STROKE IS A MAJOR cause of disability and handicap in

1046 ORIGINAL ARTICLE Contribution of Ankle Dorsiflexor Strength to Walking Endurance in People With Spastic Hemiplegia After Stroke Shamay S. Ng, PhD, Christina W. Hui-Chan, PhD ABSTRACT. Ng SS, Hui-Chan CW. Contribution of ankle dorsiflexor strength to walking endurance in people with spastic hemiplegia after stroke. Arch Phys Med Rehabil 2012;93: 1046-51. Objectives: (1) To determine the relationships of ankle dorsiflexor strength, ankle plantarflexor strength, and spasticity of the ankle plantarflexors with walking endurance; (2) to determine whether affected ankle dorsiflexor strength makes an independent contribution to walking endurance; and (3) to quantify its relative contribution to the walking endurance of people with spastic hemiplegia after stroke. Design: A cross-sectional study. Setting: University-based rehabilitation center. Participants: Subjects (N 62) with spastic hemiplegia. Interventions: Not applicable. Main Outcome Measures: Walking endurance was measured by the distance covered in the six-minute walk test (6MWT). Ankle dorsiflexor and plantarflexor strength were measured using a load-cell mounted on a custom-built foot support. Plantarflexor spasticity was measured using the Composite Spasticity Scale. Results: The six-minute walk distances showed stronger positive correlation with affected dorsiflexor strength (r.793, P.000) when compared with affected plantarflexor strength (r.349, P.005). Results of the regression model showed that after adjusting for basic demographic and stroke-related impairments, affected ankle dorsiflexor strength remained independently associated with six-minute walk distance, accounting for 48.8% of the variance. Conclusions: This is the first study, to our knowledge, to document the importance of ankle dorsiflexor strength as an independent determinant of walking endurance in stroke survivors with spastic plantarflexors. Our findings suggest that stroke rehabilitation programs aiming to improve walking endurance should include strengthening exercises for the ankle dorsiflexors. Key Words: Ankle; Rehabilitation; Stroke; Walking. 2012 by the American Congress of Rehabilitation Medicine From the Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hong Kong (SAR), China (Ng); and the Department of Physical Therapy, University of Illinois at Chicago, Chicago, IL (Hui-Chan). Supported by the Development of Niche Areas Funding to Centre for East-meets West in Rehabilitation Sciences of The Hong Kong Polytechnic University. 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 authors or on any organization with which the authors are associated. Reprint requests to Christina W. Hui-Chan, BPT, MSc, PhD, Dept of Physical Therapy, University of Illinois at Chicago, 1919 W Taylor St (MC 898), Chicago, IL 60612, e-mail: chuichan@uic.edu. In-press corrected proof published online on Mar 22, 2012, at www.archives-pmr.org. 0003-9993/12/9306-01057$36.00/0 doi:10.1016/j.apmr.2011.12.016 STROKE IS A MAJOR cause of disability and handicap in older adults, with incidence approximately doubling each decade after age 55. 1 It often results in deficits in walking capacity, including reduced walking velocity and endurance. 2-4 The ability to walk independently with sufficient endurance that permits participation at home and in the community is an important goal of neurologic rehabilitation after stroke, 5-9 so clinicians need to understand the major determinants of the walking endurance of stroke survivors. The six-minute walk test (6MWT) was initially introduced as a simple walking test for patients with respiratory 10 or cardiac problems. 11 Patients being tested are required to walk back and forth on a walkway continuously as far as they can in 6 minutes, and the distance covered in 6 minutes is recorded. The six-minute walk distance has shown excellent reliability, with intraclass correlation coefficients ranging from.94 to.99 in most studies of subjects with subacute stroke 12,13 and chronic stroke. 2,14 Moreover, the 6MWT has proven clinically sensitive for detecting changes because of its small standard error of measurement of 4.8%. 13 The 6MWT is now widely used as an outcome measure for walking endurance among subjects with mild to moderate, 15 acute, 16 and chronic stroke. 17,18 The importance of ankle plantarflexor strength for proper gait has been widely documented in stroke survivors. The ankle plantarflexors generate most of the energy required to move the legs forward. 19,20 Among survivors of strokes of different severity, the strength of their affected ankle plantarflexors has been shown to correlate well with their walking speed (correlation coefficients range,.83.845), 4,21 and with their Timed Up & Go test scores (correlation coefficient.86). 2 Using multiple stepwise regression analyses, Kim and Eng 4 identified the strength of the ankle plantarflexor as the single most important factor in predicting walking velocity, explaining 67% to 72% of its variance. Decreased ankle dorsiflexion of the affected leg during the swing phase is a common clinical observation in hemiparetic gait. 3 Inadequate ankle dorsiflexion can cause the toes to drag on the floor, leading to decreased step length and walking velocity and increased risks of falling. Supplementing any ankle dorsiflexor weakness after stroke, spastic ankle plantarflexors might also act as active restraints restricting ankle dorsiflexion. 22 However, the role of affected ankle dorsiflexor strength and the relationship between affected ankle dorsiflexor and plantarflexor strength in walking performance have not yet been systematically delineated in people with spastic plantarflexors after stroke. The objectives of this study were: (1) to elucidate the relationships of ankle dorsiflexor strength, ankle plantarflexor AFO CSS 6MWT List of Abbreviations ankle-foot orthosis Composite Spasticity Scale six-minute walk test

ANKLE IMPAIRMENTS AND WALKING ENDURANCE, Ng 1047 strength, and spasticity of the ankle plantarflexors with walking endurance; (2) to determine whether affected ankle dorsiflexor strength makes an independent contribution to walking performance, and (3) to quantify its relative contribution to the walking endurance of people with spastic hemiplegia after stroke. METHODS Participants Adult subjects with spastic hemiplegia after stroke were recruited through a community rehabilitation network in Hong Kong. Subjects were included if they had suffered a single stroke at least 1 year previously, were able to walk 10m unassisted with a cane without any ankle-foot orthosis (AFO), had spastic ankle plantarflexors with a Composite Spasticity Scale (CSS) score 10, 23 and had at least 10 of passive ankle dorsiflexion. Candidates were excluded if they had any medical comorbidity, receptive dysphasia, or any cognitive impairment denoted by scoring less than 7 of 10 on the Abbreviated Mental Test. 24 Other orthopedic and medical conditions that would hinder proper assessment were also grounds for exclusion. The ethics committee of the local institution approved this cross-sectional study, which was conducted in accordance with the Helsinski Declaration of 1975, as revised in 1983. Informed consent was obtained from all subjects prior to the study. Protocol Information was first collected about each subject s general characteristics including age, sex, side of hemiparesis, body mass index, and poststroke duration. These results are shown in table 1. Spasticity of the ankle plantarflexor. Ankle plantarflexor spasticity was measured using Chan s 23 CSS, which has been shown to be reliable and valid for people after stroke. 2,25 The CSS is an ordinal scale based on clinical assessment of (1) Achilles tendon jerks (0 4 for a 5-point scale), (2) resistance to passive ankle dorsiflexion (0 4 for a 5-point scale, doubly weighted), and (3) the amount and duration of ankle clonus Table 1: Descriptive Characteristics of the Subjects (N 62) Characteristics Mean SD (range) or No. (%) Age 57.4 7.8 (45 78) Sex Men 51 (82.3) Women 11 (17.7) Hemiplegic side Left 43 (69.4) Right 19 (30.6) Body mass index (kg/m 2 ) 25.0 3.1 (18.4 31.8) Years since stroke 5.2 3.7 Affected plantarflexor spasticity (CSS score) 12.2 1.7 (10 16) Abbreviated Mental Test score 9.2 1.1 (6 10) 6MWT distance (m) 183.7 84.4 (33 361) Gait velocity (cm/s) 51.5 26.1 (13.3 109.7) Peak dorsiflexion torque (Nm) Unaffected 22.8 5.6 (12.2 38.3) Affected 14.2 6.7 (6.3 35.3) Peak plantarflexion torque (Nm) Unaffected 29.9 4.8 (20.4 40.5) Affected 20.2 7.8 (10.0 36.2) (1 4 for a 4-point scale). Because the score for resistance to passive dorsiflexion most closely represents tone, it was doubly weighted. 23 These 3 scores were then summed to generate the total spasticity score. A total score ranging from 1 to 6 is considered to represent no spasticity, 7 to 9 as mild spasticity, 10 to 12 as moderate spasticity, and 13 to 16 as severe spasticity. 26 Maximum isometric voluntary contraction of the ankle muscles. Peak torque during maximum isometric voluntary contraction of the ankle dorsiflexors and plantarflexors was recorded using a load cell mounted on a foot frame. 2 The leg was fixed, keeping the knee flexed at about 50 with the ankle in a neutral position. Subjects were required to dorsiflex and plantarflex their ankles maximally and hold for about 3 seconds. The peak torques of 3 trials were recorded. Gait velocity. Gait velocity was measured using a 4.6-m instrumented mat (GAITRite II walkway system a ). Subjects were required to walk in their own comfortable footwear at their normative walking speed on the instrumented carpet for 3 trials, using their usual walking aids (ie, cane). The sampling rate was 80Hz, and data were processed by the GAITRite software (version 2.2). a The average of 3 trials was used for analysis. Walking endurance. Walking endurance was assessed in terms of the distance covered in the 6MWT. 11 The subjects walked up and down a 33-m corridor using their usual walking aids (ie, cane). Standardized encouragements were given at 1, 3, and 5 minutes during the walk: You re doing a good job (minute 1), You re halfway done (minute 3), and You have 1 minute to go (minute 5). To ensure safety, the investigator walked slightly behind but not beside the subjects, so as to avoid influencing their self-selected walking pace. Subjects were allowed to stop and rest as they deemed necessary. The distance covered, to the nearest centimeter, was recorded. The sequence of the test was administered in a random order. All measurement protocols had been tested for their reproducibility in our pilot study on chronic stroke patients, with intraclass correlation coefficients of.69 to.99. 2 Statistical Analysis Descriptive statistics were used to summarize the demographic data and all the outcome measurements. The distribution of the data was checked using a Kolmogorov-Smirnov test of normality. The relationship of the six-minute walk distance with the other variables was evaluated using Pearson correlation coefficients if the data were normally distributed; otherwise Spearman correlation coefficients were calculated. Multiple linear regression with enter method was applied to determine what proportion of the variability in the dependent variable (ie, the six-minute walk distance) could be explained by other independent variables together with ankle dorsiflexor and plantarflexor strength, and which variables had the most predictive power for the six-minute walk distance (the largest standardized regression coefficients). All the statistical analyses were performed with Statistical Package for the Social Sciences software, b mandating a confidence level of.05 (2- tailed) for significance. RESULTS Sixty-two subjects completed all the experimental procedures over a period of 4 years. Table 1 summarizes the subjects demographic characteristics. The 51 men and 11 women had a mean age SD of 57.4 7.8 years. Their average body mass index was 25.0 3.1kg/m 2. Forty-three subjects had left hemiplegia and 19 had right hemiplegia, with an average

1048 ANKLE IMPAIRMENTS AND WALKING ENDURANCE, Ng Table 2: Pearson Correlation Coefficients Between Six-Minute Walk Distance and All Variables (N 62) Variables Pearson Correlation Coefficients With Six-Minute Walk Distance Age.090.488 Body mass index.073.572 Years since stroke.066.609 Abbreviated Mental Test scores.020.875 Plantarflexor spasticity (CSS scores).062.635 Peak dorsiflexor torque Affected.793*.000 Unaffected.146.258 Peak plantarflexor torque Affected.349*.005 Unaffected.168.192 *Significance at.01% level of confidence. poststroke duration of 5.2 3.7 years. The mean CSS scores were 12.2 1.7, suggesting that all participating subjects had moderate to severe spasticity in their affected plantarflexors. 23,26 The correlations between the six-minute walk distance and the other variables are shown in table 2. The six-minute walk distance showed the highest positive correlation with affected ankle dorsiflexor strength (r.793, P.000). The six-minute walk distance also showed significant positive correlation with affected ankle plantarflexor strength (r.349, P.005). The demographic data including age, body mass index, ankle plantarflexor spasticity (CSS scores), and unaffected ankle dorsiflexor and plantarflexor strength accounted for 4.9% of the six-minute walk distance (model 1 in table 3). The addition of affected ankle plantarflexor strength accounted for an additional 11.3% of the six-minute walk distance (model 2 in table P 3). With further addition of affected ankle dorsiflexor strength (model 3 in table 3), the multiple linear regression model showed that affected ankle dorsiflexor strength was independently associated with the six-minute walk distance, accounting for 48.8% of the six-minute walk distance variance (model 3 in table 3). The addition of ankle dorsiflexor strength significantly improved the model s predictive power (F 1,54 75.197, P.000), and the final regression model (model 3 in table 3) explained 60.4% of the variance in the six-minute walk distance. Among all the variables, the strength of the affected ankle dorsiflexor was the most potent factor determining the six-minute walk distance, as reflected by the magnitude of the standardized regression coefficient (.750) (model 3 in table 3) and the Pearson correlation coefficient (r.793) (see table 2) when basic demographic data, plantarflexor spasticity, and unaffected dorsiflexor and plantarflexor strength were taken into account. DISCUSSION The findings of this study contribute to a growing body of results highlighting the importance of ankle dorsiflexor strength in improving gait velocity and gait symmetry in stroke rehabilitation. 3,27,28 In addition to being the most important factor determining walking speed and temporal symmetry during walking among all tested variables in a previous study, 3 ankle dorsiflexor strength is also an independent determinant of walking endurance as measured by the 6MWT. Our results show that ankle dorsiflexor strength accounts for 48.8% of the variance in six-minute walk distances when basic demographic data, stroke-related impairments, and unaffected ankle dorsiflexor and plantarflexor strength are taken into account (see table 3). Reduced Six-Minute Walk Distance and Ankle Muscle Strength After Stroke The mean six-minute walk distances covered by these subjects with spastic hemiplegia after stroke (183.7 84.4m) were Table 3: Multiple Linear Regression (Enter Method) Using Six-Minute Walk Distance as the Dependent Variable (N 62) Independent Variables R 2 (R 2 adj) R 2 Change B (SE) (unstandardized coefficient) (standardized coefficient) Model 1 0.049 ( 0.036) 0.049 Age 1.171 (1.491).108.435 BMI 0.556 (3.744).020.882 CSS scores 4.047 (6.570).082.540 Peak DF torque (unaffected) 1.508 (2.274).101.510 Peak PF torque (unaffected) 1.858 (2.702).107.494 Model 2 0.162 (0.070) 0.113 Age 1.455 (1.416).135.309 BMI 1.963 (3.666).072.595 CSS scores 5.977 (6.264).122.344 Peak DF torque (unaffected) 1.723 (2.156).115.428 Peak PF torque (unaffected) 0.451 (2.611).026.863 Peak PF torque (affected) 3.926 (1.444).361.009* Model 3 0.650 (0.604) 0.488 Age 0.530 (0.930).071.399 BMI 0.198 (2.405).008.923 CSS scores 2.657 (4.105).057.482 Peak DF torque (unaffected) 0.261 (1.425).001.989 Peak PF torque (unaffected) 0.857 (1.704).049.617 Peak PF torque (affected) 1.256 (0.991).116.210 Peak DF torque (affected) 9.461 (1.091).750.000* Abbreviations: BMI, body mass index; DF, dorsiflexion; PF, plantarflexion. *Significance at.01% level of confidence. P

ANKLE IMPAIRMENTS AND WALKING ENDURANCE, Ng 1049 shorter than those observed in previous studies (207.9 216m). 17,29 Baseline differences among the subjects in different studies may account for the variability observed, because using walking distance to measure endurance demands a certain amount of lower-limb strength and balance from the subjects. Note that six-minute walk distances are also influenced by other factors such as motivation, cardiorespiratory, and cardiovascular fitness. 11 All our subjects used a cane as a walking aid, and were not wearing any AFO during the 6MWT and gait velocity test. A previous study had investigated the effects of walking aids on walking performance. 30 Allet et al 30 found that longer walking distance in the 6MWT could be covered by stroke patients using simple canes with an ergonomic handgrip when compared with those using a 4-point cane or Nordic stick. The effects of walking aids on six-minute walk distance covered are an area that warrants further study. Decreased walking endurance as measured by six-minute walk distance has previously been reported after stroke. 2,7 The six-minute walk distance for healthy adult men and women can be predicted using equations developed by Enright and Sherrill. 31 For healthy men, they suggest six-minute walk distance in meters (7.57 height, cm) (5.02 age, y) (1.76 weight, kg) 309. For healthy women, they suggest 6MWD in meters (2.11 height, cm) (5.78 age, y) (2.29 weight, kg) 667. These equations would predict distances around 500m for the subjects in this study if they had been healthy, but subjects with stroke could complete only about 38% of the predicted distance. This is consistent with the results of a study by Mayo et al, 7 which found only 40% of the age-predicted performance in subjects with chronic stroke. This is important, as Mayo 7 further identified six-minute walk distances as the best predictor of community integration for stroke survivors available today. 7 Clearly, the increased energy consumption involved in hemiplegic gait, 32,33 physical deconditioning after stroke, and age-associated declines in fitness and muscle mass compromise stroke survivors capacity to meet the high-energy demands of the activities of daily living. 34,35 Decreased ankle dorsiflexor and plantarflexor strength on the affected side are well documented. 2,3,36 Such weakness is multifactorial, and it could be attributed to decreased activation of agonist motor units, a reduced number of functioning motor units, reduced firing rates of agonists, and increased cocontraction of the antagonists during movement of the paretic leg. 22,37 In addition, spastic ankle plantarflexors might have difficulties in generating the muscle force as agonists during ankle plantarflexion, and might also act as active restraints during ankle dorsiflexion. 22 The net ankle dorsiflexion and plantarflexion force generated was therefore decreased. It may not be valid to compare our present findings on ankle muscle strength with those of other studies, because a custom-built foot frame was used in this study to measure the ankle muscle strength with subjects in lying in a supine position and their knee fixed at 50, which was different from the experimental setups and assessment protocols of previous studies. Ankle Dorsiflexor Strength Independently Predicts Six- Minute Walk Distance The results of the multiple linear regression analyses show that ankle dorsiflexor strength was the most potent predictor of six-minute walk distance among all tested variables, explaining 48.8% of the variance. A conservative multiple regression approach was used in which ankle dorsiflexor strength was entered last. This was done to quantify the independent contribution of ankle dorsiflexor strength to walking endurance after controlling for basic demographics and known confounding variables like ankle plantarflexor strength. Ankle dorsiflexor weakness may cause inadequate foot clearance during the swing phase and insufficient eccentric contraction during the weight-transfer phase after heel-strike. The latter deficiencies may lead to increased swing time, 27 decreased single-leg support time, and increased double-leg support time, 28 thus reducing walking velocity. These findings are consistent with those of Lin et al 3 who showed that ankle dorsiflexor strength was the primary determinant of walking velocity, and that it could explain up to 30% of the variance using stepwise regression analysis. Because subjects in our study had spastic plantarflexors (mean composite spasticity score 12) (see table 1), greater ankle dorsiflexor strength and/or hip and knee flexion would be required to clear the ground during the swing phase of each stride. Total energy consumption might be increased and walking endurance decreased. Indeed, a significant correlation was observed between affected ankle dorsiflexor strength and walking velocity (r.727, P.000), and walking endurance (r.793, P.000) (see table 2). Plantarflexor spasticity did not, however, correlate with sixminute walk distance (r.062, P.635) (see table 2) or walking velocity (r.178, P.158). In previous research using the CSS, no significant correlations were found between ankle plantarflexor spasticity and comfortable walking velocity 19 or Timed Up & Go scores. 2 In contrast, ankle plantarflexor spasticity has been found to be the most important independent determinant of gait asymmetry (R 2.45.76) during walking at both comfortable and fast speeds among all tested variables. 36 Several studies have also reported that injecting botulinum toxin type A into spastic plantarflexors to reduce their spasticity can improve the walking velocity of stroke patients. 38-41 The negative results in this study may be due to the way ankle plantarflexor spasticity was measured in a relaxed supine lying position and using the CSS, while other studies used different protocols and outcome measures. Different severity of ankle plantarflexor spasticity among the subjects recruited for the different studies would also be expected to entail different degrees of disturbance to motor control of the ankle. It is surprising to note that among all variables, ankle plantarflexor strength was relatively less potent in determining the six-minute walk distance when compared with affected ankle dorsiflexor strength, as reflected by the magnitude of the standardized regression coefficient (see table 3). Previous studies have shown that walking velocity was significantly correlated with the strength of the hip flexors 19,36 and ankle plantarflexor. 2,19,20 In fact, Kim and Eng 4 identified ankle plantarflexor rather than ankle dorsiflexor strength as the most important factor in determining walking velocity. Besides methodologic differences in measuring muscle strength, these inconsistent findings among different studies may have resulted from differences in the subjects recruited. The subjects studied here had more severe ankle plantarflexor spasticity (mean CSS score, 12) and walked slightly faster (averaging 0.53cm/s 1 ) than those studied by Kim and Eng 4 (mean Modified Ashworth Scale score, 0.9; mean walking velocity, 0.45cm/s 1 ). Stronger ankle plantarflexors may have improved their subjects ability to propel their bodies forward during walking. However, in patients who walk faster, propulsion ability may not be as important. Although the six-minute walk distances observed in this study were significantly correlated with the plantarflexor strength of the affected leg (r.349, P.005), the effect of the ankle plantarflexor strength was weakened in multivariable modeling. This may be due to the significant positive correlation between affected plantarflexor strength and walking ve-

1050 ANKLE IMPAIRMENTS AND WALKING ENDURANCE, Ng locity (r.318, P.012), and the affected ankle dorsiflexor strength (r.299, P.018). Study Limitations Note that the quality of walking in performing the 6MWT might be overlooked because the distance covered was the main focus of the current study. In addition, the number of turns and rest time during the 6MWT had not been recorded, which might have affected the total distance covered in the 6MWT. All our stroke subjects used walking aids (ie, cane) and did not use AFOs during the 6MWT. The reliance on canes during the 6MWT may be due to patients insufficient muscle strength of the lower limbs and/or poor balance ability. Nevertheless, the results of our study may not be generalized to stroke patients who used different types of walking aids and AFOs. For example, Lairamore et al 42 found that use of an AFO caused a decline in tibialis anterior muscle activity during the swing phase of gait. Approximately 39.6% of the variance in the six-minute walk distances remained unexplained by the regression models. The 6MWT performance has multiple determinants, some of which were not measured in this study, including muscle strength of hip and knee flexors and extensors, lower-limb proprioception, tactile sensation, balance performance, and cardiopulmonary fitness. Because of the cross-sectional design, this study could not establish any causal relationships among the variables. Because all the subjects in this study had spastic plantarflexors, the results of this study should not be generalized to a general stroke population. CONCLUSIONS Ankle dorsiflexor strength is independently associated with the walking endurance of stroke survivors with spastic ankle plantarflexors, accounting for 48.8% of the variance in sixminute walk distance, which is the best predictor of their community integration. 7 The findings of the current study support the recommendation to incorporate ankle dorsiflexor strength training into stroke rehabilitation programs in order to improve walking endurance and community reintegration. Acknowledgment: We thank the Community Rehabilitation Network of the Hong Kong Society for Rehabilitation for their assistance in recruiting subjects. References 1. Feigin VL, Lawes CM, Bennett DA, Anderson CS. Stroke epidemiology: a review of population-based studies of incidence, prevalence and case-fatality in the late 20th century. Lancet Neurol 2003;2:43-53. 2. Ng SS, Hui-Chan CW. The timed up & go test: its reliability and association with lower-limb impairments and locomotor capacities in people with chronic stroke. Arch Phys Med Rehabil 2005;86: 1641-7. 3. Lin PY, Yang YR, Cheng SJ, Wang RY. The relation between ankle impairments and gait velocity and symmetry in people with stroke. Arch Phys Med Rehabil 2006;87:562-8. 4. Kim CM, Eng JJ. The relationship of lower-extremity muscle torque to locomotor performance in people with stroke. Phys Ther 2003;83:49-57. 5. Bohannon RW, Andrews AW, Smith MB. Rehabilitation goals of patients with hemiplegia. Int J Rehabil Res 1988;11:181-2. 6. Bohannon RW, Horton MG, Wikholm JB. Importance of 4 variables of gait to patients with stroke. Int J Rehabil Res 1991;14: 246-50. 7. Mayo NE, Wood-Dauphinee S, Ahmed S, et al. Disablement following stroke. Disabil Rehabil 1999;21:258-68. 8. Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking handicaps in the stroke population. Stroke 1995;26:982-9. 9. Lord SE, McPherson K, McNaughton HK, Rochester L, Weatherall M. Community ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil 2004;85:234-9. 10. Butland RJ, Pang J, Gross ER, Woodcock AA, Geddes DM. Two-, six- and twelve-minute walking tests in respiratory disease. Brit Med J 1982;284:1607-8. 11. Guyatt GH, Sullivan MJ, Thompson PJ, et al. The 6-minute walk: a new measure of exercise capacity in patients with chronic heart failure. Can Med Assoc J 1985;132:919-23. 12. Liu J, Drutz C, Kumar R, et al. Use of the 6-minute walk test poststroke: is there a practice effect? Arch Phys Med Rehabil 2008;89:1686-92. 13. Flansbjer UB, Holmback AM, Downham D, Patten C, Lexell J. Reliability of gait performance tests in men and women with hemiparesis after stroke. J Rehabil Med 2005;37:75-82. 14. Eng JJ, Dawson AS, Chu KS. Submaximal exercise in persons with stroke: test-retest reliability and concurrent validity with maximal oxygen consumption. Arch Phys Med Rehabil 2004;85: 113-8. 15. Duncan PW, Richards CL, Wallace D, et al. A randomized, controlled pilot study of a home-based exercise program for individuals with mild and moderate stroke. Stroke 1998;29:2055-60. 16. Pohl PS, Duncan PW, Perera S, et al. Influence of stroke-related impairments on performance in a 6-minute walk test. J Rehabil Res Dev 2002;39:1-6. 17. Dean CM, Richards CL, Malouin F. Task-related circuit training improves performance of locomotor tasks in chronic stroke: a randomized, controlled pilot trial. Arch Phys Med Rehabil 2000; 81:409-17. 18. Ng SS, Hui-Chan CW. Does the use of TENS increase the effectiveness of exercise for improving walking after stroke? A randomized controlled clinical trial. Clin Rehabil 2009;23:1093-103. 19. 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. 20. Olney SJ, Griffin MP, Monga TN, McBride ID. Work and power in gait of stroke patients. Arch Phys Med Rehabil 1991;72:309-14. 21. 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. 22. Ng SS, Shepherd RB. Weakness in patients with stroke: implications for strength training in neurorehabilitation. Phys Ther Rev 2000;5:227-38. 23. Chan CW. Motor and sensory deficits following a stroke: relevance to a comprehensive evaluation. Physiother Can 1986;38:29-34. 24. Hodkinson HM. Evaluation of a mental test score for assessment of mental impairment in the elderly. Age Ageing 1972;1:233-8. 25. Nadeau S, Arsenault AB, Gravel D, Lepage Y, Bourbonnais D. Analysis of the spasticity index used in adults with a stroke. Can J Rehabil 1998;11:219-20. 26. Levin MF, Hui-Chan CW. Relief of hemiparetic spasticity by TENS is associated with improvement in reflex and voluntary motor functions. Electroencephalogr Clin Neurophysiol 1992;85: 131-42. 27. 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. 28. Whittle MW. Gait analysis: an introduction. Oxford: Butterworth- Heinemann; 2003. 29. Salbach NM, Mayo NE, Robichaud-Ekstrand S, Hanley JA, Richards CL, Wood-Dauphinee S. Balance self-efficacy and its relevance to physical function and perceived health status after stroke. Arch Phys Med Rehabil 2006;87:364-70.

ANKLE IMPAIRMENTS AND WALKING ENDURANCE, Ng 1051 30. Allet L, Leemann B, Guyen E, et al. Effect of different walking aids on walking capacity of patients with poststroke hemiparesis. Arch Phys Med Rehabil 2009;90:1408-13. 31. Enright PL, Sherrill DL. Reference equations for the six-minute walk in healthy adults. Am J Respir Crit Care Med 1998;158: 1384-7. 32. Gerston J, Orr W. External work of walking in hemiparetic patients. Scand J Rehabil Med 1971;3:85-8. 33. Corcoran PJ, Jebson RH, Brengelman GL, Simons BC. Effects of plastic and metal leg braces on speed and energy cost of hemiparetic ambulation. Arch Phy Med Rehabil 1970;51:69-77. 34. Hagberg JM. Effects of training on the decline of VO 2 max with aging. Fed Proc 1987;46:1830-3. 35. Schwartz RS, Buchner DM. Exercise in the elderly: physiological and functional effects. In: Hazard WR, editor. Principles of geriatric medicine and gerontology. 3rd ed. New York: McGraw-Hill; 1994. p 91-105. 36. 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. 37. Bourbonnais D, Vanden Noven S. Weakness in patients with hemiparesis. Am J Occup Ther 1989;43:313-9. 38. Hesse S, Lucke D, Malezic M, Bertelt C, Friedrich H, Gregoric M. Botulinum toxin treatment for lower limb extensor spasticity in chronic hemiparetic patients. J Neurol Neurosurg Psychiatry 1994;57:1321-4. 39. Hesse S, Jahnke MT, Luecke D, Mauritz KH. Short-term electrical stimulation enhances the effectiveness of botulinum toxin in the treatment of lower limb spasticity in hemiparetic patients. Neurosci Lett 1995;201:37-40. 40. Hesse S, Krajinik J, Luecke D, Jahnke MT, Gregoric M, Mauritz KH. Ankle muscle activity before and after botulinum toxin therapy for lower limb extensor spasticity in chronic hemiparetic patients. Stroke 1996;27:455-60. 41. Reiter F, Danni M, Lagalla G, Ceravolo G, Provinciali L. Lowdose botulinum toxin with ankle taping for the treatment of spastic equinovarus foot after stroke. Arch Phys Med Rehabil 1998;79: 532-5. 42. Lairamore C, Garrison MK, Bandy W, Zabel R. Comparison of tibialis anterior muscle electromyography, ankle angle and velocity when individuals post stroke walk with different orthoses. Prosthet Orthot Int 2011;35:402-10. Suppliers a. CIR Systems Inc, 60 Garlor Dr, Havertown, PA 19083. b. Version 16.0; SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.