Kozol, Menachem Z. PhD, PT; Filer, Maya PT; Ring, Haim MD
Cerebral vascular accident (CVA) is a leading cause of disability for persons older than 65 years.1,2 Many studies have focused on impaired performance of daily activities, such as gait,3-5 rising to standing up,6 or upper-extremity function.7 However, far less research attention has been devoted to the ability to perform pelvic lifting, often referred to as “bridging.” Bridging capability is essential for maintaining the quality of life of adults with hemiplegia post-CVA in that it is a component of common daily activities, such as moving in bed, transferring, dressing, and using a bed pan.8-10 It may also serve as a measure in preventing bed sores11 and assessing the fitness of bedridden individuals.12
Bridging is initiated from a hook-lying position with hips and knees flexed and ankles plantarflexed against the supporting surface. From this position, the individual raises the pelvis and lumbar and lower thoracic areas off the bed by mainly extending the hips and knees. This is accomplished by transmitting force through the weight-bearing lower extremities.8-10 After a CVA, the ability of the paretic lower extremity to accommodate the biomechanical demand of weight bearing is compromised. As a result, the limb often slides into extension, interfering with the ability to lift the pelvis.8,9
In the past, clinicians attributed this sliding to hypertonicity or spasticity of the quadriceps femoris, acting as extensors of the knee.8,9 Recent investigators consider this explanation to be anecdotal13 and have attributed the limb-sliding phenomenon to reduced strength and motor control over the agonists, rather than to excessive activity of the spastic antagonists.2 Other investigators doubt the mutuality of tone and muscle strength, suggesting instead that they may affect different properties of performance. For instance, hyperextension of the knee and gait asymmetry have been associated with spasticity,2,14 whereas low walking speed,4,5,15 asymmetry during the transition from sitting to standing,6,16 and inability to lift the upper extremity7 have been associated with reduced muscle strength.
When the knees are bent so that the feet are close to the hips and when the friction between the supporting surface and the feet is high, the tendency for sliding during bridging performance is minimal such that the sliding control system is barely challenged.9,10,17 On the other hand, bridging on a slippery low-friction surface stresses the motor control mechanism and may provide information about its integrity (M. Z. Kozol, unpublished data, November 2007). Therefore, comparing the impairments of adults with hemiparesis who can perform the exercise on a slippery as well as on a rough surface with those who can prevent sliding only on the rough surface may assist to explore the differences between the motor control statuses of both groups.
In addition, identifying the relationships between the various impairments (tone, motor control, muscle strength) and between the impairments and bridging performance may point to the neural commonality of the deficits and to their manifestation in functional performance.18 Hence, the purposes of this study were to (1) explore the association between the limb-sliding phenomenon and the various motor impairments associated with hemiparesis and (2) examine possible interrelationships between these impairments.
A convenience sample of 27 adults with hemiparesis was recruited from the patient population of a rehabilitation center (Table 1). Subjects included 23 men and 4 women, with a mean age of 57.7 (SD = 9.5) years and a mean time since stroke onset of 6.2 (SD = 2.8) weeks. Fifteen subjects had right hemiparesis and 12 had left hemiparesis. Inclusion criteria were a single CVA in the territory of the middle cerebral artery 3 to 10 weeks prior to the investigation, ability to follow simple instructions, a medically stable condition with no major sensory or perceptional impairment, and no known major orthopedic or neurological disorders prior to the infarct. According to the attending physical therapist, all participants were able to perform functional, nonstandardized bridging on a regular firm vinyl-covered mattress, during the physical therapy sessions.
Three instruments were used for evaluating impairment of the lower extremities: (1) the Fugl-Meyer (FM) Motor Control Scale,19 (2) the Modified Ashworth Spasticity Rating Scale,13 and (3) a digital handheld dynamometer (HHD).
The FM lower-extremity (FM-LE) Motor Control subscale is an ordinal measure composed of 6 sections that rate muscular control, coordination, and reflexes. The premise beyond the FM assessment is that it is easier to control gross flexion or extension synergy of the limb as a unit than to isolate a single joint movement out of the total synergy. Each section of the FM-LE test grants from 4 to 6 points, with a maximum total of 34 points for unimpaired performance. The validity and reliability of the FM score has previously been evaluated and supported.19-21 In this study, the FM-LE score was used to quantify subjects' level of motor control.
The Modified Ashworth Scale (MAS)13 is an ordinal scale used to rate the severity of spasticity or hypertonicity, based on resistance to passive movement (Table 2). The reliability of the MAS is considered moderate but sufficient for clinical studies.13,22,23 In this study, the MAS ratings indicated severity of hypertonicity.
A digital HHD CSD100 (John Chatillon & Sons, Inc, Ametek-Chatillon, Largo, Florida), used for evaluating muscle strength, is composed of a digital screen, a strain gauge, and a pad at the edge. The tester holds the dynamometer and places the pad on the limb, distal to the attachment of the evaluated muscles. Then 1 of 2 techniques—either the “make” technique or the “break” technique—is applied. In the “make” technique, the subject is requested to push as hard as possible against the dynamometer while it is being held in a static position by the examiner. In the “break” technique, the subject must try to maintain a static position while the examiner pushes the dynamometer against the limb until the resistance is broken. Multiplication of the dynamometer's reading by the distance from the force application point to the joint axis provides the moment created by the resisting muscles. The measurements of the HHD were found to be reliable for evaluating the muscle strength of healthy individuals24 and individuals with hemiparesis.25-27
The experiment was approved by the local ethics committee, and an informed consent was obtained from all participants. The participants were evaluated by 2 physical therapists with more than 5 years of experience in neurological rehabilitation. One investigator used the MAS to rate spasticity and the other evaluated the maximum moment of the knee muscles and the level of motor control by using the HHD and the FM-LE, respectively.
The evaluation was conducted in 2 separate sessions. The initial brief session of 20 minutes confirmed the ability of the subjects to perform the bridging from a hook-lying position on a regular vinyl-covered mattress, with the hips and knees at 45° to 50° and 95° to 100° of flexion, respectively, as verified by a goniometer. Then the participants were asked to lift their pelvis as high as they comfortably could, putting equal weight on both lower extremities. The standardized initial position ensured that at the elevated position, the angle of the knees would reach approximately 75° to 80° of flexion. This final position was confirmed only by observation to avoid interference with the fluency of the performance.
The second session of the study took place 1 to 2 days later. During this 60-minute session, the symmetrical weight bearing was reinforced by feedback from the evaluator, who monitored the weight bearing by using 2 bathroom scales placed under the participants' feet. Once participants were able to assume symmetrical weight bearing on a regular vinyl-covered mat, they were then evaluated for their ability to perform the same exercise on a smoother surface of 2 layers of plastic sheets under their feet. Only half of the participants who could perform bridging independently on the regular mat were able to prevent limb sliding on a smoother surface. At this stage, the participants were divided into 2 groups according to their ability to perform bridging on the slippery surface.
Figure 1 illustrates a biomechanical model of the bridging task. The reaction force vector from the supporting surface is composed of vertical (normal) and tangential (frictional) components. The reduced magnitude of friction vector on the smoother surface changes the inclination of the resultant reaction vector (R), shifting it away from the knee axis, and increases the extension moment arm around the joint. As a result, the knee flexors have to work harder to resist a larger extension moment. This biomechanical model was supported by an electromyography study of one of the present investigators (M. Z. Kozol, unpublished data, November 2007) on 4 healthy adults. The results showed that in comparison with bridging on a regular vinyl cover, bridging on a smooth plastic sheet increased the muscular activity of the knee flexors to a larger extent than that of the extensors.
Following the screening of the bridging performance on both surfaces, the spasticity of the knee extensors, the motor control level of the affected lower extremity, and the moments of the knee muscles were evaluated by the MAS, the FM-LE, and the HHD, respectively. The spasticity evaluation was conducted in a supine position. The therapist lifted the affected limb, holding one hand beneath the heel and the other hand behind the knee. Participants were instructed to remain relaxed while the knee and hip were flexed passively 4 times at a rate of about 1 Hz. The MAS rating was determined after 4 trials by the overall impression of the resistance to flexion.
The maximum moments of the knee muscles were recorded with subjects seated in a short sitting position on the edge of a treatment table with the feet flat on the floor. The participants' trunk and shoulders were stabilized from behind by an assisting physical therapist.28,29 To evaluate the strength of the knee extensors, the participants were asked to extend the knee to slightly less than full extension, to avoid excessive hamstrings resistance and knee locking.30 In this position, the shank's inclination was approximately 30° below the horizontal plane, as verified by a goniometer. The pad of the dynamometer was positioned on the anterior aspect of the shank, slightly proximal to the malleoli. The distance from the center of the pad's area of contact with the skin to the knee joint line was measured (lever length). Participants were instructed to maintain the knee position by resisting a gradually increasing flexing force, applied perpendicular to the shank's long axis with the goal of overcoming the knee's extensors (ie, the “break” technique).
The evaluation of the flexors' moment was performed in a similar sitting position, but with the knee at 90° of flexion. To prevent friction of the sole against the floor, the plinth was elevated slightly and the contralateral foot was supported on a low stool in a position of about 90° flexion of the hip and the knee. Three testing trials were recorded for each muscle group, beginning arbitrarily with the flexors or the extensors of the paretic limb. There was a brief rest period of 1 to 2 minutes between the trials within each muscle group and of 3 to 5 minutes between the muscle groups. For each group, the mean of the peak force of the 3 trials was used for the analysis.
The moments were calculated by multiplying the mean peak force by the lever length, as described by Ada et al.31 The calculated moment of the extensors was corrected by adding the gravitational moment of the inclined shank and foot, based on Dempster's anthropometric data.32 The peak moments, motor control level, and muscular tone of both groups were compared, using independent t tests for the moments and independent Mann-Whitney U nonparametric tests for the FM and MAS scores.
Comparisons were followed by performing a point biserial correlation33 between the dichotomous parameter (able/not able) of the bridging performance on a smooth surface and the other parameters. The predictability of sliding by the other parameters and their interrelationships were determined by binary logistic regression and Pearson product moment correlation, respectively. All analyses were carried out with SPSS, version 10.0 (SPSS, Inc, Chicago, Illinois). A P value of less than 0.05 was considered statistically significant.
DATA ANALYSIS AND RESULTS
Of the 27 participants who could perform the bridging exercise on a regular mattress cover, only 13 (11 men and 2 women) could also perform it on the low-friction surface. A comparison of both groups' parameters revealed that the FM-LE motor control scores and the knee flexion moments of the “sliding” group were significantly lower than those of the “nonsliding” group (P = .001), while there were no between-group differences in the demographic data, MAS scores, or other moments (Table 3). The point biserial correlation between the bridging performance on a smooth surface and the motor parameters (Figure 2) also illustrated that only the FM-LE motor control score (r = 0.74, P < .001) and the maximal flexion moment of the paretic knee (r = 0.61, P < .001) were correlated with the bridging performance on the smooth surface.
The predictability of bridging without sliding was determined by binary logistic regression using the independent variables of the FM-LE score, MAS score, and all knee moments (ordinal parameters were categorized). This analysis revealed that although the global model was statistically significant (P < .001), none of the individual independent variables were significant. This finding suggests strong multicollinearity that destabilizes the model and may interfere with the ability to get good estimate of the distinct effect of the independent (explanatory) variables.34 The multicollinearity was confirmed by collinearity diagnostics (linear regression, SPSS) and managed by dropping the FM-LE variable from the logistic regression equation.34 After removal of the FM-LE variable, the moment of the paretic knee flexors was found to be statistically significant (P = .023) (Table 4) and the global model correctly predicted 88.9% of the observations. The FM-LE parameter was then reentered into the model and the moment of the paretic flexors dropped. This replacement did not alter the predictability of events, as evidenced by identical classification table of both models (Table 5), but the FM-LE score became the only significant individual variable (P = .009) (Table 6).
The correlation matrix of all the parameters demonstrated that the extension of the paretic knee had the highest intermoment correlation with the extension of the nonparetic knee (r = 0.85, P < .001) and was also correlated with the contralateral moment of the flexors (r = 0.64, P < .001) (Table 7). The moment of the paretic flexors was correlated with the contralateral moment of the extensors (r = 0.47, P = .013), but not with that of the flexors. Moderate correlations were also found between the agonist and antagonist moments of the paretic (r = 0.50, P = .007) and nonparetic (r = 0.76, P < .001) knees. The MAS and the FM-LE scores were not correlated with any of the other parameters except the significant correlation between the FM-LE score and the moment of the paretic flexors (r = 0.60, P = .001).
A 2-way repeated-measures analysis of variance was conducted to determine the differences between the involvements of both sides and both muscle groups (Figure 3). As expected,35 the significant main effect (P < .001) revealed that the moments of the paretic knee were lower than those of the contralateral one. In addition, there was a significant side-by-moment interaction (P = .04), suggesting that the paretic flexors' moment was more affected than that of the extensors.
Participants who could perform the bridging exercise on a low-friction surface had higher motor control scores and larger maximum flexor moments of the affected knee than their peers who could not perform the exercise (Table 3). This finding was supported by the significant correlation of the FM-LE score and the paretic flexors' moment with the bridging performance (Figure 2). Moreover, the binary logistic regression model to predict limb-sliding control demonstrated that the odds ratio of the flexor moment, controlled by the other independent parameters, was small but still significant (P = .023) and larger than the null value of 1.0, by about 120% (Table 4). Thus, an increase in the moment by a single unit (1 N m) increased the odds of not sliding during bridging by about 20%.36 The FM-LE was not initially included in the logistic regression model to avoid collinearity, but after dropping the paretic flexor moment from the model and reentering of the FM-LE, it became the best predictor of successful performance (Table 6). The adjusted odds ratio for predicting bridging performance by the FM-LE score was 2.781 (P = .009), meaning that an increase in the FM-LE score by a single point almost tripled the odds of bridging without limb sliding. Apparently, the FM-LE score and the paretic flexor moment are both important predictors of limb sliding during bridging. Although the FM-LE score appears to be a more dominant determinant of successful bridging, considering that active knee flexion is an integral component of the FM-LE test,19 and that both parameters are measured by different units, it is difficult to determine their relative practical impact.
In contrast to the traditional clinical model,8,9 the MAS score was not correlated with any of the measured parameters (−0.24 < r < 0.17) (Table 7), suggesting that spasticity of the antagonists is not a determinant of agonist strength, motor control level, or bridging performance, in agreement with previous investigations.2,7 Interestingly, although the participants were not selected by their level of spasticity, the range of their MAS scores was extremely narrow (between 0 and 2), in line with the low sensitivity of the test in the subacute stage.23,35 The limited sensitivity of the Ashworth test might have also diminished the association of the MAS scores with the other parameters (Table 7). However, it is noteworthy that the MAS determines spasticity solely by the resistance to passive movement. Expanding the evaluation to include the abnormal posture and motor behavior, often clinically associated with spasticity,8,9 may alter the relationships.
The significant correlations between the agonist and antagonist moments of the paretic (r = 0.5, P = .007) and nonparetic (r = 0.76, P < .001) knees (Table 7) probably related to their connectivity within the central nervous system.18,37–39 In other words, the association of various parameters may reflect interacting neural circuits. However, the neural connectivity of the knee flexors appeared to be affected more than that of the extensors, as evidenced by the lower correlation of the flexors' moment with contralateral moments of agonists and antagonists. This disassociation from contralateral moments could also affect movement and posture control.39,40
The significant main effect of the analysis of variance (Figure 3) confirmed previous reports35,41 that the muscles of the paretic knee are significantly weaker than the contralateral ones (P < .001). However, considering that the muscle strength of the nonparetic limb was also reduced, the actual decline in the paretic muscles strength might have been even more severe.42 There was also a significant side-by-moment interaction (P = .04), suggesting that the paretic flexors' moment was affected more than that of the extensors.
These findings are consistent with previous investigations using isokinetic dynamometers35,41 but contradict the findings of Andrews and Bohannon.43 Using HHD, those investigators found that although initially the decline in the strength of the paretic knee flexors was more prominent than that in the paretic knee extensors, the difference leveled out at 1 month poststroke. The apparent contradiction between the studies may be explained by the difference in measuring techniques, as Andrews and Bohannon43 used the “make” technique, while the present study used the “break” technique.44 Although reliability of both techniques is supported, the magnitude of the measurements may differ slightly.44,45
In addition, Andrews and Bohannon43 measured the extension moment of the knees in a short sitting position of 90° flexion of the hip and knee, whereas in the present study the knee was almost completely extended. The extended position, similar to the position used in standard manual muscle testing,30 was employed to minimize inhibition of the extensors by the flexed position of the entire limb (flexion synergy).19 The difference between the knee angle of both studies may also affect the length-related maximum strength31 and the moment arm32 of the extensors.
The pronounced impairment of the flexors and their sole association with the level of motor control (r = 0.60, P = .001) (Table 7) may reflect a stronger linkage of the corticospinal tract with the flexors than with the extensors. This finding is in line with studies that found a more pronounced effect of transcranial magnetic stimulation of the motor cortex on the activity of the tibialis anterior (flexor) than of the soleus (extensor),46 as well as by the preservation of flexor strength after the microgravity of space flight.47 Those researchers attributed the more significant decline in extensors strength after space flight to the higher impact of diminished proprioceptive afferent flow on extensors' activation and to the stronger linkage of the descending corticospinal tract with the spinal motoneurons of the flexors as compared with the extensors. Further support for the different interaction of higher neural centers with the flexors and the extensors has been provided by perturbation gait studies, demonstrating that the skill of the swinging subtask (dominated by the flexors) was more effectively transferred to the contralateral lower extremity than the skill of the weight-bearing subtask (dominated by the extensors).48,49 Van Hedel et al48 attributed the superior transfer of motor learning between contralateral flexors to their direct modulation by a higher level of the neural hierarchy. Alternatively, beyond the neural connectivity, the relative robustness of the extensors to central neural damage may be a result of their facilitated recovery by daily upright activities and by weight-bearing stimulations, challenging the extensors.17,47
Clinicians should be aware of the predominant post-CVA decline in the muscle strength of the paretic knee flexors. Considering that common upright daily activities activate the extensors to a greater extent than the flexors, it is recommended that evaluation and training protocols should include activities that specifically challenge the knee flexors, such as controlled bridging on smooth surfaces. Understanding of the association between the friction of the supporting surface and the muscular activity during bridging may assist to reveal any misleading masking of impairments by frictional forces and to focus evaluation and training on the actual disorders.
The impact of the reduced flexors' moment on limb sliding lends support to previous claims that physical therapy intervention should concentrate more on improving motor control and muscle strength than on reducing spasticity of the antagonists.7 Moreover, the disrupted associations between contralateral moments post-CVA suggest that evaluation and intervention should include simultaneous bilateral activation of agonists and antagonists in line with their normal functional interaction.39,40 This recommendation requires further experimental substantiation.
The study was conducted on a small sample of convenience with confined inclusion criteria to lesion localization and motor status, thus limiting generalization to similar populations. It is noteworthy that since the inclusion criteria did not relate to spasticity and because the tone of most participants was normal or slightly elevated, the findings cannot be generalized to individuals with marked increase in tone. Additional research taking spasticity into account may further increase our ability to generalize the results.
1. Zorowitz RD, Gross E, Polinski DM. The stroke survivor. Disabil Rehabil
2. Patten C, Lexell J, Brown HE. Weakness and strength training in persons with poststroke hemiplegia: rationale, method, and efficacy. J Rehabil Res Dev
3. Knutsson E, Richards C. Different types of disturbed motor control in gait of hemiparetic patients. Brain
4. 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
5. Kim MC, Eng JJ. The relationship of lower-extremity muscle torque to locomotor performance in people with stroke. Phys Ther
6. Engardt M, Knutsson E, Jonsson M, Sternhag M. Dynamic muscle strength training in stroke patients: effects on knee extension torque, electromyographic activity and motor function. Arch Phys Med Rehabil
7. Gowland C, deBruin H, Basmajian J, Plews N, Burcea I. Agonist and antagonist activity during voluntary upper limb movement in patients with stroke. Phys Ther
8. Davies PM. Steps to Follow
. Berlin, Germany: Springer-Verlag; 1985.
9. Bobath B. Adult Hemiplegia
: Evaluation and Treatment
. 3rd ed. London, England: Butterworth-Heinemann; 1991.
10. Carr JH, Shepherrd RB. A Motor Relearning Programme for Stroke
. London, England: Heinemann Physiotherapy; 1987.
11. Stewart P, Wharton GW. Bridging: an effective and practical method of preventive skin care for the immobilized person. South Med J.
12. Tsuji T, Liu M, Hase K, et al. Physical fitness in persons with hemiparetic stroke: its structure and longitudinal changes during an inpatient rehabilitation programme. Clin Rehabil
13. Gregson JM, Leathley M, Moore AP, Sharma AK, Smith TL, Watkins CL. Reliability of the Tone Assessment Scale and the Modified Ashworth Scale as clinical tools for assessing poststroke spasticity. Arch Phys Med Rehabil
14. 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
15. Bohannon RW. Walking after stroke: comfortable versus maximum safe speed. Int J Rehabil Res
16. Mazzà C, Stanhope SJ, Taviani A, Cappozzo A. Biomechanic modeling of sit-to-stand to upright posture for mobility assessment of persons with chronic stroke. Arch Phys Med Rehabil
17. Carr JH, Shepherd RB. Neurological Rehabilitation: Optimizing Motor Performance
. Oxford, England: Butterworth Heinemann; 1998.
18. Bohannon RW, Andrews AW. Relationships between impairments in strength of limb muscle actions following stroke. Percept Mot Skills
19. Fugl-Meyer AR, Jääskö L, Leyman I, Olsson S, Steglind S. A method for evaluation of physical performance. Scand J Rehabil Med
20. Sanford J, Moreland J, Swanson LR, Stratford P, Gowland C. Reliability of the Fugl-Meyer assessment for testing motor performance in patients following stroke. Phys Ther
21. Rödén-Jüllig A, Britton M, Gustafsson C, Fugl-Meyer A. Validation of four scales for the acute stage of stroke. J Intern Med
22. Bohannon RW, Smith MB. Inter-rater reliability of a Modified Ashworth Scale of muscle spasticity. Phys Ther
23. Blackburn M, van Vliet P, Mockett SP. Reliability of measurements obtained with the Modified Ashworth Scale in the lower extremities of people with stroke. Phys Ther
24. Li RC, Jasiewicz JM, Middleton J, et al. The development, validity, and reliability of a manual muscle testing device with integrated limb position sensors. Arch Phys Med Rehabil
25. Bohannon RW. Test-retest reliability of hand-held dynamometry during a single session of strength assessment. Phys Ther.
26. Bohannon RW, Andrews AW. Inter-rater reliability of hand-held dynamometry. Phys Ther.
27. Riddle DL, Finucane SD, Rothstein JM, Walker ML. Intrasession and intersession reliability of hand-held dynamometer measurements taken on brain-damaged patients. Phys Ther
28. Andrews AW, Thomas MW, Bohannon RW. Normative values for isometric muscle force measurements obtained with hand-held dynamometers. Phys Ther.
29. Bohannon RW. Reference values for extremity muscle strength obtained by hand-held dynamometry from adults aged 20 to 79 years. Arch Phys Med Rehabil
30. Hislop HJ, Montgomery J. Daniels and Worthingham's Muscle Testing
: Techniques of Manual Examination
. 8th ed. Philadelphia, PA: WB Saunders; 1995:224–225.
31. Ada L, Canning C, Dwyer T. Effect of muscle length on strength and dexterity after stroke. Clin Rehabil
32. LeVeau BF. Williams & Lissner's Biomechanics of Human Motion
. Philadelphia, PA: WB Saunders; 1992:297–307.
34. Allison PD. Logistic Regression Using the SAS System—Theory and Application
. Cary, NC: SAS Institute; 1999:48–51.
35. Newham DJ, Hsiao SF. Knee muscle isometric strength, voluntary activation and antagonist co-contraction in first six months after stroke. Disabil Rehabil
36. Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice
. 2nd ed. London, England: Prentice Hall; 2000:597–603.
37. Gandevia SC. Assessment of corticofugal output: strength testing and transcranial stimulation of the motor cortex. In: Gandevia SC, Burke D, Anthony M, eds. Science and Practice in Clinical Neurology
. Cambridge, England: Cambridge University Press; 1993:78.
38. Capaday C, Devanne H, Bertrand L, Lavoie BA. Intracortical connections between motor cortical zones controlling antagonistic muscles in the cat: a combined anatomical and physiological study. Exp Brain Res
39. Gauthier J, Boubonnais D, Filiatrault J, Gravel D, Arsenault AB. Characterization of contralateral torques during static hip efforts in healthy subjects and subjects with hemiparesis. Brain
40. Brunnstrom S. Movement Therapy in Hemiplegia: A Neurophysiological Approach
. Cambridge, England: Harper & Row; 1970:111–128.
41. Sharp SA, Brouwer B J. Isokinetic strength training of the hemiparetic knee: effects on function and spasticity. Arch Phys Med Rehabil
42. Sunnerhagen KS, Svantesson U, Lönn L, Krotkiewski M, Grirmby G. Upper motor neuron lesions: their effect on muscle performance and appearance in stroke patients with minor motor impairment. Arch Phys Med Rehabil
43. Andrews AW, Bohannon RW. Distribution of muscle strength impairments following stroke. Clin Rehabil
44. Bohannon RW. Make tests and break tests of elbow flexor muscle strength. Phys Ther
45. Burns SP, Breuninger A, Kaplan C, Marin H. Hand-held dynamometry in persons with tetraplegia: comparison of make- versus break-testing techniques. Am J Phys Med Rehabil
46. Capaday C, Lavoie BA, Barbeau H, Schneider C, Bonnard M. Studies on the corticospinal control of human walking. I. Responses to focal transcranial magnetic stimulation of the motor cortex. J Neurophysiol
47. Recktenwald MR, Hodgson JA, Roy RR, et al. Effects of spaceflight on rhesus quadrupedal locomotion after return to 1 G. J Neurophysiol
48. Van Hedel HJ, Biedermann M, Erni T, Dietz V. Obstacle avoidance during human walking: transfer of motor skill from one leg to the other. J Physiol
49. Prokop T, Berger W, Zijlstra W, Dietz V. Adaptational and learning processes during human split-belt locomotion: interaction between central mechanisms and afferent input. Exp Brain Res