The Effects of Constraint-Induced Therapy on Kinematic Outcomes and Compensatory Movement Patterns: An Exploratory Study

Article Outline

• Abstract
• Methods
  • Participants
  • Intervention
  • Evaluations
    • Kinematic motion analysis acquisition
    • Functional assessments
  • Statistical Analysis
• Results
  • Kinematic Outcomes
    • Unaffected reach
    • Stroke-affected reach
    • Shoulder abduction as a compensatory movement
    • Spatiotemporal parameters
    • Functional outcomes
• Discussion
  • Implications for Intervention
  • Study Limitations and Future Directions
• Conclusions
• Acknowledgment
• References
• Copyright

Abstract 

Massie C, Malcolm MP, Greene D, Thaut M. The effects of constraint-induced therapy on kinematic outcomes and compensatory movement patterns: an exploratory study.

Objective

To determine changes in kinematic variables and compensatory movement patterns of survivors of stroke completing constraint-induced therapy (CIT).

Design

Pre-post, case series.

Setting

Clinical rehabilitation research laboratory.

Participants

Men (n=7) and women (n=3) with unilateral stroke occurring at least 9 months prior to study entry with moderate, stable motor deficits.

Intervention

Participants completed 10 consecutive weekdays of CIT for 6 hours a day comprised of trainer-supervised, functionally based activities using massed practice.

Main Outcome Measures

Kinematic measures included movement time, average velocity, trajectory stability, shoulder abduction, and segmental contribution. Functional measures included Wolf Motor Function Test (WMFT) performance time and functional ability scores and Motor Activity Log (MAL) “how-well” scores. All measures were administered before and after the 2-week CIT intervention.

Results

Movement time, average velocity, and trajectory stability significantly improved after CIT. Participants used more shoulder flexion to reach after CIT, but also demonstrated increased compensatory shoulder abduction. Functional scores also significantly improved, including WMFT performance time and functional ability and MAL scores. There was no change in trunk movement or amount of elbow extension.

Conclusions

CIT improved motor capacities in the hemiparetic arm as reflected in the functional outcomes and in some kinematic measures. Participants' reliance on common compensatory movements was not beneficially affected by CIT. The results of this study demonstrate that while functional capacity and some movement strategies in the hemiparetic arm improve after CIT, participants may not overcome their reliance on common compensatory movement patterns. Based on these findings, this study suggests that CIT may encourage subjects to generate movement through compensatory and/or synergy-dominated movement rather than promote the normalization of motor control. This outcome highlights the need to develop CIT further as an intervention that improves functional capacity and more normative movement strategies.

Key Words: Rehabilitation, Stroke

List of Abbreviations: CIT, constraint-induced therapy, EXCITE, Extremity Constraint Induced Therapy Evaluation, MAL, Motor Activity Log, WMFT, Wolf Motor Function Test

STROKE IS THE LEADING cause of adult long-term disability in the United States, often resulting in reaching impairments that may limit autonomy in activities of daily living and quality of life.¹ After a stroke, the capacity for central control of movement is typically severely compromised because of damage to neural mechanisms that control voluntary movement.² This damage leads to weakness, abnormal muscle tone, and stereotypical movement synergies that collectively limit functional reach. Consequently, survivors of stroke often rely on compensatory movement strategies to accomplish reaching tasks.³, ⁴, ⁵, ⁶ Compensatory strategies are considered maladaptive and are often detrimental to recovery of necessary movement capacities (eg, use of elbow extension or shoulder flexion).⁵ Although neurorehabilitation research has recently demonstrated that structured, specific, and intensive training protocols increase the amount of hemiparetic limb use, less attention has been given to normalizing movement strategies poststroke.⁷

CIT is a more recently developed intervention designed to restore motor skill capacity through massed practice of functional activities with the stroke-affected upper extremity.⁸, ⁹ Training concepts of CIT focus on re-establishing basic limb use by increasing attention to and use of the affected side. The signature protocol of CIT is well supported as an intervention of 6 hours of therapy a day for 10 consecutive weekdays while wearing a restraining mitt on the less affected extremity for up to 90% of waking hours.¹⁰, ¹¹, ¹², ¹³, ¹⁴ The efficacy of the signature CIT intervention was demonstrated in the recently completed EXCITE trial,¹¹ which investigated over 200 survivors of stroke in a randomized controlled design. Along with other smaller CIT trials, findings from the EXCITE project demonstrated a substantial and lasting increase in amount of hemiparetic arm use. With the success of CIT, considerable efforts have also been made to develop modified CIT protocols.¹⁰, ¹⁵, ¹⁶, ¹⁷, ¹⁸ There is, however, no consensus on the most efficacious form of modified CIT.¹⁹ Although evidence supports modified CIT protocols, Sterr et al¹⁰ found that 6 hours a day of training is superior to 3 hours a day. This finding suggests that intensity may be the main factor that differentiates modified and the signature CIT protocol.¹⁹ Given the limited consistency across modified CIT protocols and a greater amount of evidence supporting signature CIT protocol, the current study employed the standard, 6 hours a day of intervention.

The bulk of CIT outcomes have focused on functional measures of change including the WMFT and the MAL. However, inconsistencies in the quality of movement ratings using the WMFT functional ability scale are reported in CIT literature.¹¹, ²⁰ The MAL is a participant-rated quality of movement scale based on different activities of daily living, and reported outcomes indicate real-world improvements after CIT.¹¹ Kunkel et al¹³ have suggested that while subject ratings of quality of movement on the MAL substantially improved after CIT, movement still exhibited a substantial deficit. One method that has been proposed to clarify movement deficits after CIT is kinematic motion analysis because it can objectively and quantitatively describe the geometry of movement in clinically relevant outcome measures.⁷ Furthermore, postintervention changes in motor control strategies involving range of motion, multijoint control, movement velocity, and timing are readily detected through kinematic analyses. Outcomes that rely on subject ratings or therapist perceptions of movement validly demonstrate improvements in functional capacity but are limited in revealing specific changes in poststroke movement strategies.

To our knowledge, such specific assessment of postintervention kinematics has not been reported on the signature CIT protocol. A limited number of studies, however, have employed kinematic motion analysis to objectively and quantitatively measure changes after modified CIT protocols.¹⁵, ¹⁶, ¹⁸, ²¹ These studies reported spatiotemporal measures such as movement duration, reaction time, normalized movement units, and normalized jerk scores. Caimmi et al¹⁵ demonstrated improvements in normalized jerk scores and movement duration but did not include assessment of any change in motor strategy in the context of compensatory or synergy-driven movement. Similarly, Wu et al¹⁸ reported significant improvement in normalized movement time after CIT. In contrast, Wu et al²¹ found no significant improvement in movement time when comparing a modified CIT protocol with traditional rehabilitation. Based on commonly reported outcome measures (both functional and kinematic), movement deficits that remain after signature or modified implementation of CIT protocol, and the extent to which such changes in movement capacities depend on compensatory motor strategies, are unclear.

Carr and Shepherd⁶ suggest that compensatory strategies are the result of using available movements given the poststroke state of the central nervous system, which leads to long-term functional limitations. Because CIT is an intervention focused on overcoming learned nonuse by massed practice of available movement strategies, limited attention may be directed to the quality of movements being performed. Therefore, distinguishing between recovery of normalized movement patterns and compensatory movement patterns is critical to understand the mechanisms underlying functional improvements after CIT (ie, overcoming learned nonuse).⁷, ²² To clarify further the role of CIT on motor recovery in the chronic stage of stroke, this study employed a detailed kinematic motion analysis to determine how CIT influenced movement patterns, including compensatory strategies and spatiotemporal parameters of movement. We hypothesized that participants would exhibit changes in movement strategies after CIT that may or may not change the extent to which survivors of stroke rely on compensatory reaching strategies. A secondary hypothesis was that participants would demonstrate significant improvement in spatiotemporal parameters of reach and in functional outcome measures (ie, WMFT and MAL).

Back to Article Outline

Methods 

Participants 

A convenience sample of participants enrolled in a separate randomized controlled CIT study was used for this study. Ten participants (3 female; 5 left cerebral vascular accident) with a mean age ± SD of 61±14.7 years participated and gave written consent in accordance with the policies of the local institutional review board. Table 1 summarizes participant demographics. Participants were recruited from the community and met the following inclusion criteria: at least 9 months poststroke of unilateral clinical presentation; at least 10° of active wrist extension and 10° of extension in at least 2 fingers and thumb; approximately 30° of active shoulder flexion; at least half the normative passive range of motion at all upper-extremity joints; ability to follow simple instructions and multistep commands; endurance to complete 6 hours of training; a score of 24 or higher on the Mini-Mental State Examination²³; the ability to sit independently without back or arm support for 5 minutes; and the ability to stand with or without the assistance of a cane, quad cane, or hemiwalker for 2 minutes. Exclusion criteria included the following: any health problems judged by the screening physician to put the client at significant risk of harm during the study, other neurologic conditions (eg, multiple sclerosis, Parkinson disease), drugs or injections treating spasticity within 3 months of participation, significant stroke-affected arm use during daily living (MAL “amount-of-use” score≥2.5), and a pain score greater than 5 on the McGill Pain Scale. These are typical selection criteria in CIT studies.¹¹ Participants were required to obtain a medical release from their primary physicians.

Table 1. Demographic Characteristics of Participants

Abbreviations: F, female; LCVA, left cerebral vascular accident; M, male; NA, not applicable; RCVA, right cerebral vascular accident.

Intervention 

Participants completed 2 weeks of CIT training based on the recently published EXCITE multicenter clinical trial.¹¹ During 10 consecutive weekdays, participants completed a daily 6-hour on-site trainer-supervised program of functionally based activities using massed practice. Task parameters (eg, spatial and/or temporal) were manipulated in each successive period of task practice requiring increased control of the affected arm and hand.²⁴ Global feedback was provided at the end of a training task, and during task practice if performance substantially strayed from the intended goal.²⁴ The therapy trainers did not verbally prompt participants to limit compensatory strategies. Examples of tasks include playing checkers, washing windows, and stacking blocks. Short rest breaks (≈5min) were taken throughout the day as needed to prevent excessive fatigue. In addition to training, participants were instructed to wear a padded mitt on their less-affected side for 90% of their waking hours. Compliance with study protocol was monitored by a home diary that participants completed daily during the 2-week intervention, and with a mitt compliance device. The mitt compliance device, housed within the padded mitt, included a capacitive sensor and timer circuit that actively recorded wearing time. These time logs were reviewed daily with a therapy-trainer to monitor compliance with the intervention and mitt-wearing protocol. If the amount of mitt wearing strayed from 90%, a trainer educated and problem-solved with the participant to increase mitt-wearing compliance within safety limitations. Safety while wearing the padded mitt was emphasized, and participants were instructed on specific times to remove the mitt (eg, while driving).

Evaluations 

Kinematic motion analysis acquisition 

A reaching task comprised of flexion-extension movements at the elbow and shoulder was used, because these movements are core components of functional reach used during daily activities.²⁵, ²⁶ See figure 1 for experimental setup. Participants sat comfortably in a chair directly in front of a table with a 10° incline. Two targets, 7.7cm in diameter, were positioned in the sagittal plane of the hemiparetic arm at the point where the tip of the middle finger made contact at maximal arm extension and at a natural returning position. If the participant was unable to make contact with the middle finger consistently, the most distal part of the hand that could make contact with the targets was used. The reaching task consisted of 4 flexion-extension movements alternating between the 2 targets, and participants were asked to reach as fast as they could.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  Schema of experimental setup. Seated subjects reached with the stroke-affected arm between a proximal target and a distal target placed at maximum reach of the stroke-affected arm in a sagittal plane. Participants were instructed to tap back and forth as fast as they could, alternating between proximal and distal targets.

Arm kinematics were recorded at 60Hz with a 3-dimensional camera-based motion analysis system.^a One camera was placed directly ahead of the participant, and 2 were placed at 45° angles diagonally ahead of the participant. Reflective markers were placed on the sternal notch, shoulder, elbow, and wrist of the paretic arm. A metal probe was taped onto a rubber finger protector placed on the middle finger of the participant (if a participant was unable to use the middle finger, the probe was attached on the part of the hand that could make contact with the target). The metal probe served as a switch to complete a simple series circuit containing a 9-V battery. Output from the series circuit was a 9-V signal going to the event synchronization unit, creating a synchronization signal. The event synchronization unit also received input from the 3 cameras, superimposed the synchronization signal, and sent the signal to videocassette recorders. Software processed the 3 camera views and computed a sequence of 3-dimensional coordinates for each reflective marker, relative to the coordinate system built into the table surface (Motus).^a Movement coordinates were calculated based on the following axes: medial-lateral was the x-axis, anterior-posterior was the y-axis, and inferior-superior was the z-axis. Joint angles and movement velocities were calculated by kinematic analysis software (Motus).^a All data were exported to Microsoft Excel^b for data reduction and then to SPSS^c for data analysis.

The primary hypothesis addressed movement pattern changes in relation to compensatory strategies. To consider the impact of compensatory trunk movement on overall reach and how multiple segments collectively produce a reaching movement, multisegment contribution was defined as the proportion of total movement accounted for by movement at the trunk, shoulder, and elbow. The relative contribution of each joint, d_jt calculated in centimeters for both shoulder and elbow, was based on the following equation:

where d_target (cm) is the distance between the targets, d_trunk (cm) is the trunk anterior displacement during reach, and jt_a and jt_b are the excursion angles from proximal to distal target contacts for each joint. Anterior trunk displacement (cm) was computed as sagittal movement (y-axis) of the sternal notch marker (see fig 1). Elbow flexion/extension was the angle between vectors formed by the shoulder to elbow and elbow to wrist where full extension equaled 180°. Shoulder flexion/extension was the angle between vectors of elbow to shoulder markers and a unit vector projecting vertically from the shoulder marker as projected on the yz-plane (arm alongside body=0°). A mean for each segment was calculated based on the 4 reaching cycles.

Excessive amounts of shoulder abduction often contribute to the compensatory reaching strategy after stroke. Shoulder abduction was defined as the angle between vectors of elbow to shoulder markers and a unit vector projecting vertically from the shoulder marker as projected on the xz-plane (arm alongside body=0°). Measured in degrees, shoulder abduction was calculated as a mean of 4 reaches at (1) proximal target contact, (2) the maximum amplitude of the wrist marker during flexion and extension movements, and (3) distal target contact.

Spatiotemporal parameters included 3 different measures: (1) trajectory variability, (2) total movement time in seconds, and (3) average reach velocity in centimeters a second. We assessed variability of the trajectory as previously defined by Thaut et al.²⁶ From the spatial distribution (frontal plane) of the wrist marker as it reached maximum amplitude, mean distance and coefficient of variation were calculated and used as a measure of variability. A tighter clustering of coordinates will result in a decrease in the coefficient of variation, suggesting a more stable trajectory (ie, the wrist follows a more consistent path). Total movement time was recorded in seconds. A reach velocity (cm/s) was calculated for each reaching movement within the trial as v_reach = d_wrist/t_reach, where d_wrist was the sagittal displacement of the wrist, and t_reach was the reaching time from proximal to distal target contact; the mean of the 4 reach velocities was calculated.

Kinematic data were collected on the unaffected side of 4 participants to provide a descriptive analysis of unaffected upper-extremity arm movements to compare with the stroke-affected side. These data were not used for statistical purposes and are reported to elucidate the differences between the stroke-affected and unaffected movement patterns.

Functional assessments 

The WMFT and MAL (“how-well” scale) were conducted at baseline and posttest. The WMFT is a laboratory-based motor assessment of 17 different tasks, including 15 timed tasks, and has established reliability.⁹, ¹³, ²⁷ The WMFT incorporates gross and fine motor tasks, integrating different upper-extremity movements such as reaching, lifting a pencil, turning a key, and folding a towel. The speed tasks are videotaped and subsequently scored for functional ability of movement on a 6-point ordinal scale. The mean performance time and functional ability scores of the stroke-affected upper extremity are reported for the WMFT. The MAL, as described by Uswatte et al,²⁸ is a reliable and valid measure of participants' perception of real-world use of the hemiparetic arm conducted as a semistructured interview. Two 6-point scales may be used: one measures amount of use, and the other measures how well participants feel they can use the hemiparetic arm. These scales are anchored at 6 points (0=never use, 5=same as prestroke). Uswatte²⁸ suggest that the amount-of-use scale may be artificially inflated because of the nature of CIT training that focuses on increasing the amount of use in the stroke-affected side. The “how-well” scale is not subject to such inflations in amount of use and therefore is a better indicator of quality of movement improvement. The “how-well” scale was used for purposes of this study with mean scores reported.

Statistical Analysis 

Kinematic variables including segmental contribution and shoulder abduction were analyzed statistically using 2-tailed, dependent-sample t tests. Spatiotemporal parameters and WMFT performance time measures were analyzed using 1-tailed, dependent-sample t tests. Ordinal data from functional quality of movement measures (WMFT functional ability and MAL “how-well”) were analyzed using a 1-tailed Wilcoxon signed-rank test. The significance level was set at α equal to .05 for all statistical comparisons and was not adjusted considering multiple comparisons in light of the preliminary nature and size of the study.

Back to Article Outline

Results 

Kinematic Outcomes 

Unaffected reach 

Data from the unaffected side (n=4; participants 2, 7, 8, 10) are presented to illustrate differences between unaffected and stroke-affected reaching patterns. Figure 2A (top panel) illustrates the reaching strategy of a representative participant (participant 10, presented as a mirror-image) and the segmental contribution to the total reaching movement. With the unaffected arm, the trunk remained relatively stable in a neutral position and contributed very little to the overall reaching movement. During reach, shoulder flexion and adduction occurred as the elbow extended. Shoulder abduction was greatest when the hand contacted the proximal target and decreased as the shoulder flexed and elbow extended (fig 3). The trajectories were smooth and followed a consistent and stable trajectory (fig 4C).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 2. 

  Graphical representation of group means (unaffected, n=4; stroke-affected, n=10) of the relative amount of movement accounted by trunk movement, shoulder flexion, and elbow extension (left) and schematic reaching of a representative subject's (subject 10) strategy of the affected limb pre-CIT and of the unaffected limb (right). Trunk and arm configurations are shown at proximal contact (gray) and at distal contact (dashed lines). (A) Reach with the unaffected side occurred with shoulder flexion and elbow extension with little movement at the trunk. Schematic (subject 10) is shown as a mirror-image. (B) Pre-CIT reach with the stroke-affected arm. The bird's-eye view of a representative reaching strategy (subject 10) illustrates that the trunk and arm moved together as fixed unit with large trunk displacement and little shoulder flexion.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 3. 

  Schematic illustration of shoulder abduction regression post-CIT (gray=unaffected; solid=pre-CIT; dashed=post-CIT). Participants used significantly more shoulder abduction post-CIT during the early stages of reach (P<.05). Compared with the unaffected side, clearly illustrated is the increased use of shoulder abduction as a compensatory strategy that increased post-CIT.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 4. 

  (A) Scatter plot of a representative subject's trajectory variability (subject 8). (B) Clearly illustrated is closer clustering of coordinates post-CIT. Post-CIT coefficient of variation for the group mean (n=10) was 47.91%±10.5%, which was a significant post-CIT decrease (P<.017). (C) The coefficient of variation for unaffected reach group mean (n=4) was 24.57%±9.9%.

Stroke-affected reach 

With the stroke-affected side, trunk displacement accounted for a large proportion of the overall movement compared with the trunk displacement when reaching with the unaffected side (see fig 2B, bottom panel). The remaining distance was attributed to approximately equal contributions from the shoulder and elbow. In the representative participant (participant 10), the trunk, shoulder, and elbow appear to move as a unit rather than the trunk remaining stable while the shoulder flexes and the elbow extends. Figure 5 illustrates the change in contribution of each segment (ie, trunk, shoulder, elbow) after CIT. Shoulder flexion accounted for 15.47±4.69cm before CIT and significantly increased to 16.69±4.05cm post-CIT (t=–2.496; P=.034; d=1.22). Trunk movement decreased from 6.71±2.75cm pre-CIT to 6.42±2.94cm post-CIT, but this change was not significant (t=.44; P=.67). Elbow extension accounted for 12.32±5.26cm pre-CIT and decreased to 11.4±5.9cm after CIT, but also not significantly (t=1.3; P=.22).

• 
  • View Large Image
  • Download to PowerPoint

• Fig 5. 

  Segmental contribution in distance. Mean distance (cm) and SE are represented at pre-CIT (black) and post-CIT (pre). After CIT, shoulder flexion accounted for significantly more movement (*P<.05); the decreases in elbow and trunk movement were not significant.

Shoulder abduction as a compensatory movement 

Participants demonstrated excessive shoulder abduction during reach with the stroke-affected arm, as illustrated in figure 3. The amount of shoulder abduction at the time of contact with the proximal target significantly increased from 39.5°±5.58° pre-CIT to 42.74°±3.0° post-CIT (t=–2.42; P=.04; d=–0.77). At midtrajectory before CIT, the shoulder was abducted 44.95°±9.50° and significantly increased to 47.41°±6.90° after CIT (t=–2.595; P=.018; d=–0.30). There was a decrease in shoulder abduction from midtrajectory to the point of contact with the distal target. The amount of shoulder abduction at the time of contact with the distal target decreased by .28° after CIT, but this was not significant (t=.125; P=.90).

Spatiotemporal parameters 

Spatiotemporal parameters included trajectory stability, movement time, and mean reach velocity. The scatter plots in figure 4 represent the trajectory coordinates of the wrist marker. When the participant reached with the stroke-affected arm before CIT training, the coordinates of the wrist were not closely clustered around the mean and had higher maximum amplitudes. The wrist trajectory followed a more consistent path (closer clustering) after CIT. Trajectory variability, expressed as a coefficient of variation, significantly decreased from 58.85%±10.51% pre-CIT to 47.91%±17.9% post-CIT (t=2.727; P=.007; d=.85). Movement time to complete 4 reaching cycles significantly decreased from 9.17±2.7s pre-CIT to 7.58±3.14s after CIT (t=3.991; P=.002; d=1.88). The mean reaching velocity also increased significantly from 26.05±14.33cm/s pre-CIT to 34.31±17.19cm/s post-CIT (t=–2.95; P=.01; d=1.39). Results for movement effectiveness outcomes are displayed in table 2.

Table 2. Descriptive and Inferential Statistics for Spatiotemporal Kinematic and Clinical Assessments

⁎Statistic was 1-tailed, paired-samples t test. ES was Cohen d and was corrected for appropriate sign.

†Statistic was Wilcoxon signed-rank pairs. ES was calculated based on r=z/.

‡P<.05.

Functional outcomes 

The results of the functional outcomes are reported in table 2. WMFT performance time significantly decreased from 37.66±32.03s pre-CIT to 25.06±27.3s post-CIT (t=2.614; P=.014; d=1.23), and functional ability as scored on WMFT significantly improved from 2.45±0.66 units to 2.69±0.54 units after CIT (z=–2.19; P=.014; r=0.49). A score of 3 on the WMFT functional ability scale indicates that the participant accomplishes the task, but movement is influenced by synergy or performed slowly or with effort. MAL “how-well” scores significantly improved after CIT from 2.45±0.80 units to 2.90±0.74 units (z=–1.68; P=.047). A score of 3 on the MAL indicates that participants perceived they were able to use their weaker arm for that activity, but movements were slow or made only with some effort.

Back to Article Outline

Discussion 

A limited number of studies have employed objective and quantitative measures to investigate change in movement patterns after CIT, and these studies have focused primarily on modified CIT protocols and spatiotemporal parameters of movement.¹⁵, ¹⁶, ¹⁸, ²¹ The goal of our study was to expand on this previous work by examining those parameters that are perhaps most clinically meaningful—for example, how motor patterns and strategies change in relation to common compensatory movements. While we found that after CIT, the timing and trajectory control improved during hemiparetic reach, our results also revealed that the intervention promoted increased reliance on compensatory movements.

Relating to our hypothesis of changes in movement strategy after CIT, we sought to determine how trunk movement combines with shoulder flexion and elbow extension to accomplish goal-directed reach. Previous research indicates that stroke disrupts interjoint coordination between the shoulder and elbow²⁹, ³⁰; however, the contribution of trunk movement relating to this interjoint coordination has not been examined. Conversely, all 3 segmental contributions to movement have been studied independently.³¹ Less clear is how these 3 segments collectively interact and contribute to movement as part of a compensatory movement strategy. Our kinematic reaching task was set up to determine how each segment contributes collectively to movement and how CIT influences change in contribution.

In order for a participant to produce the same overall reaching distance, an increase in contribution from all of the segments together was not expected; and if a change occurred in 1 segment, an opposite change would occur in at least 1 other segment. For example, a pre-post decrease in trunk contribution might be paralleled by an increase in shoulder flexion. The findings from our study indicated that the amount of shoulder flexion significantly accounted for more of the reaching movement after CIT. As a result of accomplishing more of the reach with shoulder flexion, less movement was needed at the elbow and trunk during the reaching task. Accordingly, the amount of trunk flexion and elbow extension decreased slightly, although these decreases were not significant. The kinematic data revealed that CIT did not significantly reduce the participants' compensatory use of trunk movement during reach. Although the nature of CIT training tasks required increasing amounts of forward reach, the training may have had limited focus on improving the participant's ability to recruit both shoulder and elbow muscle groups.

In addition to excessive trunk movement, increased shoulder abduction during reach is a common compensatory movement after stroke. As is clearly illustrated in figure 3, reaching patterns substantially differ comparing the unaffected side with the stroke-affected side. Interestingly, the finding that the amount of shoulder abduction significantly increased after CIT suggests that this intervention may promote compensatory rather than normalized movements at the shoulder. The CIT protocol is predicated on improving functional capacity by dramatically increasing the amount of hemiparetic arm use through massed practice. Conversely, and perhaps explaining increased reliance on compensatory patterns, CIT typically places limited emphasis on how movements are executed. That is to say, CIT is less focused on the normalization of movement. Participants may use compensatory strategies such as shoulder abduction to accomplish training tasks with little regard for quality of movement. This strategy may be amplified as participants experience fatigue. Muscles controlling compensatory movements may be more easily recruited,³ and therefore that strategy is used to complete training tasks.

The results of the present study extend our knowledge of the impact of CIT on precise movement strategy changes. Compensatory movement strategies (ie, trunk movement and shoulder abduction) have not been studied in signature protocols of CIT, nor have they been reported in previous modified CIT kinematics studies.¹⁵, ¹⁶, ¹⁸, ²¹ Modified CIT protocols may influence compensatory movements, but the reported kinematics protocol have not systematically investigated these characterizations of movement. Additionally, movement strategies such as shoulder flexion and elbow extension generally have not been reported, with 1 exception. Caimmi et al¹⁵ reported a small but nonsignificant increase in the angle of the elbow at the end of the reaching movement. Our results parallel this finding such that CIT did not significantly improve the amount of elbow extension.

After CIT, the spatiotemporal control of movement improved, and our results parallel previous findings from modified CIT studies¹⁵, ¹⁶, ¹⁸, ²¹ showing decreased movement time during a kinematic task performed at a self-selected pace.¹⁵, ¹⁸ Our results expand on those findings by reducing movement time variability as participants were asked to perform as fast as they could. Further, in our study, movement times significantly decreased when participants were instructed to reach back and forth as fast as they could. Consistent with this finding, mean reach velocity of the wrist marker, another spatiotemporal measure, significantly increased post-CIT. The significant improvement in being able to reach more quickly can be attributed to improvement in motor skill capacity as promoted during CIT. For example, participants were asked to move faster, thus increasing temporal demands during subsequent task trials.

The findings from our study provide evidence that CIT improved the stability of reaching trajectories. Based on the methods of Thaut et al,²⁶ a decrease in trajectory variability translates to greater stability of movement during reach, such that participants had more control over the trajectory. After CIT, as participants reached back and forth between the targets, this trajectory path was more consistent. This finding supports repetitive training of movements, such as are required during CIT, to improve the ability to make smoother and steadier reaching patterns. CIT training requires repetitive reaching movements during a variety of tasks in many different planes of movement. As such, participants gain more control of their ability to reach even though the reach may still rely in part on compensatory movements.

Implications for Intervention 

Based on the present study, there are important implications of the overall impact of CIT on specific movement strategies in survivors of stroke. Improvements on the WMFT, a commonly reported functional outcome, demonstrate that CIT increases motor capacities that allow survivors of stroke to complete basic movement tasks more rapidly. Such results of the WMFT and our other functional outcomes are similar to the recently completed EXCITE trial,¹¹ and support the idea that CIT improves basic motor skills. However, and perhaps more importantly, our results extend beyond basic measures of functional capacity to explain how CIT influences motor patterns employed by participants. For example, increased functional ability may be attributed to the greater use of shoulder flexion during reach and the ability to reach faster. Although our results from functional measures (ie, WMFT functional ability scale and MAL “how-well” scale) suggest that quality of movement improves, our kinematic data present a different picture: CIT promoted reliance on compensatory movement patterns involving shoulder abduction and had little or no impact on compensations with the trunk. These findings highlight 2 important considerations for CIT and stroke rehabilitation research. First, quality of movement outcome measures commonly used have limited sensitivity to detect such specific changes in motor strategies including compensatory movement. Second, our findings also support the idea that traditionally, CIT is focused on the amount of movement rather than the execution of normalized movements. Although there is much debate over the differentiation between true motor recovery and compensatory motor recovery, there is a consensus that the mechanisms underlying recovery and the impact of therapy on these mechanisms must be better understood.³², ³³ To that end, more emphasis should be given to how interventions affect functional capacity and movement strategy, to understand further the impact of interventions on recovery of motor skills. This study is an initial step toward understanding how CIT influences movement strategies. The potential for CIT to entrain normative movement patterns better while still improving functional capacity may be improved with an evaluation of the specific techniques used in CIT.

Study Limitations and Future Directions 

We acknowledge some limitations within the current study. First, the change observed with kinematic motion analysis is not indicative of entire motor control system changes. The ability to control movement is a complex process involving many factors. Further research to study different aspects within the motor control system will enhance the understanding of how rehabilitation influences the central nervous system. For instance, future studies could combine electromyography with kinematic motion analysis. Second, the kinematic task was based on a gross-motor reaching movement requiring speed using the trunk, shoulder, and elbow. Future studies should include analysis of all degrees of freedom in the upper limb, change in total range of motion, and the influence of the self-selected versus fast movements. Finally, the small sample size limits the generalizability of our findings to the larger stroke population. As such, the next phase of investigations should include a random-control design in a larger group of survivors of stroke.

Back to Article Outline

Conclusions 

The results of this study demonstrate that spatiotemporal parameters of movement improved after CIT, and there was greater use of the shoulder during reach. However, while functional capacity and some movement strategies in the hemiparetic arm improved after CIT, participants did not overcome their reliance on common compensatory movement patterns. After CIT, shoulder abduction was more pronounced, and subjects continued to rely on trunk movement to accomplish reach. Based on these findings, our study provides initial evidence that CIT may encourage subjects to generate movement through compensatory and/or synergy-dominated movement rather than promote the normalization of motor control. This outcome highlights the need to study CIT further as an intervention that improves functional capacity and normalized movement strategies.

Suppliers

Back to Article Outline

Acknowledgment 

We thank Gary Kenyon, MS, for his support of kinematic data collection procedures.

Back to Article Outline

References 

1. American Heart Association. Heart disease and stroke statistics—2007 update. Circulation. 2007;115:e69–e171
   • View In Article
   • CrossRef
2. Krakauer JW. Arm function after stroke: from physiology to recovery. Semin Neurol. 2005;25:384–395
   • View In Article
   • MEDLINE
   • CrossRef
3. McCrea PH, Eng JJ, Hodgson AJ. Saturated muscle activation contributes to compensatory reaching strategies after stroke. J Neurophysiol. 2005;94:2999–3008
   • View In Article
   • MEDLINE
   • CrossRef
4. In:  Trombly CA,  Radomski MV editor. Occupational therapy for physical dysfunction. 5th ed.. New York: Lippincott Williams & Wilkins; 2002;
   • View In Article
5. Levin MF, Michaelsen SM, Cirstea CM, Roby-Brami A. Use of the trunk for reaching targets placed within and beyond the reach in adult hemiparesis. Exp Brain Res. 2002;143:171–180
   • View In Article
   • MEDLINE
   • CrossRef
6. In:  Carr J,  Shepherd R editor. Movement science: foundations for physical therapy in rehabilitation. 2nd ed.. Gaithersburg: Aspen Publishers; 2000;
   • View In Article
7. Krakauer JW. Motor learning: its relevance to stroke recovery and neurorehabilitation. Curr Opin Neurol. 2006;19:84–90
   • View In Article
   • MEDLINE
8. Morris D, Taub E. Constraint-induced therapy approach to restoring function after neurological injury. Top Stroke Rehabil. 2001;8:16–30
   • View In Article
   • MEDLINE
   • CrossRef
9. Taub E, Uswatte G, Pidikiti R. Constraint-induced movement therapy: a new family of techniques with broad application to physical rehabilitation—a clinical review. J Rehabil Res Dev. 1999;36:237–251
   • View In Article
   • MEDLINE
10. Sterr A, Elbert T, Berthold I, Kolbel S, Rockstroh B, Taub E. Longer versus shorter daily constraint-induced movement therapy of chronic hemiparesis: an exploratory study. Arch Phys Med Rehabil. 2002;83:1374–1377
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (68 KB)
    • CrossRef
11. Wolf SL, Winstein CJ, Miller JP, et al. Effect of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke—the EXCITE randomized clinical trial. JAMA. 2006;296:2095–2104
    • View In Article
    • CrossRef
12. Blanton S, Wolf SL. An application of upper-extremity constraint-induced movement therapy in a patient with subacute stroke. Phys Ther. 1999;79:847–853
    • View In Article
    • MEDLINE
13. Kunkel A, Kopp B, Muller G, et al. Constraint-induced movement therapy for motor recovery in chronic stroke patients. Arch Phys Med Rehabil. 1999;80:624–628
    • View In Article
    • Abstract
    • Full-Text PDF (1593 KB)
    • CrossRef
14. Taub E, Uswatte G, King DK, Morris D, Crago JE, Chatterjee A. A placebo-controlled trial of constraint-induced movement therapy for upper extremity after stroke. Stroke. 2006;37:1045–1049
    • View In Article
    • CrossRef
15. Caimmi M, Carda S, Giovanzana C, et al. Using kinematic analysis to evaluate constraint-induced movement therapy in chronic stroke patients. Neurorehabil Neural Repair. 2008;22:31–39
    • View In Article
16. Lin KC, Wu CY, Wei TH, Lee CY, Liu JS. Effects of modified constraint-induced movement therapy on reach-to-grasp movements and functional performance after chronic stroke: a randomized controlled study. Clin Rehabil. 2007;21:1075–1086
    • View In Article
    • CrossRef
17. Page SJ, Levine P, Leonard A, Szaflarski JP, Kissela BM. Modified constraint-induced therapy in chronic stroke: results of a single-blinded randomized controlled trial. Phys Ther. 2008;88:333–340
    • View In Article
    • CrossRef
18. Wu CY, Chen CL, Tsai WC, Lin KC, Chou SH. A randomized controlled trial of modified constraint-induced movement therapy for elderly stroke survivors: changes in motor impairment, daily functioning, and quality of life. Arch Phys Med Rehabil. 2007;88:273–278
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (141 KB)
    • CrossRef
19. Wolf SL. On “Modified constraint-induced therapy …” Page et al. Phys Ther. 2008;88:333–340 [Letter]. Phys Ther. 2008;88:680–684
    • View In Article
    • CrossRef
20. Bonifer NM, Anderson KM, Arciniegas DB. Constraint-induced therapy for moderate chronic upper extremity impairment after stroke. Brain Inj. 2005;19:323–330
    • View In Article
    • MEDLINE
    • CrossRef
21. Wu CY, Lin KC, Chen HC, Chen IH, Hong WH. Effects of modified constraint-induced movement therapy on movement kinematics and daily function in patients with stroke: a kinematic study of motor control mechanisms. Neurorehabil Neural Repair. 2007;21:460–466
    • View In Article
22. Cirstea AC, Levin ME. Improvement of arm movement patterns and endpoint control depends on type of feedback during practice in stroke survivors. Neurorehabil Neural Repair. 2007;21:398–411
    • View In Article
23. Folstein M, Folstein S, McHugh P. Mini-Mental State—practical method for grading cognitive state of patients for clinician. J Psychiatr Res. 1975;12:189–198
    • View In Article
    • Full-Text PDF (635 KB)
    • CrossRef
24. Winstein CJ, Miller JP, Blanton S, et al. Methods for a multisite randomized trial to investigate the effect of constraint-induced movement therapy in improving upper extremity function among adults recovering from a cerebrovascular stroke. Neurorehabil Neural Repair. 2003;17:137–152
    • View In Article
    • MEDLINE
25. Michaelsen SM, Jacobs S, Roby-Brami A, Levin MF. Compensation for distal impairments of grasping in adults with hemiparesis. Exp Brain Res. 2004;157:162–173
    • View In Article
    • MEDLINE
26. Thaut MH, Kenyon GP, Hurt CP, McIntosh GC, Hoemberg V. Kinematic optimization of spatiotemporal patterns in paretic arm training with stroke patients. Neuropsychologia. 2002;40:1073–1081
    • View In Article
    • MEDLINE
    • CrossRef
27. Morris DM, Uswatte G, Crago JE, Cook EW, Taub E. The reliability of the wolf motor function test for assessing upper extremity function after stroke. Arch Phys Med Rehabil. 2001;82:750–755
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (48 KB)
    • CrossRef
28. Uswatte G, Taub E, Morris D, Light K, Thompson PA. The Motor Activity Log-28—assessing daily use of the hemiparetic arm after stroke. Neurology. 2006;67:1189–1194
    • View In Article
    • CrossRef
29. Levin MF. Interjoint coordination during pointing movements is disrupted in spastic hemiparesis. Brain. 1996;119:281–293
    • View In Article
    • CrossRef
30. Cirstea MC, Mitnitski AB, Feldman AG, Levin MF. Interjoint coordination dynamics during reaching in stroke. Exp Brain Res. 2003;151:289–300
    • View In Article
    • MEDLINE
    • CrossRef
31. Michaelsen SM, Dannenbaum R, Levin MF. Task-specific training with trunk restraint on arm recovery in stroke—randomized control trial. Stroke. 2006;37:186–192
    • View In Article
    • CrossRef
32. Kwakkel G, Kollen B, Lindeman E. Understanding the pattern of functional recovery after stroke: facts and theories. Restor Neurol Neurosci. 2004;22:281–299
    • View In Article
    • MEDLINE
33. Latash ML, Anson JG. What are “normal movements” in atypical populations?. Behav Brain Sci. 1996;19:55–57
    • View In Article
    • CrossRef

• a Peak Performance Technologies, 7388 S Revere Pky, Englewood, CO 80112.
• b Microsoft Corp, One Microsoft Way, Redmond, WA 98052-6399.
• c SPSS, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.