Secondary Logo

Journal Logo

TRAUMA AND REHABILITATION: Edited by S. Thomas Carmichael

Predicting and accelerating motor recovery after stroke

Stinear, Cathy M.a,c; Byblow, Winston D.b,c

Author Information
doi: 10.1097/WCO.0000000000000153



Stroke results in motor impairment for most patients [1], and is a leading cause of long-term adult disability [2]. The ability to live independently after stroke critically depends on the recovery of motor function [3,4], particularly of the hand and arm [5]. This review is, therefore, focused on the recovery and rehabilitation of voluntary upper limb movement after stroke.

Box 1
Box 1:
no caption available


Large-scale observational studies indicate that recovery is most rapid during the first month after stroke, and motor function typically reaches a plateau within 3 months [1,6–8]. Even the most severely affected patients can expect no further improvements in activities of daily living beyond 6 months poststroke [9]. This is not to say that patients at the chronic stage cannot benefit from therapy, discussed below. However, improvements during the first few weeks after stroke map onto the nature and time course of the neurobiological mechanisms of spontaneous recovery (Fig. 1), which have been reasonably well characterized using animal models [11,12]. Rehabilitation is delivered during the first few weeks after stroke, to interact with these neurobiological processes and shape the patient's recovery. Better outcomes are observed when therapy is initiated earlier [13–15], possibly because this maximizes the opportunity for therapy to interact with the spontaneous recovery process.

Time course of plasticity mechanisms and neurological recovery after stroke. Most patients reach their motor recovery plateau within 3 months of stroke [6,9]. Cellular dysfunction (red) begins immediately after stroke, and is the target of neuroprotection treatments aimed at reducing oedema, inflammation, and apoptosis. Cell-based therapies target the subsequent mechanisms of cell repair and genesis (grey), which persist for several weeks following stroke [10]. Functional plasticity, in the form of altered neuronal excitability and synaptic efficacy, begins within hours of symptom onset (green), and can enhance use-dependent functional plasticity and relearning. This is closely followed by the initiation of structural plasticity processes in the form of dendritic spine remodelling, axonal sprouting and synaptogenesis, which are maximally active around 1 week after stroke, and plateau by 3–4 weeks (blue) [10].

A recent study evaluated the accuracy of experienced therapists’ predictions of the upper limb motor recovery plateau [16▪▪], measured with the Action Research Arm Test (ARAT). Therapists were asked at 72 h after stroke whether patients would be severely impaired (ARAT score <10), normal (ARAT score = 57), or somewhere in between (ARAT score 10–56) at 6 months after stroke. Predicting an ARAT score between 10 and 56 may not be clinically helpful, as these scores represent a very wide range of upper limb function. Overall, the therapists made accurate predictions for 60% of patients, but were no better than chance for predicting those who would achieve an ARAT score between 10 and 56. When predictions were made at discharge from the acute stroke unit, the accuracy rate increased to 72% overall, but remained lower (61%) for patients predicted to achieve an ARAT score between 10 and 56. The authors also used bivariate analyses of clinical measures to identify 11 candidate predictors of ARAT score at 6 months. These included type of stroke, visual inattention, gaze palsy, hemianopia, sensory loss, urinary incontinence, shoulder abduction and elevation, paretic leg strength, finger extension, and sitting balance. Subsequent multivariate modelling found that only shoulder abduction dichotomized with the motricity index (9 point cut-off) and finger extension dichotomized with the Fugl–Meyer scale (1 point cut-off) remained as significant determinants in the final model. The model was no more accurate than the therapists when making predictions at discharge from the acute stroke unit (Fig. 2). This study is important because it indicates that clinical impression and clinical measures are not always sufficient to accurately predict the motor recovery plateau for individual patients, and other approaches may be needed in order to set realistic rehabilitation goals for each patient.

Accuracy of predictions of upper limb motor function. Experienced therapists were asked to predict Action Research Arm Test (ARAT) score at 6 months poststroke when patients were 72 h poststroke (Therapist @ 72 h) and at discharge from the acute stroke unit (Therapist @ discharge). Their overall accuracy rates were 60 and 72%, respectively [16▪▪]. Multivariate modelling was used to develop a computational prediction model based on dichotomized shoulder abduction and finger extension scores. The model was also used to predict 6-month ARAT score when patients were 72 h poststroke (Model @ 72 h) and at discharge from the acute stroke unit (Model @ discharge). The overall error rates were 65 and 69%, respectively [16▪▪]. The predicting recovery potential (PREP) algorithm sequentially combined Medical Research Council grades for shoulder abduction and finger extension, with neurophysiological and neuroimaging measures of descending motor pathway integrity, as required for each patient during the first 2 weeks poststroke (PREP algorithm). The algorithm was used to stratify patients according to ARAT score at 3 months poststroke, with an overall accuracy rate of 82% [17].

These recent results are in keeping with an earlier study that measured upper limb impairment with the Fugl–Meyer scale, and found that patients improved to 70% of their maximum possible improvement on this scale by 3 months poststroke [18]. For example, a mildly impaired patient with an initial Fugl–Meyer score of 46/66 will improve by around 14 points (70% of 20) to 50/66 by 3 months poststroke. A more impaired patient with an initial Fugl–Meyer score of 26/66 will improve by around 28 points (70% of 40) by 3 months poststroke. Given that patients were studied in two different rehabilitation systems, in two different countries, this finding seems to reflect a fundamental characteristic of the resolution of motor impairment after stroke [18]. However, the 70% rule did not hold for patients with initially more severe impairment and Fugl–Meyer scores less than 11/66, some of whom recovered to a lesser extent than predicted. This again highlights the limitations of using clinical measures to predict motor recovery for individual patients. The authors speculated that the extent of damage to the corticospinal tract may have been an important factor for these patients, an idea supported by subsequent studies examining recovery of upper limb function.

By way of example, an algorithm (PREP, for predicting recovery potential) has been recently developed to predict the motor function recovery plateau, incorporating measures of corticomotor pathway integrity [17]. The algorithm sequentially combines an assessment of shoulder abduction and finger extension, transcranial magnetic stimulation (TMS) to test the functional integrity of the corticomotor pathways, and MRI to detect the extent of damage to the posterior limbs of the internal capsules. The algorithm is designed to predict individual patients’ potential to make a complete, notable or limited recovery, or no recovery, of upper limb function within 3 months, measured with the ARAT.

In a recent study, patients with first-ever ischaemic stroke advanced through the steps of the PREP algorithm, with approximately 60% requiring TMS and 20% requiring MRI [17]. First, shoulder abduction and finger extension strength are graded using the Medical Research Council scale, and the scores summed to form the Shoulder Abduction, Finger Extension (SAFE) score. Patients with a SAFE score of 8 or more within 72 h are predicted to have potential for a complete recovery of upper limb function within 12 weeks, and no further testing is required. Those with a SAFE score less than 8 proceed to a TMS evaluation. Patients from whom motor-evoked potentials can be recorded from the paretic wrist extensors are predicted to have potential for a notable recovery of upper limb function within 12 weeks, and no further testing is required. Patients from whom TMS cannot elicit motor-evoked potentials proceed to an MRI evaluation, wherein diffusion-weighted imaging is used to determine the mean fractional anisotropy of the posterior limbs of the internal capsules, and the asymmetry between them. If the fractional anisotropy asymmetry index between these structures is low (<0.15), this indicates residual integrity of the ipsilesional descending pathways, and the patient is predicted to have potential for a limited recovery of upper limb function. If the fractional anisotropy asymmetry index is higher, the patient is predicted to have no potential for recovery of upper limb function within 12 weeks.

The PREP algorithm made correct predictions for 82% of patients, which is higher than the best case 72% accuracy rate of therapists [16▪▪], but there is still room for improvement. The algorithm also produced four stratifications, rather than two groups at the extreme ends of function and a group in the middle with a wide range of ARAT scores. Importantly, it was able to distinguish between patients with more severe initial impairment, whose outcomes cannot be predicted by clinical measures alone. This study extends previous findings demonstrating that the extent of damage to descending motor pathways is a key factor that limits motor performance [19] and outcome [20▪,21▪]. Further work is needed to validate and improve the PREP algorithm, and determine its clinical usefulness.

In summary, recent work indicates that the level of motor impairment or function achieved by the end of the spontaneous recovery period cannot be accurately predicted for individual patients using clinical impression or measures [16▪▪,18]. However, the motor function recovery plateau can be more accurately predicted by combining clinical measures with an objective evaluation of the extent of damage to key motor pathways in the brain [17].


A recent systematic review found that only 30 motor rehabilitation randomized controlled trials (RCTs) were of good quality and recruited all participants within 30 days of stroke [22▪]. Half of the 30 trials were negative, and these were more likely to recruit fewer than 40 patients and make no follow-up assessments, compared with positive trials. This indicates that very little is known about the effectiveness of interventions delivered during the first few weeks after stroke, when most rehabilitation takes place.

Only three of the positive trials made follow-up assessments more than 3 months after stroke. One study of amphetamine found no difference between the treatment and control groups at 3, 6 and 12 month follow-ups [23]. The second study administered five daily doses of real or sham repetitive transcranial magnetic stimulation (rTMS). Those in the treatment groups had greater upper limb strength at 1, 3 and 12 months after stroke. However, there were no benefits reported immediately after the intervention, and the treatment groups’ advantage seemed to increase over time [24]. This is difficult to reconcile with the short-term physiological effects of rTMS. The third study delivered extra upper or lower limb training, or a control intervention, during the first 5 months after stroke [25]. Patients in the lower limb group had better gait function than those in the control group at the end of the intervention, but this was not sustained at follow-up 6 weeks later. Patients in both the upper and lower limb groups had better upper limb function immediately after the intervention and 6 weeks later. This indicates that better upper limb outcomes were not specifically related to extra upper limb training, and may have arisen because of underlying between-group differences in the potential for recovering upper limb function.

One of the limitations of nearly all subacute motor rehabilitation RCTs is that experimental groups are matched at baseline on clinical measures, rather than measures of the brain. As described above, clinical measures are often inaccurate when predicting an individual's capacity to recover motor function. Two groups of subacute stroke patients can have similar baseline clinical scores, but patients in one group may have less damage to key motor pathways, and therefore, more potential for recovery, than patients in the other group. This can create between-group differences in outcome that are not due to the experimental intervention. It is possible that the two studies reporting an intervention's benefits outlasting the spontaneous recovery period [24,25] had more patients with higher recovery ceilings in the treatment groups, simply by chance. Large sample sizes are likely to overcome this potential problem; however, most subacute RCTs have included fewer than 50 patients [22▪].

Overall, there is little evidence available to support the idea that new therapies delivered during the spontaneous recovery process can improve upper limb functional outcomes beyond the first 3 months poststroke. Increasing the rate of recovery, rather than extent, may be a more useful aim. This may be particularly appropriate for experimental techniques designed to enhance neural plasticity, which could accelerate plastic reorganization within the limits of recovery imposed by the extent of motor pathway damage.


Two recent studies have addressed this question. The first randomized 50 acute stroke patients to receive either sham or real transcranial direct current stimulation with the anode over the ipsilesional motor cortex to facilitate its excitability [26▪]. Motor impairment was evaluated using the Fugl–Meyer scale immediately before and after the 5-day intervention period, and again 3 months poststroke. The intervention was found to be safe, but had no effect on any outcome measures. The authors note that their study tested only one of several possible stimulation protocols, and that TDCS may be beneficial for selected patients [26▪].

The second study randomized 57 acute stroke patients to receive either real or sham bilateral priming [27▪]. Bilateral priming involves a device that mechanically couples the two hands. Patients actively produce rhythmic flexion–extension of the nonparetic wrist, and mirror-symmetric movements of the paretic hand are generated through a mechanical linkage that confers an inertial advantage. The device allows patients to produce several hundred movement cycles without fatigue. Bilateral priming facilitates ipsilesional corticomotor excitability for at least 30 min, potentially creating a window of heightened plasticity during which therapy can be delivered [28,29]. Patients completed 15 min of real or sham priming immediately prior to daily upper limb therapy sessions during the 4-week intervention period [27▪]. Motor function was measured with the ARAT at baseline, immediately after the 4-week intervention period, and at 3 and 6 months poststroke. Bilateral priming accelerated the rate of recovery of upper limb function (Fig. 3). A greater proportion of patients in the treatment group achieved their motor recovery plateau within 3 months, and final motor outcomes were not different between the groups at 6 months, as expected [27▪].

Bilateral priming accelerates the rate of recovery of upper limb motor function. The proportion of patients achieving their upper limb motor function plateau at 12 weeks was higher for patients who had bilateral priming (Primed), compared with those who had sham priming (Control) (χ2 test, P < 0.05) [27▪]. The grey box indicates the intervention period. The primed and control groups were balanced at baseline for predicted motor recovery plateau based on principles from the predicting recovery potential algorithm [17], and motor outcomes at 26 weeks did not differ between groups, as expected.

The designs of these two trials differ in ways that may account for their contrasting results. TDCS was initiated 2 days after stroke, and delivered for only 5 days, whereas bilateral priming was initiated 2 weeks after stroke and delivered for 4 weeks. The longer intervention period in the latter study provided more opportunity for bilateral priming to facilitate the neurobiological mechanisms of spontaneous recovery. TDCS was delivered at an unspecified time each day, rather than immediately before or during physiotherapy, which could capitalize on the transient physiological effects of stimulation. In contrast, bilateral priming was delivered immediately before upper limb therapy, so that therapy took place during the period of motor cortex facilitation produced by the intervention. Finally, therapy dose was not controlled in the TDCS trial, whereas it was carefully matched between groups in the bilateral priming trial, to ensure that differences in the rate of recovery were not because of differences in therapy dose. The TDCS trial's negative result may therefore be related to the nonspecific timing of the short-duration intervention, and systematic between-group differences in therapy dose, rather than limitations of the intervention. For both trials, the specific interventions are less important than the idea of accelerating recovery.

These two studies are the first reports of trials intended to increase the rate, rather than extent, of motor recovery during the subacute stage of stroke. They therefore represent a potential shift towards a new paradigm for rehabilitation during the spontaneous recovery process. Given that there is strong evidence in favour of a predictable motor recovery plateau, and little evidence that rehabilitation interventions can modify this plateau, future trials could usefully build upon these two recent studies and test whether interventions modify the rate of motor recovery after stroke.


More than 250 RCTs of motor rehabilitation have been conducted at the chronic stage, and many are positive [30]. This has produced strong arguments against the concept of a recovery plateau, and in favour of delivering ongoing therapy [30,31,32▪]. However, these arguments overlook the deterioration typically experienced by patients after they reach their recovery plateau [33–35]. Gains at the chronic stage may reflect a reversal of deterioration, rather than further recovery beyond the initial motor function plateau. It is also possible that patients continue to improve at the chronic stage if they failed to reach their full potential for motor recovery during the spontaneous recovery period, because of an inadequate rehabilitation dose.

Although there is little doubt that patients can benefit from therapeutic input at the chronic stage [32▪], improvements are generally more modest and heavily dependent on repetition, compared with those made initially. For example, the Extremity Constraint Induced Therapy Evaluation (EXCITE) trial of constraint-induced movement therapy delivered 6 h of upper limb therapy daily for 2 weeks. The average time to complete the Wolf Motor Function Test improved by 10.0 s in the treatment group and 6.4 s in the control group, a small but statistically significant improvement [36]. Participants in a trial of the MIT-Manus robotic device completed around 1000 repetitions of task-oriented upper limb movements in each of 36 sessions delivered over 12 weeks. There were no differences in primary outcome between robotic therapy, intensive comparison therapy and standard care, as Fugl–Meyer score increased between 1 and 4 points on average across groups [37]. It seems that high-intensity therapy is required to produce modest improvements at the chronic stage of recovery. This is probably because gains at this stage depend on basal levels of neural plasticity and ‘everyday’ motor learning. Perhaps greater gains could be made if some of the permissive neurobiological conditions of the spontaneous recovery period could be reinstated. This is the motivation for studies of adjuvants that alter cellular properties, such as noninvasive brain stimulation techniques [38]. However, recent meta-analyses [39,40▪] and Cochrane reviews [41▪▪,42▪▪] have drawn conflicting conclusions, indicating that further trials are needed to establish the effectiveness of these techniques. Future work will need to consider the residual integrity of key motor pathways when selecting stimulation protocols, to avoid interfering with the adaptive reorganization that took place during spontaneous recovery [43▪].

Overall, it is currently unclear whether improvements at the chronic stage reflect a reversal of deterioration, continued improvement towards a motor recovery plateau that was not initially achieved, or a genuine raising of the motor recovery plateau. Good-quality longitudinal studies are needed to explore these possibilities.


The first few weeks after stroke provide a unique opportunity to interact with the neurobiological processes underlying spontaneous recovery, but at present this ‘critical period … is all but ignored’ ([44], p. 927). More good-quality motor rehabilitation RCTs that initiate the trial intervention early after stroke are urgently needed [22▪].

Cellular therapies may hold the key to modifying the recovery plateau, by promoting neuronal repair and regeneration in lesioned pathways [44–46]. These therapies could theoretically produce outcomes that exceed patients’ predicted recovery ceilings, profoundly changing stroke rehabilitation theory and practice. In the meantime, enhancing plasticity at the subacute stage may accelerate plastic reorganization within the limits on recovery imposed by the lesion. Trials of interventions designed to promote neural plasticity could therefore make measures throughout the spontaneous recovery period, to determine whether the intervention increases the rate of recovery. Trials will need to match groups on the extent of damage to key motor pathways at baseline, so they are balanced for predicted motor recovery plateau. Groups will also need to have matched therapy doses, for both standard rehabilitation and the intervention. Factors that limit the patient's engagement in rehabilitation may prevent them from reaching their maximum possible level of motor function during the spontaneous recovery period. These factors include comorbidities, sensory deficits, and impaired communication and cognition, which also need to be balanced between experimental and control groups.


The unique neurobiological processes at work after stroke support recovery of motor function, with maximum recovery achieved by most patients within 3 months. However, neural plasticity is usually unable to fully compensate for lost descending motor connections, which are currently untreatable. Therapies that accelerate the rate of recovery could improve the efficiency of rehabilitation, and the first reports of trials taking this approach have recently appeared.



Conflicts of interest

There are no conflicts of interest.


Papers of particular interest, published within the annual period of review, have been highlighted as:

  • ▪ of special interest
  • ▪▪ of outstanding interest


1. Langhorne P, Coupar F, Pollock A. Motor recovery after stroke: a systematic review. Lancet Neurol 2009; 8:741–754.
2. WHO. World Health Report 2003. Geneva: World Health Organization; 2003.
3. Patel AT, Duncan PW, Lai SM, Studenski S. The relation between impairments and functional outcomes poststroke. Arch Phys Med Rehabil 2000; 81:1357–1363.
4. Schiemanck SK, Kwakkel G, Post MW, et al. Predicting long-term independency in activities of daily living after middle cerebral artery stroke: does information from MRI have added predictive value compared with clinical information? Stroke 2006; 37:1050–1054.
5. Veerbeek JM, Kwakkel G, van Wegen EE, et al. Early prediction of outcome of activities of daily living after stroke: a systematic review. Stroke 2011; 42:1482–1488.
6. Kreisel SH, Hennerici MG, Bazner H. Pathophysiology of stroke rehabilitation: the natural course of clinical recovery, use-dependent plasticity and rehabilitative outcome. Cerebrovasc Dis 2007; 23:243–255.
7. Nakayama H, Jorgensen HS, Raaschou HO, Olsen TS. Recovery of upper extremity function in stroke patients: the Copenhagen Stroke Study. Arch Phys Med Rehabil 1994; 75:394–398.
8. Kwakkel G, Kollen BJ. Predicting activities after stroke: what is clinically relevant? Int J Stroke 2013; 8:25–32.
9. Jorgensen HS, Nakayama H, Raaschou HO, et al. Outcome and time course of recovery in stroke. Part II: Time course of recovery. The Copenhagen Stroke Study. Arch Phys Med Rehabil 1995; 76:406–412.
10. Chen J, Chopp M. Neurorestorative treatment of stroke: cell and pharmacological approaches. NeuroRx 2006; 3:466–473.
    11. Wieloch T, Nikolich K. Mechanisms of neural plasticity following brain injury. Curr Opin Neurobiol 2006; 16:258–264.
    12. Murphy TH, Corbett D. Plasticity during stroke recovery: from synapse to behaviour. Nat Rev Neurosci 2009; 10:861–872.
    13. Horn SD, DeJong G, Smout RJ, et al. Stroke rehabilitation patients, practice, and outcomes: is earlier and more aggressive therapy better? Arch Phys Med Rehabil 2005; 86:S101–S114.
    14. Maulden SA, Gassaway J, Horn SD, et al. Timing of initiation of rehabilitation after stroke. Arch Phys Med Rehabil 2005; 86:S34–S40.
    15. Paolucci S, Antonucci G, Grasso MG, et al. Early versus delayed inpatient stroke rehabilitation: a matched comparison conducted in Italy. Arch Phys Med Rehabil 2000; 81:695–700.
    16▪▪. Nijland RH, van Wegen EE, Harmeling-van der Wel BC, Kwakkel G. Accuracy of physical therapists’ early predictions of upper-limb function in hospital stroke units: The EPOS study. Phys Ther 2013; 93:460–469.

    This study asked therapists with at least 20 years experience in stroke rehabilitation to predict whether patients’ upper limb function would be severely impaired, normal, or somewhere in between at 6 months poststroke. Overall accuracy rates were 72% at best, with the most errors made for patients predicted to be somewhere between severely impaired and normal.

    17. Stinear CM, Barber PA, Petoe M, et al. The PREP algorithm predicts potential for upper limb recovery after stroke. Brain 2012; 135:2527–2535.
    18. Zarahn E, Alon L, Ryan SL, et al. Prediction of motor recovery using initial impairment and fMRI 48 h poststroke. Cereb Cortex 2011; 21:2712–2721.
    19. Lindenberg R, Renga V, Zhu LL, et al. Structural integrity of corticospinal motor fibers predicts motor impairment in chronic stroke. Neurology 2010; 74:280–287.
    20▪. Puig J, Blasco G, Daunis IEJ, et al. Decreased corticospinal tract fractional anisotropy predicts long-term motor outcome after stroke. Stroke 2013; 44:2016–2018.

    This study found that asymmetry in the integrity of the corticospinal tract at the level of the pons, measured 30 days after stroke, was related to stratified Motricity Index scores 2 years after stroke. These data indicate that degeneration of the corticospinal tract, distal to the stroke lesion, could be a useful marker of long-term motor outcome.

    21▪. Cheng B, Forkert ND, Zavaglia M, et al. Influence of stroke infarct location on functional outcome measured by the modified Rankin scale. Stroke 2014; 45:1695–1702.

    This study found that lesions affecting the corona radiata and internal capsule had a strong influence on functional outcome measured with the modified Rankin scale at 1 month poststroke.

    22▪. Stinear C, Ackerley S, Byblow W. Rehabilitation is initiated early after stroke, but most motor rehabilitation trials are not: a systematic review. Stroke 2013; 44:2039–2045.

    This review found that there are only 30 good-quality motor rehabilitation trials that enrolled all patients within 30 days of stroke, representing around 6% of all motor rehabilitation RCTs. This indicates that very little is known about the effectiveness of interventions delivered during the first few weeks after stroke, when most rehabilitation takes place.

    23. Walker-Batson D, Smith P, Curtis S, et al. Amphetamine paired with physical therapy accelerates motor recovery after stroke Further evidence. Stroke 1995; 26:2254–2259.
    24. Khedr EM, Etraby AE, Hemeda M, et al. Long-term effect of repetitive transcranial magnetic stimulation on motor function recovery after acute ischemic stroke. Acta Neurol Scand 2010; 121:30–37.
    25. Kwakkel G, Wagenaar RC, Twisk JW, et al. Intensity of leg and arm training after primary middle-cerebral-artery stroke: a randomised trial. Lancet 1999; 354:191–196.
    26▪. Rossi C, Sallustio F, Di Legge S, et al. Transcranial direct current stimulation of the affected hemisphere does not accelerate recovery of acute stroke patients. Eur J Neurol 2013; 20:202–204.

    This study investigated the effects of transcranial direct current stimulation delivered for 5 days, beginning 2 days after stroke. Although the intervention appeared to be safe, it did not modify upper limb impairment at 3 months poststroke.

    27▪. Stinear CM, Petoe MA, Anwar S, et al. Bilateral priming accelerates recovery of upper limb function after stroke: a randomized controlled trial. Stroke 2014; 45:205–210.

    This study investigated the effects of bilateral coordinated movement delivered for 4 weeks, beginning 2 weeks after stroke. A higher proportion of patients in the treatment group reached maximum upper limb function by 3 months poststroke than in the control group. This is the first study to report a positive intervention effect on the rate of motor recovery.

    28. Byblow WD, Stinear CM, Smith MC, et al. Mirror symmetric bimanual movement priming can increase corticomotor excitability and enhance motor learning. PLoS One 2012; 7:e33882.
    29. Stinear CM, Barber PA, Coxon JP, et al. Priming the motor system enhances the effects of upper limb therapy in chronic stroke. Brain 2008; 131:1381–1390.
    30. Teasell R, Mehta S, Pereira S, et al. Time to rethink long-term rehabilitation management of stroke patients. Top Stroke Rehabil 2012; 19:457–462.
    31. Page SJ, Gater DR, Bach YRP. Reconsidering the motor recovery plateau in stroke rehabilitation. Arch Phys Med Rehabil 2004; 85:1377–1381.
    32▪. Korner-Bitensky N. When does stroke rehabilitation end? Int J Stroke 2013; 8:8–10.

    This commentary summarizes the evidence that chronic stroke patients can benefit from a range of therapies, targeting aerobic fitness, movement, communication, and cognition.

    33. Thorsen AM, Holmqvist LW, de Pedro-Cuesta J, von Koch L. A randomized controlled trial of early supported discharge and continued rehabilitation at home after stroke: five-year follow-up of patient outcome. Stroke 2005; 36:297–303.
    34. Dhamoon MS, Moon YP, Paik MC, et al. Trajectory of functional decline before and after ischemic stroke: the Northern Manhattan Study. Stroke 2012; 43:2180–2184.
    35. Sonde L, Kalimo H, Fernaeus SE, Viitanen M. Low TENS treatment on poststroke paretic arm: a three-year follow-up. Clin Rehabil 2000; 14:14–19.
    36. 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.
    37. Lo AC, Guarino PD, Richards LG, et al. Robot-assisted therapy for long-term upper limb-impairment after stroke. N Engl J Med 2010; 362:1772–1783.
    38. Ackerley SJ, Stinear CM. Stimulating stimulation: Can we improve motor recovery following stroke using repetitive transcranial magnetic stimulation? Phys Ther Rev 2010; 15:302–308.
    39. Hsu WY, Cheng CH, Liao KK, et al. Effects of repetitive transcranial magnetic stimulation on motor functions in patients with stroke: a meta-analysis. Stroke 2012; 43:1849–1857.
    40▪. Butler AJ, Shuster M, O’Hara E, et al. A meta-analysis of the efficacy of anodal transcranial direct current stimulation for upper limb motor recovery in stroke survivors. J Hand Ther 2013; 26:162–170.

    This meta-analysis of eight RCTs found that anodal TDCS may benefit upper limb function in chronic stroke patients.

    41▪▪. Elsner B, Kugler J, Pohl M, Mehrholz J. Transcranial direct current stimulation (tDCS) for improving function and activities of daily living in patients after stroke. Cochrane Database Syst Rev 2013; 11:CD009645.

    This systematic review of 15 RCTs concluded that the evidence in favour of TDCS is of low quality, and that further work is required to determine its effectiveness for enhancing outcomes after stroke.

    42▪▪. Hao Z, Wang D, Zeng Y, Liu M. Repetitive transcranial magnetic stimulation for improving function after stroke. Cochrane Database Syst Rev 2013; 5:CD008862.

    This systematic review of 19 RCTs concluded that there is currently no evidence to support the routine use of repetitive TMS in stroke treatment.

    43▪. Bradnam LV, Stinear CM, Byblow WD. Ipsilateral motor pathways after stroke: implications for noninvasive brain stimulation. Front Hum Neurosci 2013; 7:184.

    This perspective article summarized recent empirical studies that demonstrate how noninvasive brain stimulation (NIBS) may influence excitability of neural circuits at subcortical and spinal levels. For stroke recovery, protocols that aim to suppress excitability of the contralesional hemisphere may be contraindicated for some patients – a reminder that NIBS is not ‘one-size-fits-all’.

    44. Krakauer JW, Carmichael ST, Corbett D, Wittenberg GF. Getting neurorehabilitation right: what can be learned from animal models? Neurorehabil Neural Repair 2012; 28:923–931.
    45. Savitz SI, Chopp M, Deans R, et al. Stem cell therapy as an emerging paradigm for stroke (STEPS) II. Stroke 2011; 42:825–829.
    46. Savitz SI, Cramer SC, Wechsler L, et al. Stem cells as an emerging paradigm in stroke 3: enhancing the development of clinical trials. Stroke 2014; 45:634–639.

    motor; recovery; rehabilitation; stroke

    Supplemental Digital Content

    © 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins