Trunk control is an important predictor of poststroke functional locomotor recovery and has been the main focus of physical therapy interventions.1,2 However, the trunk movement quality, defined as thoracic and pelvic segmental range of motion (ROM) and the coordination between segmental movements during functional activities, is not routinely assessed in the clinic. Most clinical measures focus on trunk muscle strength, such as the Trunk Control Test,3 and individuals with stroke are often reported to have weaker trunk muscles than age-matched subjects.3–6
Trunk kinematics are related to walking speed; in young subjects without disability, pelvic transverse (yaw) plane ROM increased during treadmill walking with walking speeds of 0.7 to 1.0 m/s and decreased with walking speeds of 1.1 to 1.3 m/s.7 Greater pelvic ROM has also been linked to longer stride lengths, while relationships between thoracic ROM and walking speed remain unclear.8,9 Intersegmental phase relationships (ie, continuous relative phase [CRP] or cross-correlations) between movements of the thorax and pelvis during gait are also affected by walking speed, where CRPs near 180° indicate an in-phase coordination and those close to 0° denote an antiphase coordination.7–11 At slower walking speeds, transverse plane thoracic and pelvic movements are close to in-phase (25°-44°) in subjects without disability, but faster walking speeds beyond 0.83 m/s are associated with more antiphase coordination (110°-126°), indicating a bifurcation point in the stability of the coordination pattern.8,9,12 In the coronal plane (roll), speed has a similar effect on CRP, while in the sagittal plane (pitch), the effect is opposite such that the faster the subjects walk, the more in-phase the coordination pattern.13
Intersegmental trunk coordination during gait has also been characterized using CRP in adults with musculoskeletal and neurological deficits, such as low back pain14–16 and Parkinson disease.17 In both groups, transverse CRP values are lower than those in subjects without disability walking at similar speeds. In subjects with chronic stroke and in those without disability, transverse CRP as well as thoracic and pelvic ROM values are related to walking speeds between 0.25 and 1.5 m/s.7 Faster walking is associated with higher CRP (more antiphase) and trunk ROM values in both the groups. Overall, mean values of CRP, pelvic ROM, and total ROM in the transverse plane are similar between groups, but participants with stroke use more thoracic ROM. Movements in the coronal plane have not been investigated further.
In spite of the emphasis of physical therapy treatment on improving trunk control in patients with stroke, the relationship between intersegmental coordination and functional gait deficits in this group has not been established. A better understanding of this functional relationship may provide support, or suggest new directions, for current clinical practice in stroke rehabilitation. We investigated this relationship, using CRP in individuals with chronic stroke. We hypothesized that individuals with stroke would have deficits in intersegmental coordination between thoracic and pelvic movements during gait, reflected in lower CRP values, compared with subjects without disability. We further hypothesized that individuals with stroke who have better intersegmental trunk coordination would perform better on functional gait and balance measures. Preliminary results have appeared in the abstract form.18
Eleven subjects with stroke and 11 age-matched subjects without disability participated in this study (Table 1). The sample size calculation was based on trunk-rotation data taken from subjects without disability and subjects with stroke walking on a treadmill.7 A sample size of 10 subjects per group was required to have an effect size of 0.7 at an α level of 0.05 and a power of 0.95. Individuals with stroke were included if they (1) had a single unilateral ischemic stroke more than 6 months back; (2) were aged 40 to 75 years; (3) were able to walk independently (ie, without walking aids) on a treadmill; (4) had impaired postural control (≤6 of 7 on the Chedoke-McMaster [CM] Impairment Inventory)19; (5) had impaired walking speed (comfortable speed ≤ 0.95 m/s); (6) had residual arm movement (the CM arm scale ≥ 3 of 7); and (7) were able to understand instructions. Exclusion criteria were the presence of (1) marked visuospatial neglect (Bell Test < 26 of 35)20 or visual-field deficits (medical chart); (2) significant deficits in upper- or lower-limb proprioception (<14 of 16 Fugl-Meyer Scale)21; or (3) orthopedic or rheumatic conditions interfering with walking. Subjects were recruited from discharge lists of rehabilitation centers within the Centre for Interdisciplinary Research in Rehabilitation (CRIR) network Greater Montreal (Quebec, Canada). All subjects signed informed consent forms approved by the Research Ethics Committee of the Establishments of CRIR before participating.
To prevent falling, subjects wore a ceiling-mounted safety harness that moved over the length of a self-paced motorized treadmill. The light-weight harness did not restrict body movements or support any amount of the body weight during the locomotor task. Treadmill speed was determined by the length of a cord attached to the harness and connected to a potentiometer located 1.88 m behind the treadmill (Figure 1). Treadmill speed changed in relation to the anteroposterior distance of the subject from the potentiometer with the intent being to maintain the subjects' position in the middle of the length of the treadmill belt. Safety switches on the treadmill and the control box could stop the treadmill if necessary.
Subjects walked at their comfortable speed and approximately 20% faster (for subjects with stroke) or 20% slower (for subjects without disability). To minimize variability in walking speed, subjects followed metronome pacing matched to their walking cadence at each speed. After becoming familiar with self-paced treadmill walking, subjects performed four 30-second walking trials (baseline), which were used to calculate their mean comfortable and faster/slower (equivalent) walking speeds. Participants walked naturally in the middle of the belt while looking at a fixed point at eye height at the end of the walkway. Rest periods were allowed between trials, and heart rate was monitored so that it did not exceed 70% of maximum (220 − age) in each trial.
Kinematic data were collected using a 12-camera motion capture system (Vicon-512, Los Angeles, California, USA) with reflective markers at a sampling rate of 120 Hz for four 30-second walking trials at each speed. Thoracic and pelvic rotations in 3 planes (sagittal, coronal, and transverse) were recorded using rigid body clusters of 3 and 4 noncoaxial markers, respectively. Markers were placed on the right and left acromion processes and midsternum for recording the thoracic movement and on the right and left anterior and posterior superior iliac spines for recording the pelvic movement. Markers were also placed on shoe toes and heels for determining gait temporal and distance factors, as well as on the third fingertip of each hand for arm-swinging parameters.
For all subjects, the ability to perform functional locomotor activities was measured with the Functional Gait Assessment (FGA),22 and additional aspects of balance were measured using the BesTest.23 The CM Stroke Assessment was used to measure impairment in subjects with stroke. Clinical evaluations were done by experienced physiotherapists blinded to study goals.
Measurement of Impairment
The BesTest is a clinical balance assessment of 36 items grouped into 6 sections. This test offers a comprehensive assessment of different aspects of balance. Items are evaluated by subscales with maximal scores ranging from 15 to 21 points, where maximal values in each section indicate no impairment or disability. Sections evaluate (1) biomechanical constraints (eg, base of support), (2) stability limits during standing (eg, functional reach), (3) anticipatory postural adjustments, (4) reactive postural responses, (5) sensory orientation (eg, Foam and Dome Test),24 and (6) gait stability, an activity measure similar to FGA that includes the Timed Up and Go Test.25 The BesTest has excellent interrater reliability (ICC = 0.91) and, although not specifically validated in people with stroke, is valid for people with balance disorders.23 The CM assesses physical impairment and disabilities that may occur after stroke. It includes 6 dimensions (shoulder pain, postural control, arm, hand, leg, and foot) on 7-point scales, where a maximum score of 7 indicates no impairment or disability. All dimensions have excellent interrater reliability (ICC = 0.93–0.98).19
Measurement of Function
Gait performance was measured using FGA, in which each of the 10 items is scored on a 4-point scale (0–3), where 3 indicates normal function. Functional Gait Assessment includes 7 of 8 items from the Dynamic Gait Index and 3 additional items assessing gait function (“gait with narrow base of support,” “ambulating backwards,” and “gait with eyes closed”). Concurrent validity was shown for people with stroke by comparing the results of the FGA with the Berg Balance Scale, the 10-m timed walking test, the Timed Up and Go Test, and the Activities-Specific Balance Confidence rating score with r-values ranging from 0.63 to 0.83. Functional Gait Assessment was chosen over Dynamic Gait Index because of its superior standardization of instructions and the smaller probability of a ceiling effect for high-functioning walkers due to the 3 additional items.26
Gait distance and temporal factors, trunk-segment kinematics, and arm-swing distance were analyzed from a sequence of 10 consecutive strides (gait cycles) selected from each of the 4 trials for each walking speed. Selection was based on the observation of consistent amplitudes and frequencies of bilateral heel movements. Factors identified were step length, step width, stride frequency (cadence), and walking speed. Step length was defined as the anteroposterior distance. Step width was defined as the mediolateral distance from the midline to the left- and right-heel markers during double support in stance. Stride frequency/cadence was calculated as the number of right and left steps divided by time. Walking speed was computed as the inverse of the treadmill speed.
For each gait cycle, thoracic and pelvic angular ROMs were measured in 2 planes (coronal and transverse). Rotations in the sagittal plane were not considered since they are influenced by head position and gaze direction, which were not rigidly controlled in this study.27 Range of motion was defined as the absolute angular difference in degrees from the maximal to the minimal rotation within each cycle (heel-off to heel-off on the same leg). Outcomes were measured separately for the left and right gait cycles for subjects without disability and for the paretic and nonparetic gait cycles in the stroke group. From rigid bodies placed on the thorax and the pelvis, segment reference frames were defined and transformed into a global coordinate system (x, y, and z). In this reference frame, rotation around the y and z axes corresponded to the coronal and transverse planes, respectively. Arm-swing symmetry was computed as the ratio of the paretic/left to the nonparetic/right arm-swing amplitude. Amplitudes were defined as the arm-swing length (y, mm) in each cycle from the maximal backward to the maximal forward displacement of each fingertip, regardless of the presence of rotational bias toward the paretic side.10
The primary outcome measure was the CRP of the thoracic (Th) and pelvic (Pl) trunk segments in 2 planes (coronal and transverse). Continuous relative phase describes the instantaneous differences in both position and velocity between the 2 body segments.28 For each time point, the inverse tangent of the ratio between velocity and position was obtained for each segment and the pelvic phase angle was subtracted from that of the thorax, which can be represented by the following formula:
A phase difference of 180° or 0° represents perfect antiphase or in-phase movement, respectively. For all measures, mean (SD) values of all gait cycles over the 4 trials were determined. Variability of CRP values for each subject was expressed as the coefficient of variability (CV) ([SD/mean] × 100).
Comparisons between spatial and temporal variables were done with 2×2×2 analyses of variance with factors group (nondisabled and stroke), speed (comfortable and equivalent), and side (right/nonparetic and left/paretic). For CRP values, a 2×2 analysis of variance was used (factors: group—nondisabled and stroke; speed—comfortable and equivalent). Minimal significance levels were P < 0.05. Correlations were evaluated with the Spearman tests.
All subjects walked on the treadmill at 2 different speeds. Subjects with stroke walked approximately 36% slower than subjects without disability, at comfortable speed (nondisabled: 1.22 ± 0.21 m/s; stroke: 0.78 ± 0.20 m/s; P < 0.001). When asked to walk faster, walking speed in subjects with stroke remained more than 20% slower than the comfortable speed of subjects without disability (Table 1). One subject with stroke could not walk faster, and her data were included in the comfortable speed group only.
Mean absolute thoracic and pelvic ROM values did not differ between right/left or paretic/nonparetic gait cycles; therefore, data from only 1 side were considered (right/nonparetic). Subjects with stroke used approximately 15% more transverse thoracic ROM than subjects without disability (F 1,78 = 6.436, P < 0.02; Table 2), illustrated for a single trial in one subject from each group in Figures 2A and C. Only in the stroke group, thoracic ROM was larger than pelvic ROM (thorax: 12.2° ± 3.9°; range: 6.7°-13.9°; pelvis: 7.9° ± 2.7°; range: 4.9°-18.5°; P < 0.001; Figures 3A and B). There were no differences in ROM in the coronal plane for either group.
Overall, subjects with stroke had a higher cadence (∼24%, F 1,80 = 12.347, P < 0.001) and used wider (∼22%, F 1,80 = 4.256, P < 0.03) and shorter steps (∼20%, F 1,80 = 37.816, P < 0.001) than the subjects without disability (Table 2). The wider step width was more marked for the nonparetic leg in the stroke group than in the control group without disabilities.
Walking speed affected transverse pelvic ROM (F 1,78 = 4.168, P < 0.05; Figures 3A and B) and cadence (F 1,78 = 4.987, P < 0.03) in the stroke group. While these were similar in both groups at comfortable speed, pelvic rotation was lower (P < 0.05) and cadence was higher (P < 0.001) in the stroke group at equivalent speeds (Table 2).
In subjects without disability, arm-swing amplitude of the nondominant arm was 92% of that of the dominant arm at comfortable speed but was equal at the slower (equivalent) speed. However, in the stroke group, the ratio was 56:59 for the paretic arm compared with the nonparetic arm at both speeds and was significantly less than that in the group without disabilities (F 1,78 = 8.759, P < 0.005). The difference in arm-swing amplitude between the arms was significant only in the stroke group (F 1,78 = 4.514, P < 0.05; Table 2).
Continuous Relative Phase
Continuous relative phase within a single cycle did not vary more than 15% in the most severely affected subject with stroke, and lead/lag relationships between the segments were relatively stable across trials. Across trials, transverse plane CRP values were lower (more in-phase) in the group with stroke than in the group without disabilities at comfortable speed (F 1,39 = 4.082, P < 0.05; Figure 3C). Continuous relative phase values were higher (more antiphase) in participants with stroke who had lower gait impairment (r = 0.63, P < 0.05; Table 3). This is illustrated in Figure 3D, in which data are arranged according to the subject's FGA scores from high (left) to low gait impairment (right). Continuous relative phase variability was similar between groups with CVs ranging from 6.0% to 57.7% (equivalent) and 8.9% to 24.5% (comfortable) in subjects without disability and from 6.1% to 46.2% (equivalent) and 3.6% to 42.2% (comfortable) in participants with stroke. For movements in the coronal plane, CRP was closer to antiphase in both groups (not shown in Figure 3), with no difference between groups.
Correlations Between Kinematic Data and Clinical Scores
Since clinical evaluations were performed at comfortable speed, only correlations between kinematic data recorded at this speed and clinical scores were evaluated. Total BesTest scores ranged from 85 to 107 of 108 points in subjects without disability and ranged from 75 to 104 points in subjects with stroke. In subjects without disability and subjects with stroke, the FGA scores ranged from 26 to 30 and 17 to 28 of 30 points, respectively.
In subjects without disability, there were markedly fewer significant correlations between kinematic and clinical scores than in the stroke group (Table 3), and where significant correlations were observed, these occurred for measures in the transverse plane. Walking speed was positively correlated with CRP for each speed condition. Continuous relative phase was also moderately correlated with the “Biomechanical Constraints” section of the BesTest. Pelvic ROM and thoracic ROM were negatively correlated with the FGA and the “Sensory Orientation” section of the BesTest, respectively.
Although the range in walking speed was similar between groups, in contrast to subjects without disability, walking speed was not correlated with CRP in subjects with stroke. However, in subjects with stroke, there were substantially more correlations between transverse CRP and clinical measures (BesTest 1, 3, 4, and 6: r = 0.53–0.68; FGA: r = 0.63) than in subjects without disability. Correlation values indicated that a more antiphase coordination was related to better locomotor function. In addition, greater segmental ROM was associated with poorer functional performance. Pelvic ROM was negatively correlated with the “Anticipatory Postural Adjustment” section of the BesTest (BesTest 3: r = −0.56) and the FGA (r = −0.54) as well as with the CM arm (r = −0.64) and foot (r = −0.67) scores. Thoracic ROM was negatively correlated with the BesTest subscales 1, 3, 4, and 6 (r = −0.58 to −0.83), the FGA (r = −0.54), and 3 CM subscales (arm, leg, and foot: r = −0.71 to −0.78).
Comfortable walking speed was slower in subjects with stroke than in subjects without disability,7 and there were differences in kinematics at that speed between groups despite the similarity in cadence; subjects with stroke took shorter and wider steps and tended to have a more in-phase compared with antiphase thorax-pelvis coordination. Also, in contrast to subjects without disability, subjects with stroke used more thoracic than pelvic ROM at both speeds. Consistent with a previous study, differences in gait kinematics were also present when walking speeds and cadences were equivalent between groups.29 Although CRP values at equivalent speeds were similar between groups, the usual increase in the CRP values with an increase in the walking speed9 was not found in the stroke group, and CRP values were more often associated with functional gait deficits.
Kinematic Measures Correlate With Clinical Function in Subjects With Stroke
All subjects with stroke were high functioning, could walk independently, and participate in community activities. Nevertheless, they performed significantly worse than age-matched subjects without disability in some clinical gait and balance tests. Subjects with stroke who had a more antiphase pattern of trunk intersegmental coordination (higher CRP) when walking at their comfortable pace had better balance and gait functional performance scores.
While this is the first demonstration of a relationship between the CRP and the functional gait performance, the relationship between walking speed and functional performance has been well-documented.30–32 For example, faster walking ability has been correlated with the accomplishment of more walking-related functional activities.33 Few studies have identified the underlying movement deficits or biophysical mechanisms related to deficits in walking speed; however, Daly et al34 found strong correlations between disruptions in hip and knee interjoint coordination and deficits in walking speed.
Faster walking has been associated with better antiphase transverse thorax-pelvis coordination in both subjects without disability and subjects with stroke.7–10 This relationship was also observed in our group of subjects without disability for a restricted range of walking speeds (0.98–1.22 m/s). In contrast, most subjects with stroke did not increase their CRP when challenged to walk approximately 20% faster than their comfortable speed (from 0.78 to 0.97 m/s; see Figure 3B). This suggests that increases in walking speed were achieved by using compensatory movement patterns instead of by “normalizing” movement patterns (cf Kim and Eng35). The possibility that a limitation in transverse intersegmental coordination may be a factor limiting higher walking speed in some persons with stroke merits further investigation.
Increase in Relative Thoracic ROM May Be a Compensatory Mechanism in Stroke
Thorax and pelvis ROMs were negatively correlated with clinical measures of locomotor impairment (the FGA, the BesTest, and the CM), suggesting that larger thoracic and pelvic movements were associated with greater functional gait impairment. The increased thoracic ROM may be similar to compensatory mechanisms previously described in subjects with stroke, such as the recruitment of excessive trunk movements to assist reaching, pointing, and grasping from the sitting position36–38 and to assist rhythmical bilateral arm swinging while standing. For example, during in-phase arm swinging in standing, Ustinova et al39 found a negative relationship between displacement of the thorax and arm in subjects without disability, such that anterior arm swinging was inversely correlated with posterior displacement of the ipsilateral shoulder. However, this correlation was positive in subjects with stroke, indicating that forward thoracic movement was used to assist arm swinging. The greater thoracic movement also resulted in a forward shift of the center of mass. Participants with stroke in our study may have used a similar mechanism (greater thoracic rotation) to compensate for the deficit in the paretic arm swing amplitude (Table 2) and to facilitate anterior shifting of the center of mass to assist forward gait progression. Thus, our finding that changes in spatial movement characteristics (thoracic ROM) persist regardless of walking speed and are related to clinical locomotor function suggests that the nature of trunk compensation is spatial rather than temporal since the temporal coordination is, to a large extent, dictated by walking speed as found in this study and in a previous study.9 Nevertheless, a role for temporal movement factors in compensation cannot be ruled out since subjects with stroke with larger (more antiphase) transverse CRP values also performed better on functional locomotor scales (Table 3).
Achieving better gait performance is one goal of stroke rehabilitation. This study showed that better functional gait performance in individuals with stroke was related to using less segmental transverse rotation and more antiphase CRP during walking. While not different from subjects without disability at equivalent walking speeds, transverse CRP values were correlated with deficits in functional gait only in the stroke group, and an inability to increase CRP may be a limiting factor in increasing walking speed. Data support the emphasis on trunk reeducation approaches in stroke rehabilitation. Our findings may suggest that improving intersegmental coordination and reducing excessive thoracic rotation may lead to improved functional outcomes in patients with stroke which merits further investigation.
LIMITATIONS OF THE STUDY AND GENERALIZABILITY
Since the study was conducted using a treadmill, the applicability of the results to overground walking may be limited.40 We used a self-paced treadmill that responds to the individual's walking speed, rather than imposing a set speed, in order to minimize the difference between treadmill walking and overground walking. The use of metronome pacing may also have affected the results. However, gait patterns41 would be similarly affected by metronome pacing in both groups.42 Aside from some elements (eg, items of the FGA and the Timed Up and Go Tests), the BesTest has not been validated in stroke; therefore, results with respect to balance function should be interpreted with caution. Finally, the selection criteria for the stroke group, which required them to be high functioning, as well as the small sample size limit the generalizability of results.
We thank Melanie Banina, Valeri Goussev, Gevorg Chilingaryan, and Rhona Guberek for their assistance and the subjects who participated in the study.
1. Duarte E, Marco E, Muniesa J, et al. Trunk control test as a functional predictor in stroke
patients. J Rehabil Med. 2002;34:267–272.
2. Carr JH, Shepherd RB. Stroke
Rehabilitation: Guidelines for Exercise and Training to Optimize Motor Skill. Butterworth-Heinemann; 2003.
3. Verheyden G, Nieuwboer A, Mertin J, Preger R, Kiekens C, De Weerdt W. The Trunk Impairment Scale: a new tool to measure motor impairment of the trunk after stroke
. Clin Rehabil. 2004;18:326–334.
4. Verheyden G, Nieuwboer A, Feys H, Thijs V, Vaes K, De Weerdt W. Discriminant ability of the Trunk Impairment Scale: a comparison between stroke
patients and healthy individuals. Disabil Rehabil. 2005;27:1023–1028.
5. Verheyden G, Vereeck L, Truijen S, et al. Trunk performance after stroke
and the relationship with balance, gait and functional ability. Clin Rehabil. 2006;20:451–458.
6. Karatas M, Çetin N, Bayramoglu M, Dilek A. Trunk muscle strength in relation to balance and functional disability in unihemispheric stroke
patients. Am J Phys Med Rehabil. 2004;83:81–87.
7. Wagenaar RC, Beek WJ. Hemiplegic gait: a kinematic analysis using walking speed as a basis. J Biomech. 1992;25:1007–1015.
8. Bruijn SM, Meijer OG, van Dien JH, Kingma I, Lamoth CJC. Coordination
of leg swing, thorax rotations, and pelvis rotations during gait: the organisation of total body angular momentum. Gait Posture. 2008;27:455–462.
9. Van Emmerik REA, Wagenaar RC. Effects of walking velocity on relative phase dynamics in the trunk in human walking. J Biomech. 1996;29:1175–1184.
10. Lamontagne A, De Serres SJ, Fung J, Paquet N. Stroke
affects the coordination
and stabilization of head, thorax and pelvis during voluntary horizontal head motions performed in walking. Clin Neurophysiol. 2005;116:101–111.
11. Lamoth C, Beek P, Meijer O. Pelvis-thorax coordination
in the transverse plane during gait. Gait Posture. 2002;16:101–114.
12. Haken H, Kelso JAS, Bunz H. A theoretical model of phase transitions in human hand movements. Biol Cybern. 1985;51:347–356.
13. Van Emmerik R, McDermott W, Haddad J, Van Wegen E. Age-related changes in upper body adaptation to walking speed in human locomotion
. Gait Posture. 2005;22:233–239.
14. Lamoth CJC, Meijer OG, Daffertshofer A, Wuisman PIJM, Beek PJ. Effects of chronic low back pain on trunk coordination
and back muscle activity during walking: changes in motor control. Eur Spine J. 2006;15:23–40.
15. Lamoth CJC, Daffertshofer A, Meijer OG, Beek PJ. How do persons with chronic low back pain speed up and slow down? Trunk-pelvis coordination
and lumbar erector spinae activity during gait. Gait Posture. 2006;23:230–239.
16. Seay JF, Van Emmerik REA, Hamill J. Low back pain status affects pelvis-trunk coordination
and variability during walking and running. Clin Biomech. 2011;26(6):572–578.
17. Van Emmerik REA, Wagenaar RC, Winogrodzka A, Wolters EC. Identification of axial rigidity during locomotion
in Parkinson disease. Arch Phys Med Rehabil. 1999;80:186–191.
18. Levin MF, Hacmon RR, Krasovsky T, Lamontagne A, Levin MF. Inter-segmental coordination
of the trunk and functional locomotor deficits in chronic stroke
. Soc Neurosci Abstr. 2011;587:03.
19. Gowland C, Stratford P, Ward M, et al. Measuring physical impairment and disability with the Chedoke-McMaster Stroke
20. Gauthier L, Dehaut F, Joanette Y. The Bells Test: a quantitative and qualitative test for visual neglect. Int J Clin Neuropsychol. 1989;11:49–54.
21. Fugl-Meyer A, Jääskö L, Leyman I, Olsson S, Steglind S. The post-stroke
hemiplegic patient. 1. A method for evaluation
of physical performance. Scand J Rehabil Med. 1975;7:13–31.
22. Wrisley DM, Marchetti GF, Kuharsky DK, Whitney SL. Reliability, internal consistency, and validity of data obtained with the functional gait assessment. Phys Ther. 2004;84:906–918.
23. Horak FB, Wrisley DM, Frank J. The Balance Evaluation
Systems Test (BESTest) to differentiate balance deficits. Phys Ther. 2009;89:484–498.
24. Shumway-Cook A, Horak B. Assessing the influence of sensory interaction on balance. Phys Ther. 1986;66:1548–1550.
25. Podsiadlo D, Richardson S. The Timed Up & Go: a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39:142–148.
26. Lin JH, Hsu MJ, Hsu HW, Wu HC, Hsieh CL. Psychometric comparisons of 3 functional ambulation measures for patients with stroke
27. Mulavara AP, Bloomberg JJ. Identifying head-trunk and lower limb contributions to gaze stabilization during locomotion
. J Vestib Res. 2002;12:255–270.
28. Stinear JW, Byblow WD. Phase transitions and postural deviations during bimanual kinesthetic tracking. Exp Brain Res. 2001;137:467–477.
29. Chen G, Patten C, Kothari DH, Zajac FE. Gait differences between individuals with post-stroke
hemiparesis and non-disabled controls at matched speeds. Gait Posture. 2005;22:51–56.
30. Perry J, Garrett M, Gronley JAK, Mulroy SJ. Classification of walking handicap in the stroke
31. Richards C, Malouin F, Dumas F, Tardif D. Gait velocity as an outcome measure of locomotor recovery after stroke
. In:Craik RL, Oatis CA, eds. Gait Analysis: Theory and Application. St Louis, MO: Mosby; 1995:355–364.
32. Robinson CA, Shumway-Cook A, Matsuda PN, Ciol MA. Understanding physical factors associated with participation in community ambulation following stroke
. Disabil Rehabil. 2011;33:1033–1042.
33. Schmid A, Duncan PW, Studenski S, et al. Improvements in speed-based gait classifications are meaningful. Stroke
34. Daly JJ, Sng K, Roenigk K, Fredrickson E, Dohring M. Intra-limb coordination
deficit in stroke
survivors and response to treatment. Gait Posture. 2007;25:412–418.
35. Kim CM, Eng JJ. Magnitude and pattern of 3D kinematic and kinetic gait profiles in persons with stroke
: relationship to walking speed. Gait Posture. 2004;20:140–146.
36. Archambault P, Pigeon P, Feldman A, Levin M. Recruitment and sequencing of different degrees of freedom during pointing movements involving the trunk in healthy and hemiparetic subjects. Exp Brain Res. 1999;126:55–67.
37. Cirstea M, Levin MF. Compensatory strategies for reaching in stroke
. Brain. 2000;123:940–953.
38. Michaelsen SM, Luta A, Roby-Brami A, Levin MF. Effect of trunk restraint on the recovery of reaching movements in hemiparetic patients. Stroke
39. Ustinova KI, Goussev VM, Balasubramaniam R, Levin MF. Disruption of coordination
between arm, trunk, and center of pressure displacement in patients with hemiparesis. Mot Control. 2004;8:139–159.
40. Bayat R, Barbeau H, Lamontagne A. Speed and temporal-distance adaptations during treadmill and overground walking following stroke
. Neurorehabil Neural Repair. 2005;19(2):115–124.
41. Ford MP, Wagenaar RC, Newell KM. The effects of auditory rhythms and instruction on walking patterns in individuals post-stroke
. Gait Posture. 2007;26:150–155.
42. Roerdink M, Lamoth CJ, Kwakkel G, van Wieringen PC, Beek PJ. Gait coordination
: benefits of acoustically paced treadmill walking. Phys Ther. 2007;87:1009–1022.