SENSORIMOTOR DYSFUNCTION, such as impaired sensation,

197 Motor Function and Joint Position Sense in Relation to Gait Performance in Chronic Stroke Patients Sang-I Lin, PhD, PT ABSTRACT. Lin S-I. Motor function and joint position sense in relation to gait performance in chronic stroke patients. Arch Phys Med Rehabil 2005;86:197-203. Objectives: To determine the association between joint position sense (JPS) and motor function of the lower extremity and gait performance of patients with chronic stroke. Design: Single-group cross-sectional design. Setting: University gait laboratory. Participants: Twenty-one patients with stroke onset of more than 6 months. Interventions: Not applicable. Main Outcome Measures: The isometric strength, Fugl- Meyer Assessment (FMA) motor status, and JPS of the lower extremity were tested. Spatiotemporal gait characteristics were measured using a 6-camera motion analysis system, with patients walking at their comfortable speeds while using their usual devices. Results: Lower-extremity muscle strength and FMA motor score correlated significantly with the spatiotemporal characteristics of gait and contributed significantly to the variance in gait velocity, stride length, and double-stance time. The JPS was not related to gait performance, except that the ankle JPS contributed significantly to the variance in gait velocity and stride length. Conclusions: For patients with chronic stroke, motor function was significantly related to gait performance. Although the JPS was not, that of the ankle joint made a significant contribution. When enhancing gait performance is desired, improving the motor function is recommended, and the role of JPS should also be taken into consideration. Key Words: Gait; Motor skills; Rehabilitation; Sensory motor performance; Stroke. 2005 by American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation SENSORIMOTOR DYSFUNCTION, such as impaired sensation, muscle weakness, altered muscle tone, and/or lack of isolated movement control, is commonly seen in patients with stroke. 1,2 For these patients, the ability to perform functional activities is often compromised. One of the functional disabilities that can greatly affect the life of many patients is poor ambulatory ability. Ambulatory ability determines to a large degree the level of independence of daily living and has From the Department of Physical Therapy, National Cheng Kung University, Tainan, Taiwan. Supported by the National Health Research Institute, Taiwan (grant no. NHRI-EX 91-8902). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the author(s) or on any organization with which the author(s) is/are associated. Reprint requests to Sang-I Lin, PhD, PT, Dept of Physical Therapy, National Cheng Kung University, Tainan, 701, Taiwan, e-mail: lin31@mail.ncku.edu.tw. 0003-9993/05/8602-8861$30.00/0 doi:10.1016/j.apmr.2004.05.009 been reported by patients to be the most important goal for rehabilitation. 3,4 However, despite intensive rehabilitation, gait disturbances, such as slower walking velocity and asymmetric gait pattern, can often be seen in patients. 5-11 It is imperative to identify factors contributing to gait performance so that better rehabilitation protocols can be developed. One of the most extensively studied contributing factors in gait performance is muscle strength. A large quantity of research has been conducted to reveal the relationships between muscle strength and gait performance in patients with stroke. Various patient populations, measurements, and gait variables were examined in different studies. To date, information derived from the existing literature is far from conclusive. For example, although positive correlations between leg muscle strength and gait velocity have been reported, 10,12-17 there were also studies showing weak or nonsignificant correlations. 18-20 This inconclusiveness in research findings is not surprising, considering that the ability to recruit a single group of skeletal muscles under well-supported posture in a stable, predictable environment, such as in a strength test, does not necessarily guarantee a similar ability of muscle recruitment during walking. Control of functional locomotion requires continuous sensory afferent input. 21 Impaired sensory function may impede the ability of muscle recruitment during walking. Empirically, patients who suffer from impaired sensory function tend to be slower to regain functional ability, and, very often, they fail to achieve optimal functional recovery. 22-24 This anecdote is contradicted by many studies 12,20,22,25,26 that found nonsignificant or low correlations between lowered sensory function and poorer gait performance. However, a recent study, 17 reported that the Fugl-Meyer Assessment (FMA) sensory scores of patients with mild to moderate stroke were correlated significantly with gait velocity and made small but significant contributions to gait velocity. The controversies surrounding the role of sensory function in gait performance have been partly attributed to the method used in sensory function assessment. Dettmann et al, 25 after failing to find a significant correlation between sensory function and gait performance, questioned their method of sensory function test. The FMA sensory score, which was used by Dettmann 25 and in other studies, 12,14,17,20 rates light touch and the position sense of various areas of the leg based on a 3-level scale and uses the sum of these scores to represent sensory function. 27 This method thus assesses 2 different sensory modalities. However, research evidence has shown that tactile sensation and joint position sense (JPS) may have unequal roles in locomotion control. Although the vibration (activates Ia afferent group) threshold was related to the peak pressure of the foot during walking, tactile threshold was not. 28 Thus, although the FMA sensory test is a reliable clinical test, 29,30 the appropriateness of using it to relate to gait performance is not clear. When proprioception was specifically investigated, experimentally induced changes in its function in healthy adults has been shown repeatedly to lead to altered gait patterns. 28,31 Furthermore, an animal study 32 also shows that impaired knee position sense was related to marked reduction in muscle recruitment in walking. These findings suggest that JPS, whose

198 SENSORIMOTOR FUNCTION AND GAIT IN CHRONIC STROKE, Lin function has not been quantitatively and separately assessed in the previous stroke-related studies, may have significant contribution to gait performance and thus is in need of further investigation. Another issue that deserves attention in this area of research is the composition of the participants, in terms of onset durations. There are at least 2 mechanisms contributing to functional recovery after stroke, neurologic recovery, and functional adaptation. 33-35 It is generally believed that neurologic recovery is most likely to occur in the acute stage, leading to significant functional recovery. 36,37 In the chronic stage, when most patients are discharged to home, functional recovery can still be gained, probably via adaptations to their home environment or developing compensatory strategies to overcome their physical limitations. 38,39 These different functional recovery mechanisms make it reasonable to assume that the relationships between sensorimotor function and gait performance in the acute versus chronic populations may not be identical. Previous research that investigated contributing factors to gait performance has primarily been concerned with patients with acute and chronic onset as a homogenous group. 12,14,16-20,22,25 Doing so may risk the chance of failing to detect the specific relationships in the patient group of interest. To address the limitations in the previous studies regarding the relationships between sensorimotor function and gait performance in patients with stroke, this study specifically focused on patients with chronic onset durations and used quantitative measures of muscle strength and JPS. METHODS Participants Subjects were recruited from the community on a volunteer basis through local newspaper and poster advertising. All subjects were screened by a physical therapist to determine whether they met the following inclusion criteria: (1) first time unilateral stroke for at least 1 year; (2) having residual unilateral weakness and gait deviations; (3) ability to ambulate independently with or without assistive devices for at least 10m; (4) ability to follow 3-step commands; (5) ability to complete the JPS test (described below); (6) no other known neurologic or orthopedic problems that could affect sensorimotor function or gait; and (7) medically stable to walk and perform all the tests. Twenty-six patients were contacted and screened, and all but 5 met the inclusion criteria and participated in the study. All patients suffered from first-time stroke, were between the ages of 51 and 79 years, and lived in the community. The onset of stroke was from 18 to 247 months. All patients were ambulatory, and half used assistive devices (table 1). This study was approved by the institutional human subject review board of the institution at which the study was conducted. Informed consent was obtained from all patients. Gait Evaluation Patients were instructed to walk at their comfortable speeds using their usual assistive device (if required) on a 10-m walkway for 3 repetitions. Typically, a rest period of about 2 minutes was given after each walking trial. A Vicon motion analysis system a with 6 cameras was used to record gait kinematics at 60Hz. Reflective markers were placed on the bilateral second metatarsal head, lateral malleolus, midshank, knee joint center, midthigh, anterior superior iliac spine, and shoulder tip and at the midpoint between bilateral posterior superior iliac spine. For each walking trial, data collection did not begin until Table 1: Characteristics of Subjects With Chronic Stroke Characteristics Age (y) Weight (kg) Height (m) Onset (mo) Characteristics Mean SD 65.2 9.1 65.0 10.0 1.61 0.08 63.2 55.5 Subjects (n) Gender Male 15 Female 6 CVA type Infarct 12 Hemorrhage 9 Hemiplegic side Left 13 Right 8 Assistive device None 11 Regular cane 1 Quadricane 9 Abbreviations: CVA, cerebrovascular accident; SD, standard deviation. at least 3 steps were taken by the patients to ensure that a steady state of gait velocity had been achieved. 40 Motor Function Evaluation The isometric strength of the bilateral hip flexors, knee extensors, and ankle dorsiflexors was tested using a hand-held dynamometer. b Each strength test was performed 3 times. The testing position for each muscle group was standardized. 41 The motor function of the lower extremity was also assessed by the FMA lower-extremity motor assessment, 27 which primarily rates reflex activity, muscle synergy pattern, and active motion based on a 3-level scale. The higher the score (maximum, 34), the better the motor status. JPS Evaluation The JPS of the knee and ankle was evaluated using a computerized 2-inclinometer system, c with the inclinometer attached along the axis of the fibular bone for the knee joint test and attached to the fifth metatarsal bone for the ankle joint test. In this test, the proximal segment of the joint being tested was kept motionless, while the patient moved the distal segment of the unaffected joint to match the corresponding joint angle of the affected side that was moved to a specific angle by the examiner. The rationale for having the unaffected leg match the affected joint angle was to avoid confounding proprioceptive impairments with erroneous matching movement because of poor motor control in the affected leg. For the ankle joint, the patient sat with the leg hanging vertically to the ground. The examiner moved the affected foot from a neutral position to 10 of dorsiflexion or plantarflexion and then signaled the patient to actively move the unaffected foot to match the joint angle of the affected ankle angle. For the knee joint, the patient sat with the thighs fully supported by the seat. The examiner moved the affected leg from 90 of knee flexion to 100 or 80 of flexion and then asked the patient to move actively the unaffected leg to match the affected knee angle. The absolute values of the differences in the angle between the 2 matching joints (joint-diff) was recorded.

SENSORIMOTOR FUNCTION AND GAIT IN CHRONIC STROKE, Lin 199 Fig 1. Gait characteristics and muscle strength. Legends: Velocity is in percentage of height/s; stride length and step length are in percentage of height; SW, SS, and DBS are in percentage of gait cycle; muscle strength is in percentage of body weight. Abbreviations: DBS, double-leg stance; dorsi, dorsiflexion; ext, extension; flex, flexion; SS, single-leg stance; SW, swing. *P<.000. Greater joint-diff indicated poorer JPS. Each joint was tested 3 times. Data Reduction Vicon Clinical Manager software a was used to derive joint angles from the reflective marker data using Euler angles. Specific gait events, including initial affected foot contact, unaffected foot-off, unaffected foot contact, affected foot-off, and second affected foot contact, were identified by visual inspection of the vertical displacement of the ankle (lateral malleolus) or toe (second metatarsal head) markers. Based on the temporal and spatial information of these gait events, the Vicon Clinical Manager software calculated velocity (average advancing velocity in the sagittal plane), stride length (distance between the initial and second foot contact), step length (distance between initial foot contact and foot-off), single-leg stance time (time elapsed between foot-off and foot contact of the opposite leg), swing time (time elapsed between foot-off and second foot contact of the same leg), and double-leg stance time (total gait cycle time minus swing and single-stance time). Data Analysis The means of 3 repetitions were used for data analysis. Unless otherwise specified, the data of the affected side were used for analysis. The muscle strength was normalized to the body weight. All the spatial gait variables, including velocity, stride length, and step length, were normalized to the body height, whereas the temporal gait variables were presented in percentage gait cycle. For the JPS, the joint angle difference (joint-diff) was used as a continuous variable for correlation and regression analyses. In addition, the JPS was classified into intact or impaired, based on 2 criteria: range and variation of joint-diff. Unpublished data from this laboratory showed that, for healthy adults, the joint-diff ranged from 1 to 13 for the knee joint and 1 to 14 for the ankle joint. The greatest within-subject variation was 6 for both knee and ankle. Thus, when 2 of the 3 repeated joint position tests of the same joint showed joint-diff greater than 13 for the knee or 14 for the ankle, the patient would be classified as having impaired JPS. In addition, when the withinsubject variation was greater than 6, the JPS would also be classified as impaired. Computer software d was used for statistical analysis. Paired t tests were used to compare the gait variables (speed, stride length, step length, swing time, single-stance time, doublestance time) and leg muscle strength of the 2 legs. Multivariate analysis of variance (MANOVA) was used to determine the differences in the gait variables, in association with the use of assistive devices and JPS classification. Pearson correlation tests were performed to determine the relationships between gait variables and sensorimotor function (FMA score, muscle strength, joint-diff). To give a comprehensive understanding of the contribution of the sensorimotor function to gait, we performed linear regression analyses between gait and all the sensorimotor function variables. Specifically, each of the 6 gait spatiotemporal characteristics (velocity, stride length, step length, swing time, single-stance time, double-stance time) was entered separately as a dependent variable, whereas independent variables, including age, use of an assistive device (as dummy variable), FMA motor score, strength of bilateral hip flexor, knee extensor, ankle dorsiflexion, and knee and ankle joint-diff, were entered at the same time. A type I error rate of.05 was used to indicate statistical significance. RESULTS Gait Characteristics Asymmetric gait was observed, with the affected leg showing significantly greater step length, shorter single-stance time, and longer swing time than the unaffected leg (fig 1). With regard to the use of assistive devices, MANOVA showed that patients walking without them differed significantly from those walking with them (table 2). Follow-up univariate analysis showed significantly slower walking velocity, smaller step and stride length, longer double-leg stance time, and shorter swing and single-leg stance time in patients walking with devices (fig 2). Sensorimotor Function Motor function. Overall, the affected leg showed significant muscle weakness compared with the unaffected side (fig 1). A wide range of lower-extremity motor function existed among the patients, shown by the FMA lower-extremity motor Table 2: Statistics for Multivariate Analysis on the Use of Assistive Devices Item Univariate Wilks Follow-Up F P F P Gait (affected leg) 8.737.000 Velocity (% height/s) 46.892.000 Stride length (% body 21.626.000 height) Step length (% body 8.108.000 height) SW time (%GC) 21.055.000 SS time (%GC) 50.129.000 DBS time (%GC) 41.124.003 Motor function (affected leg) 3.272.034 Hip flexion (% body 7.561.013 weight) Knee extension (% body 5.993.024 weight) Ankle dorsiflexion (% 7.355.014 body weight) FMA 16.517.001 Abbreviations: DBS, double-leg stance; GC, gait cycle; SS, single-leg stance; SW, swing.

200 SENSORIMOTOR FUNCTION AND GAIT IN CHRONIC STROKE, Lin Fig 2. Means and SDs of gait and motor function in patients ambulating with and without assistive devices. The between-group differences were all statistically significant. scores (mean, 21 8), ranging from very low (score 5) to maximum (score 34). Patients who walked without assistive devices had significantly greater muscle strength and higher FMA scores (fig 2) than those using devices (table 2). Joint position sense. Six patients were classified as having impaired knee JPS, and 7 as having impaired ankle JPS. The means and standard deviations (SDs) of gait characteristics and motor function for patients with impaired and intact JPS are shown in figures 3 and 4. Patients with impaired knee JPS were significantly older than those with intact knee JPS (t 2.907, P.01). Other than the age differences, there was no significant difference in muscle strength or FMA motor score (table 3). There was also no significant correlation between joint-diff and motor function (table 4). Relation Between Sensorimotor Function and Gait Performance Motor function. The strength of the affected hip flexion, knee extension, and ankle dorsiflexion correlated significantly with all the gait variables, except for step length (table 4). The FMA lower-extremity motor scores also correlated significantly with all the gait variables. Joint position sense. The gait characteristics of the affected leg did not differ significantly between patients with intact and impaired knee or ankle JPS (table 3). The number of patients using assistive devices was equally distributed for patients with and without ankle sensory impairments (table 3). However, patients with impaired knee JPS were more likely to use assistive devices for walking than those with intact knee joint sensation ( 2 test 4.333, P.05). Correlation analysis showed that the knee and ankle joint-diff was not significantly related to any of the gait variables (table 4). Regression Analysis For each gait variable, a regression analysis was conducted entering the use of an assistive device as dummy variable; age, FMA motor score, strength of bilateral hip flexor, knee extensor, ankle dorsiflexion, and knee and ankle proprioception (joint-diff) were entered as independent variables. The statistical findings are shown in table 5. Gait velocity. The independent variables together explained 96% (R 2 ) of the variance in gait velocity, with age, use of assistive device, FMA motor score, strength of knee extensors and ankle dorsiflexors, and ankle JPS showing statistically significant contributions. Stride length. The independent variables together explained 86% of the variance in stride length, with FMA motor score, affected knee extensor strength, and ankle JPS showing significant contribution. Step length. The independent variables together explained 65% of the variance in step length, but the P value was statistically nonsignificant. Swing time. The independent variables together explained 84% of the variance in swing time. Although regression analysis yielded statistically significant findings, none of the independent variable showed a significant contribution. Single-stance time. The independent variables together explained 90% of the variance in single-stance time, with only the use of assistive devices showing a statistically significant contribution. Double-support time. The independent variables together explained 89% of the variance in double-support time, with affected knee extensor strength and use of assistive devices showing statistically significant contribution. DISCUSSION Impairments in the sensory and motor function are often seen in patients with stroke and can contribute to poor gait performance. Our study examined a group of patients with chronic stroke and showed that motor function, demonstrated by lower-extremity muscle strength and FMA motor score, correlated significantly with the spatiotemporal characteristics of gait and contributed significantly to the variance in gait velocity, stride length, and double-stance time. Although there was no direct relationship between JPS and gait performance, ankle JPS contributed significantly to the variance in gait velocity and stride length. Changes in gait pattern are common in patients with stroke. Compared with the accepted normative gait temporal characteristics of 40% gait cycle for swing, 40% gait cycle for single stance, and 20% gait cycle for double stance, 42,43 our study observed decreased swing (23% gait cycle) and single-stance Fig 3. Means and SDs of gait and motor function in patients with normal and impaired knee position sense. MANOVA showed a nonsignificant group effect. Fig 4. Means and SDs of gait and motor function in patients with normal and impaired ankle position sense. MANOVA showed a nonsignificant group effect.

SENSORIMOTOR FUNCTION AND GAIT IN CHRONIC STROKE, Lin 201 Table 3: Age, Use of Assistive Device, Motor Function, Gait Characteristics, and Statistics for Patients With Normal and Impaired Joint Sensation Knee Ankle Item Normal (n 15) Impaired (n 6) Normal (n 14) Impaired (n 7) Mean age SD (y) 62.5 9.1 71. 5.3* 66.1 8.8 63.4 10.2 Assistive device (no. of patients) No/yes 10/5* No/yes 1/5* No/yes 7/7 No/yes 4/3 Wilks Wilks Motor function F.741, P.580 F 1.291, P.320 Gait F 1.027, P.453 F 1.289, P.333 *P.05 between normative and impaired knee joint sensation. Independent t test. 2 test. MANOVA with strength of hip flexion, knee extension and ankle dorsiflexion, and FMA motor score as dependent variables. MANOVA with gait velocity, stride length, step length, swing time, single-stance time, and double-stance time as dependent variables. (30% gait cycle) time and increased double-support time (47% gait cycle). Spatial gait characteristics of the patients were also noted to have changed, with the step length of the 2 legs differing significantly. These altered gait patterns in patients with chronic stroke coincide with those reported in the other studies in which patients with acute and chronic stroke were pooled together. 44-47 As to which factors contributed to gait performance, the motor function of the lower extremity appears to be strongly implicated. Muscle strength of the affected hip flexors, knee extensors and ankle dorsiflexors, and FMA motor score correlated significantly with the gait spatiotemporal characteristics. These findings not only add to the existing literature to support the notion that motor function is related to gait performance, but also point out that these relationships exist in patients in the chronic stage. It should be noted that the number of muscle groups under investigation in our study was far fewer than those that contribute to locomotion. Therefore, these findings could not provide direct evidence that would indicate how the strength of a particular muscle group was related to a specific characteristic of the gait performance. What we can be sure of is that, considering the significant correlations between motor function and gait, strength maintenance and improvement should continue to be emphasized in patients in the chronic stage of stroke. In addition to impaired motor function, our study, based on categorical classification, found that JPS impairments existed in about one third of the patients, and gait performance did not differ significantly between patients with normal and impaired JPS. Furthermore, the extent of the JPS impairment (joint-diff) Table 4: Pearson Correlation Coefficients for Age, Gait, Motor Function, and JPS Item Hip Flexion Knee Extension Ankle Dorsiflexion Knee Sensation Ankle Sensation Velocity.633*.436.645*.075.021 Stride length.585*.488.653*.023.016 Step length.255.347.429.025.026 SW time.594*.388.382.111.119 SS time.644*.421.631*.147.071 DBS time.659*.431.531.018.014 NOTE. Joint difference is the absolute value of the difference between the matching affected and nonaffected joints during passive motion test (joint-diff). *P.01. P.05. of the knee and ankle was not significantly related to gait performance. These findings appear to coincide with those reported in the literature 12,20,22,25,26 that argue against the anecdotal notion that poorer sensory function is associated with poorer gait performance. The lack of significant direct relationship between JPS and gait performance has been attributed, at least partly, to the redundancy in the afferent control of locomotion. Walking extensively involves continuous controlled weight bearing and shifting and would activate the pressure sense that could, in turn, provide crucial sensory afferent input for locomotion control. 23,48 Thus, somatosensory pathways other than proprioception can be activated, such that the impact of proprioceptive loss on gait performance could be partly concealed. Despite a lack of direct relationship, the ankle JPS was a significant contributor to gait velocity and stride length. Clinically, patients who complain of not knowing where the foot is tend to walk at slower velocity and take smaller steps. Changes in gait patterns, including slower velocity or smaller step length, have been observed in healthy subjects with experimentally induced proprioceptive change around the ankle joint. 31 It has been suggested that changes in locomotor strategies, shown by slower walking velocity to compensate for Table 5: Regression Analyses for Gait Variables Gait Variable R P Significant Factor P Velocity.978.000 Age.030 Device.009 FMA.020 Affected knee ext.001 Affected ankle dorsi.023 Ankle joint-diff.008 Stride length.927.011 FMA.019 Affected knee ext.040 Ankle joint-diff.040 Step length.804.276 Swing time.915.020 None Single-stance time.947.003 Device.032 Double-stance time.942.005 Device.039 Affected knee ext.048 NOTE. For each gait variable, a linear regression analysis was performed entering age, FMA motor score, strength of bilateral hip flexion, knee extension (ext) and ankle dorsiflexion (dorsi), and JPS (joint-diff) of knee and ankle as independent variables. Use of assistive device was also entered as a dummy variable.

202 SENSORIMOTOR FUNCTION AND GAIT IN CHRONIC STROKE, Lin sensory loss, could be adopted by patients to improve walking stability. 49,50 Following this line of thinking, long-term proprioceptive impairments might lead to the development of compensatory gait changes in patients with chronic stroke. These arguments, if true, would imply that some of the gait changes in patients with chronic stroke could be required functionally. Furthermore, clinical efforts to change gait patterns should be carefully thought through, taking into consideration the possibility that the existing gait pattern could partly be the individual s solution to compensate for physical impairments. Further studies are needed to clarify this issue. For patients whose impairments prohibit them from walking safely or efficiently, ambulatory assistive devices can be provided. However, the ways in which these devices affect gait performance are still being debated. For patients who can walk without assistive advices, the addition of a cane resulted in greater stride and step length 14 or in no significant changes. 51,52 In our study, patients who used assistive devices had more pronounced atypical gait patterns, indicated by slower speed, smaller stride and step length, and markedly shorter swing and longer double-support time, compared with those without devices. In addition, use of assistive devices was a significant contributor to gait velocity and to single- and double-stance time. These findings, however, by no means suggest that use of assistive devices resulted in more pronounced atypical gait patterns. Because these 2 groups of patients differed significantly in their motor function, the leg muscle strength and FMA motor scores were significantly poorer in patients walking with devices. The disparity in motor function could also contribute to the differences in gait performance between the 2 groups. CONCLUSIONS For patients with chronic stroke, the strength of the affected hip flexor, knee extensor, and ankle dorsiflexors was directly related to their gait spatiotemporal characteristics. Although the JPS of the knee and ankle did not have a direct relationship with gait performance, the ankle JPS was a significant contributor to gait velocity and stride length. Thus, clinically maintenance or improvement of leg muscle strength should be emphasized in the chronic stage. The ability to perceive the position of the ankle joint should be taken into account in clinical decision making about the necessity of changing gait patterns in patients with chronic stroke. Acknowledgment: The author thanks Mei-Chun Hwang and Yie- Ning Wu for help in data collection. References 1. Davidoff GN, Keren O, Ring H, Solzi P. Acute stroke patients: long-term effects of rehabilitation and maintenance of gains. Arch Phys Med Rehabil 1991;72:869-73. 2. Thorngren M, Westling B. Rehabilitation and achieved health quality after stroke. A population-based study of 258 hospitalized cases followed for one year. Acta Neurol Scand 1990;82:374-80. 3. Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil 1995;76:27-32. 4. Bohannon RW, Andrews AW, Smith MB. Rehabilitation goals of patients with hemiplegia. Int J Rehabil Res 1988;11:181-3. 5. Goldie PA, Matyas TA, Evans OM. Deficit and change in gait velocity during rehabilitation after stroke. Arch Phys Med Rehabil 1996;77:1074-82. 6. 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. 7. Wall JC, Turnbull GI. Gait asymmetries in residual hemiplegia. Arch Phys Med Rehabil 1986;67:550-3. 8. Hesse SA, Jahnke MT, Bertelt CM, Schreiner C, Lucke D, Mauritz KH. Gait outcome in ambulatory hemiparetic patients after a 4-week comprehensive rehabilitation program and prognostic factors. Stroke 1994;25:1999-2004. 9. Knutsson E, Richards C. Different types of disturbed motor control in gait of hemiparetic patients. Brain 1979;102:405-30. 10. De Quervain IA, Simon SR, Leurgans S, Pease WS, Mcallister D. Gait pattern in the early recovery period after stroke. J Bone Joint Surg Am 1996;78:1506-14. 11. 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. 12. Bohannon RW. Gait performance of hemiparetic stroke patients: selected variables. Arch Phys Med Rehabil 1987;68:777-81. 13. Bohannon RW, Andrews AW. Correlation of knee extensor muscle torque and spasticity with gait speed in patients with stroke. Arch Phys Med Rehabil 1990;71:240-3. 14. 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. 15. Suzuki K, Imada G, Iwaya T, Handa T, Kurogo H. Determinants and predictors of the maximum walking speed during computerassisted gait training in hemiparetic stroke patients. Arch Phys Med Rehabil 1999;80:179-82. 16. 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. 17. 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. 18. Bohannon RW, Walsh S. Nature, reliability, and predictive value of muscle performance measures in patients with hemiparesis following stroke. Arch Phys Med Rehabil 1992;73:721-5. 19. Davies JM, Mayston MJ, Newham DJ. Electrical and mechanical output of the knee muscles during isometric and isokinetic activity in stroke and healthy adults. Disabil Rehabil 1996;18:83-90. 20. Nadeau S, Gravel D, Arsenault B, Bourbonnais D, Goyette M. Dynamometric assessment of the plantarflexors in hemiparetic subjects: relations between muscular, gait and clinical parameters. Scand J Rehabil Med 1997;29:137-46. 21. Kandel ER, Schwartz JH, Jessell TM. Principles of neural science. 4th ed. New York: McGraw-Hill; 2000. ch 37. 22. Brandstater ME, DeBruin H, Gowland C, Clark BM. Hemiplegic gait: analysis of temporal variables. Arch Phys Med Rehabil 1983;64:583-7. 23. Kusoffsky A, Wadell I, Nilsson BY. The relationship between sensory impairment and motor recovery in patients with hemiplegia. Scand J Rehabil Med 1982;14:27-32. 24. Keenan MA, Perry J, Jordan C. Factors affecting balance and ambulation following stroke. Clin Orthop 1984;Jan-Feb(182):165-71. 25. Dettmann MA, Linder MT, Sepic SB. Relationship among walking performance, postural stability, and functional assessments of the hemiplegic patient. Am J Phys Med 1987;66:77-90. 26. Kuan TS, Tsou JY, Su FC. Hemiplegic gait of stroke patients: effect of using a cane. Arch Phys Med Rehabil 1999;80:777-84. 27. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post stroke hemiplegic patients. 1. A method for evaluation of physical performance. Scand J Rehabil Med 1975;7:13-31. 28. Nurse MA, Nigg M. Quantifying a relationship between tactile and vibration sensitivity of the human foot with plantar pressure

SENSORIMOTOR FUNCTION AND GAIT IN CHRONIC STROKE, Lin 203 distributions during gait. Clin Biomech (Bristol, Avon) 1999;14:667-72. 29. Duncan PW, Propst M, Nelson SG. Reliability of the Fugl-Meyer assessment of sensorimotor recovery following cerebrovascular accident. Phys Ther 1983;63:1606-10. 30. Sanford J, Moreland J, Swanson LR, Stratford PW, Gowland C. Reliability of the Fugl-Meyer assessment for testing motor performance in patients following stroke. Phys Ther 1993;73:447-54. 31. Courtine G, Pozzo T, Lucas B, Schieppati M. Continuous, bilateral Achilles tendon vibration is not detrimental to human walk. Brain Res Bull 2001;55:107-15. 32. Ferrell WR, Baxendale RH, Carnachan C, Hart IK. The influence of joint afferent discharge on locomotion, proprioception and activity in conscious cats. Brain Res 1985;347:41-8. 33. Loewen SC, Anderson BA. Predictors of stroke outcome using objective measurement scales. Stroke 1990;21:2178-81. 34. Granger CV, Greer DS, Liset E, Coulombe J, O Brien E. Measurement of outcome of care for stroke patients. Stroke 1975;6: 34-41. 35. Wade DT, Langston-Hewer R, Wood VA, Skilbeck CE, Ismail HM. The hemiplegic arm after stroke: measurement and recovery. J Neurol Neurosurg Psychiatry 1983;46:521-4. 36. Thorngren M, Westling B, Norrving B. Outcome after stroke in patients discharged to independent living. Stroke 1990;21:236-40. 37. Ferrucci L, Bandinelli S, Guralnik JM, et al. Recovery of functional status after stroke. A postrehabilitation follow-up study. Stroke 1993;24:200-5. 38. Tangeman PT, Banaitis DA, Williams AK. Rehabilitation of chronic stroke patients: changes in functional performance. Arch Phys Med Rehabil 1990;71:876-80. 39. Waagfjord J, Levangie PK, Certo CM. Effects of treadmill training on gait in a hemiparetic patient. Phys Ther 1990;70:549-60. 40. Breniere Y, Do MC. When and how does steady state gait movement induced from upright posture begin? J Biomech 1986;19: 1035-40. 41. Kendall FP, McCreary EK, Provance PG. Muscles: testing and function. 4th ed. Baltimore: Williams & Wilkins; 1993. p 201-13. 42. Perry J. Gait analysis: normal and pathological function. New York: Slack; 1992. ch 2, p 4-6. 43. Winter DA. The biomechanics and motor control of human gait: normal, elderly and pathological. 2nd ed. Waterloo (ON): Waterloo Univ Pr; 1991. ch 2. 44. Teixeira-Salmela LF, Nadeau S, Mcbride I, Olney SJ. Effects of muscle strengthening and physical conditioning training on temporal, kinematic and kinetic variables during gait in chronic stroke survivors. J Rehabil Med 2001;33:53-60. 45. Kuan TS, Su FC. Correlation between clinical status and gait performance of hemiplegic stroke patients. J Rehabil Med Assoc ROC 2001;29:1-13. 46. Pinzur MS, Sherman R, DiMonte-Levine P, Trimble J. Gait changes in adult onset hemiplegia. Am J Phys Med 1987;66:228-37. 47. Chou SW, Wong AM, Leong CP, Tang FT, Lin TH. Postural control during sit-to-stand and gait in stroke patients. Am J Phys Med Rehabil 2003;82:42-7. 48. Nurse MA, Nigg BM. The effect of changes in foot sensation on plantar pressure and muscle activity. Clin Biomech (Bristol, Avon) 2001;16:719-27. 49. Dingwell JB, Cavanagh PR. Increased variability of continuous overground walking in neuropathic patients is only indirectly related to sensory loss. Gait Posture 2001;14:1-10. 50. Courtemanche R, Teasdale N, Boucher P, Fleury M, Lajoie Y, Bard C. Gait problems in diabetic neuropathic patients. Arch Phys Med Rehabil 1996;77:849-55. 51. Tyson SF. Hemiplegic gait symmetry and walking aids. Physiother Theory Pract 1994;10:153-9. 52. Tyson SF. Trunk kinematics in hemiplegic gait and the effect of walking aids. Clin Rehabil 1999;13:295-300. Suppliers a. Vicon 370; Vicon Motion Systems, 14 Minns Business Park, West Way, Oxford, OX2 0JB, UK. b. PowerTrack II Commander; JTech Medical Industries, 470 Lawndale Dr, Ste G, Salt Lake City, UT 84115. c. Norotrack 360; Myotronics-Noromed, 15425-53rd Ave S, Tukwila, WA 98188. d. Release 10.01; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.