Efficacy of Constraint-Induced Movement Therapy in Early Stroke Rehabilitation

Abstract

Background. There is limited evidence for the effects of constraint-induced movement therapy (CIMT) in the early stages of stroke recovery. Objective. To evaluate the effect of a modified CIMT within 4 weeks poststroke. Methods. This single-blinded randomized multisite trial investigated the effects of CIMT in 47 individuals who had experienced a stroke in the preceding 26 days. Patients were allocated to a CIMT or a usual care (control) group. The CIMT program was 3 h/d over 10 consecutive working days, with mitt use on the unaffected arm for up to 90% of waking hours. The follow-up time was 6 months. The primary outcome was the Wolf Motor Function test (WMFT) score. Secondary outcomes were the Fugl-Meyer upper-extremity motor score, Nine-Hole Peg test (NHPT) score, the arm use ratio, and the Stroke Impact Scale. Analyses of covariance with adjustment for baseline values were used to assess differences between the groups. Results. After treatment, the mean timed WMFT score was significantly better in the CIMT group compared with the control group. Moreover, posttreatment dexterity, as tested with the NHPT, was significantly better in the CIMT group, whereas the other test results were similar in both the groups. At the 6-month follow-up, the 2 groups showed no significant difference in arm impairment, function, or use in daily activities. Conclusions. Despite a favorable effect of CIMT on timed movement measures immediately after treatment, significant effects were not found after 6 months.

Keywords

stroke constraint-induced movement therapy rehabilitation

Introduction

Constraint-induced movement therapy (CIMT) is designed to improve upper-extremity motor function after stroke and consists of 3 key components: (1) repetitive, task-oriented training; (2) adherence-enhancing behavioral strategies (transfer package); and (3) constraining the use of the less-affected arm, usually by wearing a mitt.^1,2 The original protocol was developed for patients with chronic stroke and included 10 days of therapy for 6 h/d and constraining the less-affected arm during 90% of the time awake.³

There is a critical window for neuroplasticity and ability to relearn impaired activities within the first weeks after stroke.^4,5 Modified forms of CIMT have already been tested in the early stages of stroke rehabilitation (<10 weeks).⁵ In 2000, Dromerick et al⁶ published the results from a small-scale trial of 23 patients and reported that a 2-hour/10-day CIMT program was associated with less arm impairment at the end of treatment. Another study of 23 patients by Boake et al⁷ reported trends favoring a 3-hour/10-day CIMT program over standard therapy of equal duration. However, their only significant finding was improved Fugl-Meyer assessment scores immediately after treatment; no long-term effects were found. In the VECTORS study,⁸ 3 groups of stroke patients received 10 days of treatment within 4 weeks after stroke. The low-intensity CIMT group (2 h/d, 6-hour constraint) had significantly better improvement in the Action Research Arm test (ARAT), whereas the high-intensity CIMT group (3 h/d, 90% constraint) scored significantly worse than the control group. Yet another modification of CIMT was reported by Page et al,⁹ who reported increased use of the affected arm and improved motor impairment after 0.5 hours of training, 3 d/wk for 10 weeks of CIMT in a sample of 10 patients. In summary, the existing evidence on the effect of CIMT in the early stages of rehabilitation is limited to 5 trials that included 64 CIMT patients and 41 controls.¹⁰ These 5 trials included at least 3 different protocols and reported wide confidence intervals and large variations in treatment effects. Of particular concern are the negative results of the VECTORS trial.⁸ Because of the limited data and diversity of the results between previous trials, the effect of CIMT in the early stroke rehabilitation is still uncertain.

The aim of the present study was to assess the effect of a modified CIMT protocol in the early phase of rehabilitation after stroke. The primary hypothesis was that patients who completed a modified CIMT protocol in the early phase after stroke would have better arm motor function measured with the Wolf Motor Function test (WMFT) 6 months after the intervention compared with patients who received the usual care. The secondary aims were to evaluate the effect of CIMT on arm motor impairment, dexterity, arm use in daily activities, and overall health status after stroke.

Methods

Design and Participants

The Norwegian CI therapy multisite trial (NORCIMT) is a single-blinded, multicenter, randomized controlled trial that recruited participants from 5 Norwegian hospitals: The University Hospital of North Norway, The Trondheim University Hospital, Oslo University Hospital, Vestfold Hospital, and Telemark Hospital. Patients were recruited from October 2008 until June 2012. The inclusion criteria were (1) first-ever or recurrent stroke at >5 and <26 days prior, (2) modified Rankin Scale between 0 and 2 prior to the stroke, (3) persistent unilateral arm or hand paresis (Scandinavian Stroke Scale items: arm function 2-5 or hand motor function 2-4), (4) able to lift 2 fingers with the forearm pronated on the table or able to extend the wrist at least 10° from a fully flexed position, (5) able to follow a 2-step command, and (6) had Mini Mental State Examination score >20 (or >16 in combination with expressive aphasia). Patients were diagnosed according to the World Health Organization (WHO) definition¹¹ of stroke. Patients with recurrent stroke were included if they had no detectable arm weakness after the first stroke. Exclusion criteria were (1) Modified Rankin Scale >4, (2) unable to give informed consent, (3) >2 cm deviation on the line bisection test, (4) life expectancy <1 year, (5) injury or condition in the affected upper extremity that limited use prior to the stroke, or (6) other neurological conditions affecting motor function. After providing consent, the patients were allocated to either the CIMT or the control group by a computer-generated block scheme for randomization. The randomization procedure was stratified in 3 groups according to treatment center only. Two treatment centers were collated because of a low planned number of participants. The Regional Committee of Medical Ethics and the Commission of privacy rights at the University Hospital of North Norway reviewed and approved the study (REK NORD 39/2008).

Interventions

CIMT group participants were admitted to the rehabilitation department of 1 of the 4 treatment sites as soon as they were enrolled in the study. They received CIMT for 10 consecutive workdays. The CIMT focused on repetitive task-oriented training (2.0 and 0.5 hours of shaping and task practice, respectively, per treatment day) and adherence-enhancing behavioral strategies (0.5 hours per treatment day). The therapists recorded the treatment time and the time spent on each activity. In addition, the patients were encouraged to wear a constraining mitt on the more affected arm for up to 90% of their waking hours. The behavioral strategies included a treatment contract and daily use of the motor activity log, home diary, and home skill assignments.² The home skill assignments were prescribed for 0.5 h/d. The therapists completed a 4-day training program on the study procedures before they could administer the treatment.

The control group was treated according to the Norwegian guidelines for stroke patient treatment.¹² All patients took part in the multidisciplinary treatment approaches in their respective stroke units. Patients suitable for inpatient rehabilitation service were admitted to units. The rest were discharged home or to community-based rehabilitation.

Outcome Measures

All participants were examined by a blinded examiner at their respective treatment site after inclusion but before randomization, after the 2-week treatment period, and after 6 months. The primary outcome measure, the WMFT, is a 17-item test of arm motor function. It consists of 15 movement tasks and 2 strength measures. Each movement task was filmed, and the time to complete a task was determined by analyzing each frame in the video. The maximum time to complete an item was 120 s. If a trial was not completed, the trial result was recorded as 121 s. The median time for all 15 tasks was used for analysis.^13,14 In addition, the movement quality was rated from the videos using a 6-point functional ability scale (0 = does not attempt; 5 = normal movement). The video analyses were assigned to 1 of 2 specially trained raters at the end of the project period to ensure the reliability of time and quality measures. The test has previously been shown to be valid and reliable in stroke populations.^13,15-19

Secondary outcome measures were the Fugl-Meyer upper-extremity motor assessment (FMA upper-extremity),²⁰ the Nine-Hole Peg test (NHPT), the arm use ratio,²¹ and the Stroke Impact Scale. The FMA was used to measure motor impairment.²⁰ The upper-extremity motor part includes 33 items scored on a 3-point ordinal scale (0 = cannot perform; 1 = performs partially; 2 = performs fully), for a possible score from 0 to 66. Previous studies reported excellent interrater and intrarater reliability and good construct validity.²²

The NHPT is a dexterity measure for the upper extremities.²³ We used a wooden, 9-hole pegboard (Weston Home Health/Medical Equipment, West Sussex, UK) and previously described procedures.^24,25 A stopwatch was used to measure the time taken to place the 9 pegs. The maximum time allowed for each trial was increased from 50 to 120 s to reduce the floor effect.²⁵ The number of pegs per second for the more affected arm was calculated and used in the analysis. Adequate to excellent reliability of the NHPT has been reported in acute stroke patients.²⁴

Arm use/arm activity in daily activities was measured using ActiGraph GT1M or GT3X Accelerometers (ActiGraph Inc, Pensacola, FL). The patients were asked to wear 1 accelerometer on each arm for 24 hours. The arm use ratio calculated between the durations of movement in the more-affected and the least-affected arm was used as to measure actual arm use. Several investigations have validated this method in chronic phases and subacute phases (3-9 months) after stroke.^21,26,27 The measurement was adapted for acute stroke by using a 1-s epoch time.²⁸ The arm use ratio correlates well with the FMA and self-care dependence in acute stroke patients.²⁹

The Stroke Impact Scale is a self-report questionnaire with 64 items divided into 8 domains. In the present study, the domains of hand function (5 items), activities of daily living/instrumental activities of daily living (ADL/IADL, 10 items), participation/role function (8 items), and global perception of recovery (1 item) were reported because they were considered relevant for upper-extremity function. Each item is rated on a 5-point Likert scale. The scores in each domain are generated from an algorithm. Possible scores range from 0 to 100. The global perception of recovery was reported on a continuous scale from 0 to 100. The reliability of the included domains is reported as adequate to excellent, and the measures are valid and sensitive to change in stroke patients.³⁰

Statistical Analysis

SPSS for Windows 19.0 (SPSS Inc, Chicago, IL) was used for all statistical analyses. Means and standard deviations (SDs) were calculated for pretreatment, posttreatment, and the 6-month follow-up for each group. Changes within each group were evaluated with paired t tests. Statistical interferences between the groups were examined with analyses of covariance (ANCOVAs). The WMFT time variable was log transformed using the LG10 function to fit the assumptions of linearity, homoscedasticity, and homogeneity of variance. A per-protocol analysis was performed for all outcomes. The ANCOVA models for each dependent variable were then controlled for the influence of each of the cofactors: age, sex, prior stroke, National Institutes of Health Stroke Scale (NIHSS) total score, NHISS affected arm motor score, and days since stroke. Changes in the partial η² were used to assess confounding. The Stroke Impact Scale scores on 6-month follow-up were analyzed with independent sample t tests. Two-sided P values of .5 were considered significant for all predefined hypotheses. Intention-to-treat analyses were performed for the log-transformed WMFT time scores and the WMFT functional ability score. Each of the outcomes was examined separately with a multiple imputation technique. Pretest and posttest scores of the outcome measure were entered as predictors. A total of 5 new data sets were imputed; 5 imputations is seen to be necessary to avoid underestimation of the standard errors (SEs).³¹ ANCOVA procedures were first run on each data set, and pooled baseline adjusted means were then calculated across all 5 data sets.

Power Analysis

Pretrial power calculations were performed to determine the likelihood of detecting a true difference between the treatment and control groups on the primary outcome measures at the 6-month examination. We used the SD (logWMFT = 1.02) treatment effects (logWMFT = 0.55) from the EXCITE study treatment group (P. A. Thompson and S. L. Wolf, Means and standard deviations from the EXCITE trial, personal communication, 2008)³² for the power calculations. A power of 0.8 would require 53 participants in each group. After adding a fourth treatment center and extending the trial for 2.5 years, the study was halted before the required 106 patients were included. To evaluate the effect of the low inclusion rate, we calculated the observed power of the ANCOVA analyses in the actual sample using the standard SPSS subcommand.

Results

In total, 47 participants (11 women and 36 men) took part in this study. Overall, 96% had experienced an ischemic stroke, and 22% had had a prior stroke. The CIMT group included 24 participants, and 23 were allocated to the control group. Table 1 shows the participants’ characteristics. There were no significant differences between the groups for any of the variables at baseline. Figure 1 shows the CONSORT (Consolidated Standards of Reporting Trials) diagram of participant recruitment. One of the patients in the CIMT group withdrew from the study after treatment day 2. Another patient declined further participation after 8 treatment days. As shown in Figure 1, a total of 7 of the randomized patients—4 from the CIMT group and 3 controls—dropped out of the study before the 6-month follow-up assessment. One patient did not complete 10 of the WMFT items on the 6-month follow-up examination because of shoulder pain, and we were unable to analyze the video from another patient. The drop-out rates were 17% and 13% in the CIMT and control groups, respectively.

Table 1.

Baseline Characteristics of Study Participants.

	Total Sample (n = 47)		CI Therapy Group (n = 24)		Control Group (n = 23)
Age (years), mean (SD)	63.2	11.9	65.3	8.0	61.0	14.8
Females, n (%)	11	23%	5	21%	6	26%
Days poststroke, mean (SD)	17.3	6.9	16.6	7.2	18.0	6.5
Range		7-32		7-32		7-29
NIHSS, mean (SD)	1.7	1.8	1.7	1.9	1.8	1.8
NIHSS affected arm
0	32	68%	16	67%	16	70%
1	11	23%	5	21%	6	26%
2	3	6%	2	8%	1	4%
3	1	2%	1	4%	0	0%
4	0	0%	0	0%	0	0%
Modified Rankin Scale, mean (SD)	2.6	0.8	2.5	0.7	2.7	0.9
Affected side, right, n (%)	22	47%	10	42%	12	52%
Dominant side affected, n (%)	26	57%	16	67%	10	45%
Ischemic stroke, n (%)	43	96%	23	96%	20	95%
Prior stroke, n (%)	10	22%	6	25%	4	19%
Days in hospital, mean (SD)	36.9	16.3	38.7	14.1	35.0	18.5

Abbreviations: CI, constraint-induced; SD, standard deviation; NIHSS, National Institutes of Health Stroke Scale.

Figure 1.

Flow diagram of eligibility and the inclusion in the Norwegian CI-Therapy multisite trial.

Of the 22 CIMT participants available for posttreatment assessment, 19 completed all 10 treatment days, 1 missed 3 days of treatment, and 2 participants missed 1 day of treatment. The mean time in CIMT treatment was 27.0 hours (SD = 2.8 hours). Among the participants, 20 (91%) were inpatients between the inclusion and posttreatment assessments. In comparison, 15 (68%) of the participants in the control group were inpatients between the pretreatment and posttreatment assessments (P = .062). There were no significant differences in the total time in hospital between the groups. While in hospital, the control group attended a mean number of 8.8 (SD = 5.4) physical therapy sessions and 4.6 (SD = 3.7) occupational therapy sessions. In comparison, those in the CIMT group attended a mean number of 5 (SD = 3.9) physical therapy and 2.6 (SD = 3.0) occupational therapy sessions in addition to the CIMT treatment. The 2 groups had equal logWMFT time and functional ability scores at baseline, and there were no differences between the groups for the primary outcome at the 6-month follow-up. However, there was a significantly greater improvement in the CIMT group compared with controls for the logWMFT time variable at the posttreatment assessment. After the intention-to-treat analyses, the pooled baseline adjusted mean in the CIMT group was 0.31 (SE = 0.04) compared with 0.45 (SE = 0.04) in the control group. The differences between groups were significant in all 5 imputed data sets (P = .005-.028). The 3-s difference between the groups shown in Table 2 was above the 1.5- to 2-s limit considered a clinically important difference.³³ The pooled baseline adjusted means of the logWMFT time at the 6-month follow-up were 0.21 (SE = 0.03) in the CIMT group and 0.28 (SE = 0.03) in the control group (P = .032 to .198). The posttreatment pooled baseline adjusted means for the WMFT functional ability score were 4.1 (SE = 0.1) and 3.9 (SE = 0.01) in the CIMT and control groups, respectively. At the 6-month follow-up, the results were 4.4 (SE = 0.1) and 4.3 (SE = 0.1). The per-protocol analysis revealed a similar pattern (Table 2). None of the cofactors (age, sex, prior stroke, NIHSS total score, NIHSS affected arm score, days poststroke) interfered with this result. The observed power for detecting the actual differences between posttreatment results in the actual sample was 0.68 for the logWMFT time variable and 0.30 for the WMFT-Functional ability variable.

Table 2.

Per Protocol Analysis of Primary and Secondary Motor Outcomes.^a

	CI Therapy Group			Control Group			ANCOVA
	n	Mean	SD	n	Mean	SD	F	P
WMFT time (s)
Baseline	24	13.8	30.0	23	13.7	33.9
After treatment	22	2.3	1.1	22	5.2	11.3
6-Month follow-up	20	1.8	0.8	20	2.0	1.0
logWMFT time
Baseline	24	0.69	0.49	23	0.62	0.50
After treatment	22	0.32	0.19	22	0.43	0.38	6.112	.018
6-Month follow-up	20	0.22	0.18	20	0.27	0.20	1.636	.209
WMFT functional ability
Baseline	24	3.3	0.8	23	3.4	0.8
After treatment	22	4.1	0.7	22	3.9	0.8	2.170	.148
6-Month follow-up	21	4.4	0.8	20	4.3	0.7	0.452	.506
WMFT arm strength (kg)
Baseline	24	5.0	3.1	22	5.0	3.0
After treatment	22	6.9	3.9	21	6.0	2.5	1.930	.172
6-Month follow-up	18	7.9	2.2	17	7.2	2.7	1.270	.268
WMFT grip strength (kg)
Baseline	23	19.0	11.4	21	15.1	11.1
After treatment	21	24.1	11.4	20	17.8	11.2	2.539	.119
6-Month follow-up	20	29.0	14.5	19	25.1	12.7	0.008	.931
Fugl-Meyer upper extremity
Baseline	24	49.4	11.8	23	50.1	9.4
After treatment	22	56.9	10.0	22	55.1	7.1	2.583	.116
6-Month follow-up	21	61.6	5.0	21	60.0	8.8	1.120	.296
Nine-Hole Peg test (peg/s)
Baseline	24	0.16	0.16	23	0.16	0.13
After treatment	22	0.29	0.18	22	0.22	0.16	4.781	.035
6-Month follow-up	21	0.40	0.19	21	0.38	0.16	0.229	.635
Arm use ratio
Baseline	22	0.73	0.27	22	0.63	0.22
After treatment	18	0.84	0.19	18	0.73	0.16	1.103	.301
6-Month follow-up	19	0.87	0.18	17	0.90	0.18	1.595	.215

Abbreviations: CI, constraint-induced; SD, standard deviation; WMFT: Wolf motor function test.

F and P values are adjusted for baseline values. The ANCOVAs include only participants with valid baseline values. There were missing data on the strength tests and arm use ratio that reduced the number of participants available for analysis.

Among the secondary outcomes, all motor variables improved significantly in both groups from baseline to posttreatment and from posttreatment to the 6-month follow-up. There was a significant posttreatment difference for the NHPT (partial η² = 0.104; P = .035). This difference was robust for all cofactors except for days poststroke (partial η² = 0.085; P = .064). There were no significant differences between the 2 groups for WMFT shoulder strength, grip strength, FMA upper extremity, arm use ratio, or Stroke Impact Scale. Table 3 shows that SIS scores were not significantly different between groups at the 6-month follow-up. Regarding adverse symptoms, one patient in the CIMT group developed a shoulder capsulitis in the time span between posttreatment and the 6-month follow-up examination, and this patient was not able to complete the WMFT test at follow-up. Another CIMT patient developed shoulder pain during the treatment. No other adverse effects were observed in the 6-month time span.

Table 3.

Stroke Impact Scale (SIS) at the 6-Month Follow-Up.

	CI Therapy Group			Control Group			t Test (P Value)
	n	Mean	SD	n	Mean	SD	t Test (P Value)
SIS: Hand	20	85.0	15.7	19	79.2	26.2	.405
SIS: ADL/IADL	19	91.9	7.3	19	89.9	15.7	.623
SIS: Participation	16	81.9	17.7	11	81.3	11.9	.919
SIS: Overall recovery	19	70.4	20.1	19	73.7	14.7	.571

Abbreviations: CI, constraint-induced; SD, standard deviation; ADL/IADL, activities of daily living/instrumental activities of daily living.

Discussion

The main result from the present trial was that CIMT initiated within 1 month after stroke was safe and feasible but did not improve long-term motor function. However, that we observed significantly better WMFT scores in the CIMT group posttreatment indicates that CIMT might promote faster recovery compared with standard care. The 2.9-s difference in WMFT and the 0.7-peg/s difference in NHPT indicates a moderate short-term effect of CIMT on arm motor function and dexterity. There were no differences between the groups in the ability to reduce arm motor impairment or increase arm use.

An obvious limitation of this study is that it was halted early with far fewer participants than recommended from the pretrial power calculations. Despite the low power, the differences between the groups were rather small, and a larger sample might not reveal results of clinical significance. The sizes of the treatment arms in the present study are also larger than in comparable studies,^6-8 and the trial, therefore, provides important new data on CIMT in early stroke rehabilitation. Another limitation is that there was no periodic reevaluation or training of the examiners. The trial was extended for 2.5 years more than planned, and this might have introduced bias in the measurements. The drop-out rate was also relatively high, and this represents a risk of an effect-dependent bias in the results. The nature of the control group has to be considered when the effects are interpreted. The control group in this trial was treated as usual and did not follow a predefined protocol. All control participants were treated in multidisciplinary stroke units, and the majority of both groups were inpatients in the time between inclusion and the first posttreatment assessment. However, we were not able to get reliable data about the actual time spent in active upper-limb training in the control group. Therefore, we are not able to control the effects for dosage of therapy. A large proportion of the included participants had mild to moderate strokes. This caused a ceiling effect in most of the outcome measures that might have influenced the results, especially at the 6-month follow-up. The trial included mostly men. Although sex did not appear to affect the outcome, the results might not be applicable to women. In the CIMT group, 16 (67%) of the participants were affected on the dominant side, and the results might be less relevant for stroke patients with an affected nondominant arm. Unfortunately, technical problems with accelerometer logistics reduced the data for this secondary outcome.

The findings of the present study do not demonstrate the same trend toward less impairment after CIMT as shown by Ro et al³⁴ and Boake et al.⁷ On the other hand, the results are in line with those described in a meta-analysis by Nijland et al,¹⁰ who concluded that there was no significant improvement following high-intensity CIMT initiated within 1 month after stroke. Another important finding was that, despite 2 cases of shoulder pain, there was no unfavorable development in motor function in the CIMT group compared with the usual care group. This is in contrast with the results of the VECTORS study,⁸ where the high-intensity CIMT group (3 h/d, 10 days) scored significantly worse on arm motor function (ARAT) than the low-intensity CIMT and ADL practice/bilateral training groups. The ARAT and FMA are strongly correlated,^35,36 and the difference is probably not attributable to the different measurements. Possible explanations could be that different components of the CIMT treatment, like greater levels of task practice in the VECTORS study, accounted for differences in component compliance. Task practice usually leads to more continuous work during the treatment sessions, with fewer breaks, compared with shaping exercises.³⁷ The time from stroke to inclusion was also lower in the VECTOR study (9.7 days) compared with ours (17.3 days). This difference might contribute significantly to the differences in results. In addition, our control intervention was not standardized, whereas Dromerick et al⁸ studied an experimental control group that received standardized treatment. The 2-hour ADL practice and bilateral training used in the VECTORS study might have been more effective than the protocol used here. Our results showed a considerably weaker effect than the study published by Page et al,⁹ which might be attributable to the different CIMT protocol in their study (0.5 h/d, 30 days over 10 weeks). The difference might also be related to the fact that learned nonuse is developed over time and that the lower-intensity, longer-duration treatment is more effective than the 2-week program used in our trial in the early rehabilitation. Another possible reason is that Page et al⁹ only included patients with a motor activity log score <2.5. The motor activity log is highly sensitive to nonuse and including patients with low scores and equal motor impairment would result in a sample of participants with more pronounced nonuse that might respond better to the treatment. The differences between our study and others might also be attributed to another difference in inclusion criteria. Previous trials^7,8,34 excluded participants without detectable shoulder weakness (NIHSS arm score <1). Among our participants, 32 had a NIHSS arm motor score of 0, and the FMA values of our participants indicate less impairment than that of participants included in the other studies.^7,8,34 The effects of CIMT might change according to the level of upper-extremity impairment.

Implications

The CIMT protocol used in this trial can be safely applied within 28 days poststroke and might increase movement speed in the early stage. Future research should focus on the variability between treatment protocols, optimal therapy time, and the reasons for the large variations in results within 4 weeks poststroke. At the moment, application of CIMT in early stroke rehabilitation is not warranted because of limited evidence of lasting effects.

Footnotes

Acknowledgements

We would like to thank Martin Vatshaug for collating and ensuring the quality of data, Gunn Marte Vadset and Eli Langørgen for taking part in the design of the treatment protocol, and Grete Jakobsen for teaching the study protocol and acting as a site administrator. We would also like to thank Paul A. Thompson and Steven Wolf for providing the data for the power calculation and Tom Wilsgaard for statistical advice. The study was supported by the Northern Norway Regional Health Authority (SAT 544-06). The authors declare no conflict of interests.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Authors’ Note

Clinical trial registration number (): NCT00906477.

References

Mark

Taub

. Constraint-induced movement therapy for chronic stroke hemiparesis and other disabilities. Restor Neurol Neurosci. 2004;22:317-336.

Morris

Taub

Mark

. Constraint-induced movement therapy: characterizing the intervention protocol. Eura Medicophys. 2006;42:257-268.

Taub

Miller

Novack

. Technique to improve chronic motor deficit after stroke. Arch Phys Med Rehabil. 1993;74:347-354.

Dobkin

. Neurobiology of rehabilitation. Ann N Y Acad Sci. 2004;1038:148-170.

Kwakkel

Kollen

Twisk

. Impact of time on improvement of outcome after stroke. Stroke. 2006;37:2348-2353.

Dromerick

Edwards

Hahn

. Does the application of constraint-induced movement therapy during acute rehabilitation reduce arm impairment after ischemic stroke? Stroke. 2000;31:2984-2988.

Boake

Noser

. Constraint-induced movement therapy during early stroke rehabilitation. Neurorehabil Neural Repair. 2007;21:14-24.

Dromerick

AWM

Lang

CEP

Birkenmeier

RLM

. Very Early Constraint-Induced Movement during Stroke Rehabilitation (VECTORS): a single-center RCT. Neurology. 2009;73:195-201.

Page

Levine

Leonard

. Modified constraint-induced therapy in acute stroke: a randomized controlled pilot study. Neurorehabil Neural Repair. 2005;19:27-32.

10.

Nijland

Kwakkel

Bakers

van Wegen

. Constraint-induced movement therapy for the upper paretic limb in acute or sub-acute stroke: a systematic review. Int J Stroke. 2011;6:425-433.

11.

Hatano

. Experience from a multicentre stroke register: a preliminary report. Bull World Health Organ. 1976;54:541-553.

12.

Indredavik

Salvesen

Ness

Thorsvik

. National Guidelines: Treatment and Rehabilitation After Stroke. Oslo, Norway: The Norwegian Directorate of Health; 2010.

13.

Morris

Uswatte

Crago

Cook

III Taub

. The reliability of the wolf motor function test for assessing upper extremity function after stroke. Arch Phys Med Rehabil. 2001;82:750-755.

14.

Askim

Dahl

Stock

Langøren

. Nevrologisk test for funksjon i arm og hånd: Wolf Motor Function test på norsk. Fysioterapeuten. 2009;76.

15.

Wolf

Thompson

Morris

. The EXCITE trial: attributes of the Wolf Motor Function test in patients with subacute stroke. Neurorehabil Neural Repair. 2005;19:194-205.

16.

Nijland

van Wegen

Verbunt

van Wijk

van Kordelaar

Kwakkel

. A comparison of two validated tests for upper limb function after stroke: the Wolf Motor Function test and the Action Research Arm test. J Rehabil Med. 2010;42:694-696.

17.

Whitall

Savin

Jr Harris-Love

Waller

. Psychometric properties of a modified Wolf Motor Function test for people with mild and moderate upper-extremity hemiparesis. Arch Phys Med Rehabil. 2006;87:656-660.

18.

Wolf

Catlin

Ellis

Archer

Morgan

Piacentino

. Assessing Wolf Motor Function test as outcome measure for research in patients after stroke. Stroke. 2001;32:1635-1639.

19.

Ang

Man

. The discriminative power of the Wolf Motor Function test in assessing upper extremity functions in persons with stroke. Int J Rehabil Res. 2006;29:357-361.

20.

Fugl-Meyer

Jaasko

Leyman

Olsson

Steglind

. The post-stroke hemiplegic patient: 1. A method for evaluation of physical performance. Scand J Rehabil Med. 1975;7:13-31.

21.

Uswatte

Foo

Olmstead

Lopez

Holand

Simms

. Ambulatory monitoring of arm movement using accelerometry: an objective measure of upper-extremity rehabilitation in persons with chronic stroke. Arch Phys Med Rehabil. 2005;86:1498-1501.

22.

Gladstone

Danells

Black

. The Fugl-Meyer assessment of motor recovery after stroke: a critical review of its measurement properties. Neurorehabil Neural Repair. 2002;16:232-240.

23.

Mathiowetz

Weber

Kashman

Volland

. Adult norms for the 9 Hole Peg test of finger dexterity. Occup Ther Res. 1985;5:24-38.

24.

Heller

Wade

Wood

Sunderland

Hewer

Ward

. Arm function after stroke: measurement and recovery over the first three months. J Neurol Neurosurg Psychiatry. 1987;50:714-719.

25.

Sunderland

Tinson

Bradley

Hewer

. Arm function after stroke: an evaluation of grip strength as a measure of recovery and a prognostic indicator. J Neurol Neurosurg Psychiatry. 1989;52:1267-1272.

26.

Uswatte

Miltner

Foo

Varma

Moran

Taub

. Objective measurement of functional upper-extremity movement using accelerometer recordings transformed with a threshold filter. Stroke. 2000;31:662-667.

27.

Uswatte

Giuliani

Winstein

Zeringue

Hobbs

Wolf

. Validity of accelerometry for monitoring real-world arm activity in patients with subacute stroke: evidence from the extremity constraint-induced therapy evaluation trial. Arch Phys Med Rehabil. 2006;87:1340-1345.

28.

Lang

Wagner

Edwards

Dromerick

. Upper extremity use in people with hemiparesis in the first few weeks after stroke. J Neurol Phys Ther. 2007;31:56-63.

29.

Thrane

Emaus

Askim

Anke

. Arm use in patients with subacute stroke monitored by accelerometry: association with motor impairment and influence on self-dependence. J Rehabil Med. 2011;43:299-304.

30.

Duncan

Wallace

Lai

Johnson

Embretson

Laster

. The stroke impact scale version 2.0: evaluation of reliability, validity, and sensitivity to change. Stroke. 1999;30:2131-2140.

31.

Gadbury

Coffey

Allison

. Modern statistical methods for handling missing repeated measurements in obesity trial data: beyond LOCF. Obes Rev. 2003;4:175-184.

32.

Wolf

Winstein

Miller

. 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.

33.

Lin

Hsieh

Chen

Jang

Liu

. Minimal detectable change and clinically important difference of the Wolf Motor Function test in stroke patients. Neurorehabil Neural Repair. 2009;23:429-434.

34.

Noser

Boake

. Functional reorganization and recovery after constraint-induced movement therapy in subacute stroke: case reports. Neurocase. 2006;12:50-60.

35.

Platz

Pinkowski

van Wijck

Kim

di Bella

Johnson

. Reliability and validity of arm function assessment with standardized guidelines for the Fugl-Meyer test, Action Research Arm test and Box and Block test: a multicentre study. Clin Rehabil. 2005;19:404-411.

36.

De Weerdt

WJG

. Measuring recovery of arm-hand function in stroke patients: a comparison of the Brunnstrom-Fugl-Meyer test and the Action Research Arm test. Physiother Can. 1985;37:65-70.

37.

Winstein

Miller

Blanton

. 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.