003;17:68-74. 31. Perotto AO. Anatomical guide for the electromyographer: the limbs and trunk. Springfield: CC Thomas; 1994. 32. Norkin CC, White DJ. Measurement of joint motion: a guide to goniometry. Philadelphia: FA Davis; 2003. 33. Winter DA, Scott SH. Technique for interpretation of electromyography for concentric and eccentric contractions in gait. J Electromyogr Kinesiol 1991;1:263-9. 34. Gajdosik RL, Vander Linden DW, Williams AK. Influence of age on length and passive elastic stiffness characteristics of the calf muscle-tendon unit of women. Phys Ther 1999;79:827-38. 35. Vandervoort AA, Chesworth BM, Cunningham DA, Paterson DH, Rechnitzer PA, Koval JJ. Age and sex effects on mobility of the human ankle. J Gerontol 1992;47:M17-21. 36. Deshande N, Connelly DM, Culham EG, Costigan PA. Reliability and validity of ankle proprioceptive measures. Arch Phys Med Rehabil 2003;84:883-9. 37. Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking handicap in the stroke population. Stroke 1995;26:982-9. 38. Leiper CI, Craik RL. Relationships between physical activity and temporal-distance characteristics of walking in elderly women. Phys Ther 1991;71:791-803. 39. Olney SJ, Richards C. Hemiparetic gait following stroke. Part I: Characteristics. Gait Posture 1996;4:136-48. 40. Carey LM, Oke LE, Matyas TA. Impaired limb position sense after stroke: a quantitative test for clinical use. Arch Phys Med Rehabil 1996;77:1271-8. 41. Ostrosky KM, VanSwearingen JM, Burdett RG, Gee Z. A comparison of gait characteristics in young and old subjects. Phys Ther 1994;74:637-44; discussion 644-6. 42. Murray MP. Gait as a total pattern of movement. Am J Phys Med Rehabil 1967;46:290-333. 43. Lamontagne A, Malouin F, Richards CL. Viscoelastic behavior of plantar flexor muscle-tendon unit at rest. J Orthop Sports Phys Ther 1997;26:244-52. Suppliers a. CIR Systems Inc, 60 Garlor Dr, Havertown, PA 19083. b. Jtech Medical Industries Inc, 357 W 910 South, Herber City, UT 84032. c. BIOPAC Systems Inc, 42 Aero Camino, Goleta, CA 93117. d. Biometrics Ltd, PO Box 340, Ladysmith, VA 22501. e. Version 10.0; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.