Share this article on:

Effects of Constraint-Induced Movement Therapy on Gait, Balance, and Functional Locomotor Mobility

Zipp, Genevieve Pinto PT, EdD; Winning, Sue PT

doi: 10.1097/PEP.0b013e31823e0245
Research Articles

Purpose: This study reports the secondary effects of a constraint-induced movement therapy (CIMT) protocol on spatial temporal parameters of gait, balance, and functional locomotor mobility in children with cerebral palsy.

Methods: Sixteen children (4–12 years old) participated in a 3-week CIMT program. Participants were tested on the first and last day of the CIMT program using the Standardized Walking Obstacle Course (SWOC), the Pediatric Balance Scale (PBS) and the GAITRite Gold system (CIR Systems, Inc, Havertown, Pennsylvania).

Results: Wilcoxon signed rank tests were used on all pre- and posttests. Only the spatial temporal parameters of cadence and velocity differed significantly, with 12 subjects displaying a faster cadence (P = .02) and 10 subjects displaying a faster velocity (P = .05).

Conclusion: In this pilot study, CIMT was found to promote changes in 2 spatial temporal parameters of gait. However, no changes were noted in the participant's measures on the SWOC and PBS.

Examination of the effects of CIMT on gait parameters, balance, and locomotor ability revealed that cadence and velocity of gait were improved after a 6-week program. The authors recommend additional study with more sensitive tools to examine what may be additional effects.

Department of Graduate Programs in Health Sciences (Dr Zipp), Seton Hall University, South Orange, and Department of Physical Therapy and General Rehab (Ms Winning), Children Specialized Hospital, Toms River and Mountainside, New Jersey.

Genevieve Pinto Zipp, PT, EdD, Department of Graduate Programs in Health Sciences, Seton Hall University, 400 South Orange Ave, South Orange, NJ 07079 (

The authors declare no conflicts of interest.

Back to Top | Article Outline


Numerous researchers have investigated the effects of constraint-induced movement therapy (CIMT) in individuals with hemiplegia. Constraint-induced movement therapy engages the involved upper extremity (UE) in repetitive functional activity for a significant amount of time per day (4–6 hours) for several weeks, while constraining the uninvolved UE. By constraining the uninvolved UE, the involved UE engages in the performance of the task. On the basis of the need to use the involved UE in an environment that employs massed practice of functional tasks that are relevant to the performer, a rich environment, which supports neural plasticity, is created. By repeating functional movements in the involved UE, retention of the task increases via the development of new motor plans. Dean and Shepherd1 and Carr and Shephard2 demonstrated that the retention of motor performance is enhanced through the use of task repetition.

Although CIMT protocols presented in the research literature are not consistently the same, 2 key characteristics are consistent across all studies. First, the uninvolved UE is in some way controlled so that it cannot participate in the task. Second, active repetitive massed practice of functional activities using the involved UE is required. Consistently across studies the results have shown significant improvements in UE function in both adults and children regardless of their level of impairment.36

Constraint-induced movement therapy research has been focused specifically on the UE effects of this approach, and one question which has not been addressed in the literature is: “Does increasing the functional mobility of one's UE further enhance the subjects' overall balance capabilities, quality of gait, and functional locomotor mobility?” Given that the motor system functions as a dynamic system, as proposed by the dynamic systems model of motor control, it is hypothesized that improvements in UE functioning resulting from CIMT would positively influence lower extremity (LE) functioning. Thus, it is proposed that the effects of CIMT can be measured by changes in spatial temporal gait parameters, enhanced balance abilities, and functional locomotor mobility.

This hypothesis is based upon the generally accepted theory that central pattern generators (CPGs) located in the spinal cord are involved in the coordination and control of limb and body segment movements used to produce stereotypic locomotor patterns associated with gait.7 Findings from investigations using treadmill training protocols support the idea that CPG networks are activated through the sensory input produced by walking. The findings by Bermann et al 8 in a 4.5-year-old boy with a cervical level incomplete spinal cord injury, demonstrated the ability to walk following 16 weeks of intense task-specific training which included body weight–supported treadmill training and over ground walking, further support the notion that CPGs are involved in walking.

Craik et al9 demonstrated, however, that lower extremity rhythmical repetitive movements are not the sole factor responsible for the development of gait. Gait speed influences the coordination between both the legs and the arms thus creating an interdependent relationship between all 4 limbs. This interdependent relationship results in a 2:1 phase limb coordination ratio (2 arm swings to each leg stride) during slow walking with the emergence of a 1:1 ratio in phase limb coordination as gait speed increases. A similar rhythmical pattern has been noted between the trunk and pelvis, with the trunk and pelvis moving in phase at lower gait speeds and out of phase at higher speeds. Although interlimb coordination can enhance gait efficiency, it is not required for the initiation or maintenance of functional gait. If arm swing is limited, individuals will tend to shorten their stride length and decrease both stride frequency and velocity.10 When arm swing is absent or constrained in individuals with limited or absent UE motor control, such as children with hemiplegia, decreased trunk rotation is noted which may influence the efficiency of their gait.11

Therefore, the question remains as to how can we engage the UEs during gait in patients with limited UE motor control? Ford et al12 reported that when engaged in treadmill walking, patients with stroke produce a more normal arm swing. Nevertheless, for those patients who have learned not to use their involved UE the question remains, will the externally paced speed requirement set by the treadmill be enough to invoke the interlimb rhythmical coordination pattern and promote retention of interlimb coordination? Given that speed is the parameter that influenced interlimb coordination in the treadmill training experiments, the authors suggest that without the speed requirement interlimb coordination in patients who have learned not to use their UE will not be retained. However, the authors suggest that the theoretical premise surrounding CIMT offers a strong theoretical framework for the positive secondary effects that CIMT may have on gait, balance, and functional locomotor mobility. To date, the secondary effects of CIMT on gait have not been assessed and thus warrant further investigation. Therefore, the purpose of this pilot study was to determine if any secondary effects of CIMT could be seen in the participants' functional locomotor mobility, balance, and spatial temporal parameters of gait.

Back to Top | Article Outline



Participants were day campers in the Camp Open Arms summer program offered by Children's Specialized Hospital in New Jersey during the summer 2009 program. Solicitation flyers were given to the parents and children when they registered for the camp. All participants of the camp were eligible to take part in this pilot project. The camp only recruits children between 4 and 12 years of age with the diagnosis of hemiplegia confirmed by the referring physician. If camp participants were willing to take part in the assessment outlined in the flyer, they contacted the primary research investigator to initiate the assessment process. Eighteen children ranging from 4 to 12 years (M = 7.16, SD = 2.4) volunteered to take part in the Camp Open Arms program and assented to participation in the study along with their parents consent. Only 16 children completed both testing sessions, as 1 child was unable to attend the final session due to medical complications and 1 child did not complete all post session assessments due to noncompliance. Seton Hall University Institutional Review Board approved the study.

Back to Top | Article Outline


This pilot study employed 3 assessment instruments. The GAITRite Gold system was used to collect spatial temporal parameters of gait. The GAITRite Gold system provides an objective measurement of gait and can provide support for the efficacy of a treatment protocol. Electronic footprints instantly measure cadence and step length and velocity can be tracked, reported, and graphed. Thorpe et al13 report good intertrial reliability with children. Webster et al14 found intraclass correlation coefficients (ICCs) ranging from 0.92 to 0.99 for walking speed, cadence, step length, and step time, when comparing the GAITRite system to the Vicon MX system. Good test-retest reliability among younger and older adults (ICCs ranging between 0.82 and 0.92) was reported by Menant et al.15

The Standardized Walking Obstacle Course (SWOC) was used to measure functional locomotor mobility pre and post intervention. The SWOC determines ambulation capacity by measuring stability and speed during gait under different circumstances in a safe, reproducible, and efficient way.16 The SWOC was tested in 24 young adults with no known pathology and 13 older adults with arthritis and was reported to have excellent test-retest reliability and high inter tester reliability for each test item.17 Concurrent validity has also been demonstrated between the SWOC and the timed 50 Foot Walk Test.17

The Pediatric Balance Scale (PBS) is a modified version of the adult Berg Balance Scale, which assesses the balance abilities of children. The 14-item PBS was developed to measure balance in school-aged children with mild to moderate motor impairments. Franjoine et al18 determined the test-retest and interrater reliability of the PBS. Their results demonstrated that no significant differences existed in total test scores (ICC model3,1 = 0.998) or individual items (Kappa Coefficients, κ = 0.087–1.0; Spearman rank correlation coefficients, r = 0.89–1.0) as measured by 1 therapist on 2 occasions. No significant difference among ratings by different physical therapists was found on the PBS for the total test score (ICC3,1 = 0.997).

Back to Top | Article Outline


The participants engaged in 2 testing sessions on the 1st day of camp (pretest) and the last day of camp (posttest). Methods used during the pretest and posttest sessions were consistent in accord with a pretest-posttest research design. Each testing session lasted approximately 30 to 40 minutes.

The occupational therapists at Children Specialized Hospital facilities in Toms River and Mountainside, New Jersey, offered the summer “Camp Open Arms” for children with hemiplegia. The camp focused on improving UE use of the involved side in children with hemiplegia by utilizing CIMT strategies that emphasized repeating functional tasks in the involved UE in a fun and functionally relevant group environment. To encourage the use of the involved UE participants had a long arm fabric covered fiberglass splint fabricated prior to the program's start for use on their uninvolved UE during the camp. The cast was form fitted extending from mid-humerus to past the fingertips and encompassed the thumb. The splint design allowed for easy removal for hygiene purposes and was economically feasible. During all sessions of the camp the splint was worn, with supervision provided by the camp occupational therapists. Participants also took part in an occupational therapy pre- and post-camp evaluation, which assessed their UE functional abilities. This assessment, however, was not part of this study. Children's Specialized Hospital's occupational therapists planned and ran the Camp Open Arms program focusing on age appropriate bilateral and unilateral UE functional activities while the participants were standing, sitting, and walking. The camp ran for 3 weeks, 5 days per week, for 6 hours per day during the month of July. To maintain confidentiality, the camps occupational therapists were not informed of which campers volunteered to participate in this study. Table 1 presents the participants' demographic information.



Pre- and post-camp assessments occurred in a quiet location in the physical therapy gym and hallway. The primary investigator who collected the data has over 20 years of experience in pediatric physical therapy. To monitor the children throughout the testing sessions several other physical therapists assisted.

Upon arrival at the assessment session, the children were randomly assigned to begin with 1 of the 3 assessments to account for any possible effects associated with testing order. A code was assigned to all children to ensure anonymity throughout the testing session. The children's right and left LE lengths were measured from the greater trochanter to the lateral malleolus using a cloth tape measure. For the spatial temporal parameters of gait on the GAITRite, the children were instructed to walk at a comfortable pace from the beginning to the end of the travel path along the GAITRite mat. The travel path began 3.0 meters from the start of the GAITRite mat and ended 3.0 meters beyond the GAITRite mat. Each child completed 5 walking trials. Children were given the opportunity to sit and rest up to 1 minute between trials. The 5 self-paced walking trials were averaged to generate a pretest score, which was then compared with a posttest score that was the average of the 5 posttest trials. Data obtained from the GAITRite included step time, cadence, number of steps, and velocity.

Using the standard protocol outlined for the SWOC assessment, the children's functional locomotor mobility was assessed. The children completed 2 trials of the SWOC under 1 condition, which was walk hands free. Each trial included standing up, walking the course from the starting point to the ending point, and sitting down. All pretest scores were compared with the posttest score. Data obtained included time to complete the trial using a digital stopwatch, number of steps taken, number of stumbles, and the number of steps taken from the path of travel. The investigator required that the participants complete the SWOC test without the tray carrying task to allow the arms freedom to assist in the coordination of walking.

Using the standard protocol outlined for the PBS, the children's balance was assessed. The children completed 2 trials of the PBS. The overall total score on trial 2 of the pretest was compared with the overall total score on trial 2 of the posttest.

Back to Top | Article Outline

Data Analysis

Data were analyzed using SPSS version 15.0 (SPSS, Chicago, Illinois). To determine if the data were normally distributed, the Kolmogorov-Smirnov test was employed. On the basis of the significant values obtained, which indicated a deviation from normality for each variable tested, nonparametric testing was employed. To determine if the intervention provided during the camp secondarily affected the children's gait and balance, Wilcoxon signed rank tests were employed on all pretest and posttest measures including participants' spatial temporal gait parameters, specifically cadence, velocity, and number of steps; SWOC scores; and PBS scores.

Back to Top | Article Outline


No significant findings were observed on the PBS and the SWOC. However, when looking further at the mean change in the PBS (M = −2.29, SD = 5.08, N = 16) which was not significantly different, t 16 = −1.861, 1-tailed P = .060, a positive trend was noted (Table 2).



The Wilcoxon test showed that the results were significant for cadence, with 12 of 16 subjects displaying a faster cadence (number of steps or strides per unit of time) (P = .02). The Wilcoxon test showed that the results were significant for velocity, with 10 of 16 subjects displaying a faster velocity (P = .05).

Back to Top | Article Outline


This pilot study demonstrated that the engagement in CIMT positively affected spatial temporal characteristics of gait in children with hemiplegia as measured by changes in cadence and velocity. In addition, the positive trend noted in the balance data as measured by the PBS score may offer support for the possible effect of CIMT in promoting balance and thus warrants further investigation.

Surprisingly, no significant changes were noted in the children's functional capability scores as measured by the SWOC. Upon reflecting upon the results, the authors speculate that the lack of significance may have been a result of several factors. Factors may include functional differences between subjects, which were not controlled in this pilot, and low subject numbers. Also with the primary purpose of the SWOC being to determine ambulation capacity on unlevel surfaces, the question is raised as to whether or not a change in functional use of the involved UE would positively influence this more complex level of ambulation capacity required during the SWOC given this limited period of CIMT. Possibly, if given a longer period of practice the children could have become more proficient in using the involved UE routinely during their walking and demonstrated carryover when performing more complex ambulation tasks. Finally, the SWOC may not be sensitive enough to pick up subtle changes in functional locomotor mobility that may have emerged following this intervention. Future work should assess the utility of other functional measures such as the Gross Motor Function Measure in documenting functional locomotor ability following CIMT.

With the constraint of the uninvolved UE, the children, although proficient in ambulation prior to the application of CIMT, were now required to organize a different motor control strategy to meet the demands of the task of ambulation. The generation of a new motor control strategy may have forced the children into the initial stage of learning as defined by Gentile,19 which requires that the learner work to get close to what is expected in performing the task before refinement of performance seen in the later stage of learning. Thus, this question requires additional investigation to assess the effects of the stage of learning and practice schedule on the overall spatial and temporal characteristics of gait as well as balance and functional locomotor mobility. The authors suggest that it may also be plausible that the items tested in the PBS require a more integrated organization of the UE and LE to execute a motor plan and thus may require even further practice to organize effective interlimb coordination. Highly organized interlimb coordination can be seen for example in other functional tasks such as rising from a chair, retrieving objects from the floor or reaching forward and thus they may benefit from enhanced bilateral UE control and coordination activities. In future work comparing sway measures on a balance platform system to pre and post intervention PBS scores maybe insightful. Finally, using a 3-dimensional gait analysis system would have allowed the authors to evaluate changes in arm swing pre and post intervention and the influence of the UEs on gait and functional locomotor mobility.

Back to Top | Article Outline


Although the findings of this pilot study are limited by the instrumentation used to measure change, the results do raise important issues regarding the influence of CIMT on children's gait, balance, and potentially their functional abilities and participation level. Further exploration of some of the questions raised by these preliminary findings is warranted to address the limitations of this pilot study. By gaining a fuller understanding of the possible secondary effects of CIMT, therapists may be better able to enhance a patient's overall functional independence. On the basis of these pilot data, therapists using CIMT must consider and monitor the possible secondary effects of the intervention on a client's gait, balance, and functional locomotor mobility when designing and assessing a plan of care.

Back to Top | Article Outline


The authors thank all the children and their families who participated in the study, and the staff at Camp Open Arms summer program offered by Children's Specialized Hospital in Toms River and Mountainside, New Jersey, who allowed them to collect data.

Back to Top | Article Outline


1. Dean CM, Shepherd RB. Task-related training improves performance of seated reaching tasks following stroke: a randomized controlled trail. Stroke. 1997;28:722–728.
2. Carr JH, Shepherd RB. Training motor control, increasing strength and fitness and promoting skill acquisition. In:Carr JH, Shepherd RB, eds. Neurological Rehabilitation. Optimizing Motor Performance. Oxford, England: Butterworth-Heinemann; 1998:23–46.
3. Taub E, Ramey S, DeLuca S, Echols K. Efficacy of constraint-induced (CI) movement therapy for children with cerebral palsy with asymmetric motor impairment. Pediatrics. 2004;113:305–312.
4. Wolf SL, Winstein CJ, Miller JP, et al. Effects of constraint-induced movement therapy on upper extremity function 3 to 9 months after stroke: the EXCITE randomized clinical trial. JAMA. 2006;17:2095–2104.
5. Eliasson AC, Krumlinde-Sundholm L, Shaw K, Wang C. Effects of constraint-induced movement therapy in young children with hemiplegic cerebral palsy: an adapted model. Dev Med Child Neurol. 2005;47:266–275.
6. Hart H. Can constraint therapy be developmentally appropriate and child-friendly? Dev Med Child Neurol. 2005;47:363.
7. Zehr EP, Duysens J. Regulation of arm and leg movement during human locomotion. Neuroscientist. 2004;10(4):347–361.
8. Behrman A, Nair P, Bowden M, et al. Locomotor training restores walking in a nonambulatory child with chronic, severe, incomplete cervical spinal cord injury. Phys Ther. 2008;88:508–590.
9. Craik R, Herman R, Finley R. Human solutions for locomotion: interlimb coordination. In:Herman RM, Grillner S, Stein PSG, Stuart DG, eds. Neural Control of Locomotion. New York, NY: Plenum; 1976: 51–64.
10. Wagenaar R, van Emmerik RE. Resonant frequencies of arms and legs identify different walking patterns. J Biomech. 2000;33:853–861.
11. Eke-Okoro ST, Gregoric M, Larsson LE. Alterations in gait resulting from deliberate changes of arm-swing amplitude and phase. Clin Biomech. 1997;12:516–521.
12. Ford MP, Wagenaar RC, Newell KM. Arm constraint and walking in healthy adults. Gait Posture. 2007;26:135–141.
13. Thorpe DE, Dusing SC, Moore CG. Repeatability of temporospatial gait measures in children using the GAITRite electronic walkway. Arch Phys Med Rehabil. 2005;86(12);2342–2346.
14. Webster KE, Wittwer JE, Feller JA. Validity of the GAITRite walkway system for the measurement of averaged and individual step parameters of gait. Gait Posture. 2005;22:317–321.
15. Menant J, Steele J, Menz H, Munro B, Lord S. Effects of walking surfaces and footwear on temporo-spatial gait parameters in young and older people. Gait Posture. 2009;29:392–397.
16. Held S, Kott K, Young B. Standardized Walking Obstacle Course (SWOC): reliability and validity of a new functional measurement tool for children. Pediatr Phys Ther. 2006;18(1):23–30.
17. Taylor MJ, Gunter J. Standardized Walking Obstacle Course: preliminary reliability and validity of a functional measurement tool. J Rehabil Outcomes Meas. 1998;2(1):15–25.
18. Franjoine MS, Gunther JS, Taylor MJ. Pediatric balance scale: a modified version of the berg balance scale for the school-age child with mild to moderate motor impairment. Pediatr Phys Ther. 2003;15(2):114–120.
19. Gentile A. A working model of skill acquisition with application to teaching. Quest. 1972;17:3–23.

cerebral palsy/rehabilitation; child, child/preschool; exercise movement techniques; gait hemiplegia; motor skills, paresis/rehabilitation; physical therapy modalities/methods, postural balance

© 2012 Lippincott Williams & Wilkins, Inc.