Bilateral Priming Accelerates Recovery of Upper Limb Function After Stroke
A Randomized Controlled Trial
Background and Purpose—The ability to live independently after stroke depends on the recovery of upper limb function. We hypothesized that bilateral priming with active–passive movements before upper limb physiotherapy would promote rebalancing of corticomotor excitability and would accelerate upper limb recovery at the subacute stage.
Methods—A single-center randomized controlled trial of bilateral priming was conducted with 57 patients randomized at the subacute stage after first-ever ischemic stroke. The PRIMED group made device-assisted mirror symmetrical bimanual movements before upper limb physiotherapy, every weekday for 4 weeks. The CONTROL group was given intermittent cutaneous electric stimulation of the paretic forearm before physiotherapy. Assessments were made at baseline, 6, 12, and 26 weeks. The primary end point was the proportion of patients who reached their plateau for upper limb function at 12 weeks, measured with the Action Research Arm Test.
Results—Odds ratios indicated that PRIMED participants were 3× more likely than controls to reach their recovery plateau by 12 weeks. Intention-to-treat and per-protocol analyses showed a greater proportion of PRIMED participants achieved their plateau by 12 weeks (intention to treat, χ2=4.25; P=0.039 and per protocol, χ2=3.99; P=0.046). ANOVA of per-protocol data showed PRIMED participants had greater rebalancing of corticomotor excitability than controls at 12 and 26 weeks and interhemispheric inhibition at 26 weeks (all P<0.05).
Conclusions—Bilateral priming accelerated recovery of upper limb function in the initial weeks after stroke.
Clinical Trial Registration—URL: http://www.anzctr.org.au. Unique identifier: ANZCTR1260900046822.
- motor evoked potentials
- neuronal plasticity
- physical therapy techniques
- transcranial magnetic stimulation, single pulse
- upper extremity
The ability to live independently after stroke depends on the recovery of motor function, particularly of the upper limb.1 The potential for recovery is related to the extent of cerebral damage that creates a ceiling effect,2,3 with a plateau usually reached within 6 months after stroke.4 There are no treatments available that can repair the stroke lesion and raise the recovery ceiling. An alternative strategy could be the development of adjuvant techniques that accelerate recovery and help patients more efficiently reach a plateau of best possible function.
Techniques that prime the brain for a more plastic response to therapy may accelerate motor recovery after stroke. Increasing excitability and reducing inhibition are important precursors for neural plasticity,5 which may allow surviving neural elements to more easily reorganize in response to therapy.6 Active–passive bilateral priming (APBP) is a pattern of coordinated movement that disinhibits the M1 contralateral to the assisted (paretic) limb7 and facilitates its excitability for ≥30 minutes after a 15-minute session.8 In patients with ≥6 months after stroke, daily APBP followed by motor practice led to increased ipsilesional corticomotor excitability and a greater improvement in upper limb function compared with motor practice alone.9 The aim of this study was to determine the immediate and longer term effects of bilateral priming with patients with stroke at the subacute stage. We hypothesized that APBP before upper limb therapy would accelerate the recovery of hand and arm function, with a greater proportion of PRIMED participants reaching maximum recovery by 12 weeks.
Consecutive patients aged ≥18 years admitted with first-ever monohemispheric ischemic stroke were screened between November 2009 and March 2012 (Figure 1). Patients were excluded if they did not require upper limb rehabilitation or were unsuitable because of spasticity, homonymous hemianopia, blindness, visuospatial neglect, or complete somatosensory loss. Patients were also excluded if they were unsuitable for research participation because of contraindications to transcranial magnetic stimulation or MRI and cognitive or communication impairment or pre-existing conditions precluding informed consent or compliance with study assessments. The study was approved by the regional ethics committee, and all participants provided written informed consent in accordance with the Declaration of Helsinki.
Participants completed baseline assessments with the Action Research Arm Test (ARAT),10 the Fugl–Meyer scale, and the National Institutes of Health Stroke Scale to evaluate baseline upper limb function, impairment, and stroke severity, respectively. Clinical assessments were repeated after the 4-week intervention and at 12 and 26 weeks after stroke. The modified Rankin Scale and Stroke Impact Scale were used to evaluate disability and quality of life at 26 weeks. Participants were assessed by therapists who were unaware of their group allocation and did not treat them.
Participants were randomized and began the intervention within 26 days of stroke. Intervention allocation was concealed and randomized using customized software (www.rando.la) that minimized between-group differences in age, baseline ARAT score, PREP stratification,2 and brain-derived neurotrophic factor genotype derived from a single baseline blood sample because this may influence plasticity and learning11 (Table 1). To stratify patients using the PREP algorithm, we first graded shoulder abduction and finger extension strength of the paretic upper limb 72 hours after stroke. Next, transcranial magnetic stimulation was used to determine whether motor evoked potentials could be recorded from the extensor carpi radialis (ECR) muscle of the paretic upper limb. Finally, T1-weighted and diffusion-weighted images were acquired with a Siemens 1.5 T Avanto scanner (Methods in the online-only Data Supplement). The mean fractional anisotropy (FA) was calculated within the posterior limb of each internal capsule. The structural integrity of the posterior limbs of the internal capsules was quantified by calculating an asymmetry index from the mean FA values: FAAI=(FAcontra–FAipsi)/(FAcontra+FAipsi).12
Transcranial magnetic stimulation was used to evaluate corticomotor excitability in each hemisphere and interhemispheric inhibition. Motor evoked potentials were recorded from the ECR of each upper limb using standard surface electromyography techniques. Rest motor threshold was determined for each ECR and stimulus–response (S–R) curves constructed by recording blocks of 12 motor evoked potentials at intensities −5%, +5%, +15%, +25%, and +35% of maximum stimulator output relative to rest motor threshold, with the order of stimulus intensities randomized.
Interhemispheric inhibition was evaluated by delivering single magnetic stimuli at 80% maximum stimulator output to each M1 while participants maintained full voluntary extension of the ipsilateral wrist against gravity. In healthy adults, this protocol produces a period of silence in the electromyography of ipsilateral ECR that begins ≈30 ms after stimulation and lasts for 30 to 50 ms. The ipsilateral silent period reflects the excitability of interhemispheric pathways responsible for inhibition passed between the motor cortices13 (Methods in the online-only Data Supplement).
Participants allocated to the bilateral priming group (PRIMED) used a portable device to couple the 2 hands mechanically and produce rhythmic, continuous bimanual mirror symmetrical movements for 15 minutes. Participants actively flexed and extended the nonparetic wrist, with the device driving the paretic wrist in a mirror symmetrical pattern (Figure I in the online-only Data Supplement). The device confers an inertial advantage such that little force is required from the active wrist. Participants were instructed to move at a comfortable rate and were given a target of 500 to 1500 movement cycles, depending on their individual ability. The number of movement cycles completed each session was recorded from a digital counter on the device.
The CONTROL intervention (CONTROL) was intermittent cutaneous electric stimulation of the volar aspect of the paretic forearm, using a standard TENS unit delivered for 15 seconds (including 2-s ramp-up, 2-s ramp-down), once per minute, for 15 minutes. The intensity was adjusted to produce a mild cutaneous sensation for each participant and served only to control for participants’ expectations. The priming and control interventions were delivered immediately before upper limb therapy every weekday for 4 weeks. Participants continued self-directed priming and therapy at home if they were discharged from inpatient rehabilitation during the 4-week intervention period. Although participants could not be blinded to the priming, they had no reason to expect one technique was more effective.
Upper Limb Rehabilitation
A standardized dose of upper limb rehabilitation began within 4 weeks after stroke, consisting of 30 minutes of physiotherapy and occupational therapy delivered every weekday for 4 weeks. Therapy was delivered immediately after the completion of either the priming or control intervention by therapists who were unaware of group allocation. The amount of time spent priming and then in therapy was recorded for each session by therapists or by the participant using a therapy diary if they completed self-directed priming and therapy at home. Compliance was monitored with telephone calls and home visits, and diaries were returned at the end of the intervention for analysis. Participants also received standard care throughout the study.
Independent 2-sided t tests were used for linear continuous variables. Two-sided Pearson χ2 tests were used for nominal and ordinal variables, except when cell counts were <5, in which case 2-sided Fisher exact tests were used.
Primary End Point
Participants were binarized at 12 weeks according to whether they had reached their recovery plateau, defined as achieving ≥75% of their maximum recovery on the ARAT score or being within 1 point of their maximum ARAT score.14 This criterion was appropriate because the average recovery across all participants was 74% at this time. We tested whether a greater proportion of PRIMED than CONTROL participants achieved their plateau using 2-sided Pearson χ2 tests. Per-protocol and intention-to-treat analyses were performed, and odds ratios were calculated. ARAT scores from all randomized patients, with the last observation carried forward, were used for the intention-to-treat analyses. We powered the study based on these expectations: at least half the CONTROL participants would achieve plateau by 12 weeks and a between-group difference of 0.25 for the proportion of patients achieving their plateau by 12 weeks, α=0.05 and β=0.80. The required sample size was estimated to be ≤58 patients, depending on the absolute proportions achieved by each group.15
Secondary Clinical Measures
Between-group comparisons of modified Rankin Scale and Stroke Impact Scale scores at 26 weeks were made with a Mann–Whitney U test and an independent samples t test, respectively.
Bilateral priming was suitable for 292 (80%) of the 350 screened patients who required upper limb rehabilitation (Figure 1). Fifty-seven participants (26 men; mean age, 68 years; SD 25 years) were randomized; the 2 groups were well-matched at baseline (Table 1) and received balanced priming, therapy, and assessment (Table 2). Fifty-one of 57 (89%) participants completed the primary end point (Figure 1).
Primary End Point
A greater proportion of participants achieved a plateau of upper limb function by 12 weeks in the PRIMED group than the CONTROL group (Figure 2A) for both the intention-to-treat (23/29 PRIMED; 15/28 CONTROL; χ2=4.25; P=0.039) and per-protocol analyses (23/28 PRIMED; 13/23 CONTROL; χ2=3.99; P=0.046). PRIMED participants were ≈3× more likely to achieve their plateau within 12 weeks compared with CONTROL participants (intention-to-treat odds ratio, 3.32; 95% confidence interval, 1.1–10.7 and per-protocol odds ratio, 3.54; 95% confidence interval, 1.0–12.6).
Secondary Clinical Measures
At 26 weeks, there were no between-group differences in median modified Rankin Scale score (2; PRIMED range, 0–4; CONTROL range, 0–5; Mann–Whitney U test; P>0.4) or mean Stroke Impact Scale score (63.5; SE, 2.2; CONTROL, 65.4; SE, 2.4; t test P>0.2).
There were no differences between the PRIMED and CONTROL groups at baseline (t tests; all P>0.15; Table I in the online-only Data Supplement).
Bilateral priming promoted rebalancing of corticomotor excitability (Figure 2B and 2C). The S–R slope was used as a measure of corticomotor excitability in each hemisphere. There was a main effect of hemisphere and an interaction between hemisphere and time on S–R slope (Table I in the online-only Data Supplement). There was an interaction among group, hemisphere, and time, which was decomposed with RM ANOVA for each group. There were main effects of hemisphere for both groups, but only the PRIMED group had an interaction between hemisphere and time. This expected effect arose because ipsilesional slope increased, and contralesional slope decreased, over time in the PRIMED group but not in the CONTROL group.
Bilateral priming increased the excitability of transcallosal projections from the ipsilesional to contralesional M1. The excitability of these projections was indexed by the persistence and depth of ipsilateral silent periods produced in the ongoing electromyography recorded from the nonparetic ECR by stimulation of ipsilesional M1. As expected, both the persistence and depth of ipsilateral silent periods produced by stimulation of the ipsilesional M1 increased over time for the PRIMED group, but perhaps surprisingly decreased over time for the CONTROL group (Table I and Figure II in the online-only Data Supplement).
This is the first study to show that bilateral priming before upper limb therapy accelerates the recovery of upper limb function after stroke and increases the odds of reaching the recovery plateau by 12 weeks for equivalent therapy dose. Bilateral priming was suitable for 80% of patients with upper limb weakness and is feasible in the clinical setting. Recovery of motor function usually plateaus during the first 6 months after stroke, and prestroke levels of function are seldom restored.4 Conceptually, aiming to accelerate the rate of recovery during rehabilitation may be a more realistic goal than attempting to overcome the recovery ceiling.
The neurophysiological effects of APBP are the most likely mechanism underlying the observed acceleration of recovery. Excitability increased for both descending and transcallosal projections from the ipsilesional M1 in the PRIMED group but not in the CONTROL group. These effects were evident 12 and 26 weeks after stroke, indicating long-term benefits for the motor system, which may overcome a progressive neurophysiological decline (CONTROL ipsilateral silent period; Figure II in the online-only Data Supplement). Bilateral priming facilitates corticomotor excitability in the hemisphere contralateral to the assisted (paretic) upper limb for ≥30 minutes.8 This period of disinhibition7 may create a therapeutic window where plastic reorganization within the ipsilesional M1 is more likely to occur. It is unlikely that APBP constitutes an additional 15 minutes of therapy because it is not task-specific and can be completed without any active movement of the paretic upper limb. Furthermore, a 15-minute increase in daily upper limb therapy has no beneficial effects.16 Bilateral priming is a neuromodulatory adjuvant rather than a therapy and is, therefore, distinct from bilateral isokinematic training or bilateral arm training with auditory cueing,17,18 which have been found to have no additional benefits by a recent Cochrane review.19
This study has a number of potential limitations. First, the sample of 57 patients is relatively small. Sample size was limited by the neurophysiological and neuroimaging techniques used to stratify patients for research purposes, even though APBP was suitable for most patients. Second, although the duration of each therapy session was recorded, it is difficult to measure self-directed practice completed by patients during rehabilitation after stroke. This is a challenge for all rehabilitation trials and not unique to this study. Although participants are likely to have completed a wide range of total therapy doses, the randomization process makes systematic between-group differences unlikely.
This study also has a number of strengths. All participants began the intervention within 26 days of stroke, much earlier than the majority of motor rehabilitation randomized controlled trials.20 The groups’ clinical baseline characteristics were well balanced. Blinding was carefully maintained by having 3 groups of therapists who PRIMED, treated, and assessed participants. The successful PRIMED treatment of a heterogeneous sample of patients in a busy clinical setting indicates that bilateral priming is feasible in the real world, and that these findings are generalizable.
We thank Suzanne Ackerley, Yvette Baker, Claudia Barclay, Patricia Bennett, Jemma Crowe, Alison Elston, Marie-Claire Smith, Anne Ronaldson, and Anna Vette for their assistance with patient recruitment and study coordination. We thank the Centre for Advanced MRI and LabPlus for assistance with imaging and genotyping.
Sources of Funding
This study was funded by the Health Research Council of New Zealand (09/164 to Dr Byblow, Dr Stinear, S. Anwar, and P.A. Barber) and the Stroke Foundation (Northern Region) New Zealand (3623723 to P.A. Barber, Dr Stinear, and Dr Byblow).
Dr Stinear has been funded for travel by Pierre Fabre Pharmaceuticals and receives funding from Health Research Council of New Zealand (11/270). Dr Byblow is a named inventor on a patent for a training device assigned to Uniservices Ltd and receives funding from Health Research Council of New Zealand (11/270). The other authors report no conflicts.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.003537/-/DC1.
- Received September 17, 2013.
- Accepted October 1, 2013.
- © 2013 American Heart Association, Inc.
- Veerbeek JM,
- Kwakkel G,
- van Wegen EE,
- Ket JC,
- Heymans MW
- Stinear CM,
- Barber PA,
- Petoe M,
- Anwar S,
- Byblow WD
- Stinear CM,
- Barber PA,
- Coxon JP,
- Fleming MK,
- Byblow WD
- Stinear CM,
- Barber PA,
- Smale PR,
- Coxon JP,
- Fleming MK,
- Byblow WD
- 15.↵Sample size for 2 proportions tables. http://www.statstodo.com/SSiz2Props_Tab.php. Accessed September 14, 2013.
- Rodgers H,
- Mackintosh J,
- Price C,
- Wood R,
- McNamee P,
- Fearon T,
- et al
- Mudie MH,
- Matyas TA
- Whitall J,
- Waller SM,
- Sorkin JD,
- Forrester LW,
- Macko RF,
- Hanley DF,
- et al
- Coupar F,
- Pollock A,
- van Wijck F,
- Morris J,
- Langhorne P
- Stinear C,
- Ackerley S,
- Byblow W