Clinical Neurophysiology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Clinical Neurophysiology
journal homepage: www.elsevier.com/locate/clinph

Multiple sessions of transcranial direct current stimulation and upper

extremity rehabilitation in stroke: A review and meta-analysis
L. Tedesco Triccas a,⇑, J.H. Burridge a, A.M. Hughes a, R.M. Pickering b, M. Desikan c, J.C. Rothwell c,
G. Verheyden d
a
Faculty of Health Sciences, University of Southampton, UK
b
Medical Statistics Group, Faculty of Medicine, University of Southampton, UK
c
Sobell Department of Motor Neuroscience, Institute of Neurology, University College of London, UK
d
KU Leuven, Department of Rehabilitation Sciences, Belgium

a r t i c l e i n f o h i g h l i g h t s

Article history:
 tDCS and rehabilitation had small non-significant effect on upper extremity impairments.
Accepted 15 April 2015
 Varied tDCS and rehabilitation programmes were identified in selected studies.
Available online xxxx
 Future research needs to further analyse tDCS and therapy interventions in stroke.

Keywords:
Transcranial direct current stimulation
Rehabilitation a b s t r a c t
Stroke
Upper extremity
Objective: To systematically review the methodology in particular treatment options and outcomes and
Recovery the effect of multiple sessions of transcranial direct current stimulation (tDCS) with rehabilitation pro-
Non-invasive brain stimulation grammes for upper extremity recovery post stroke.
Methods: A search was conducted for randomised controlled trials involving tDCS and rehabilitation for
the upper extremity in stroke. Quality of included studies was analysed using the Modified Downs and
Black form. The extent of, and effect of variation in treatment parameters such as anodal, cathodal and
bi-hemispheric tDCS on upper extremity outcome measures of impairment and activity were analysed
using meta-analysis.
Results: Nine studies (371 participants with acute, sub-acute and chronic stroke) were included. Different
methodologies of tDCS and upper extremity intervention, outcome measures and timing of assessments
were identified. Real tDCS combined with rehabilitation had a small non-significant effect of +0.11
(p = 0.44) and +0.24 (p = 0.11) on upper extremity impairments and activities at post-intervention respec-
tively.
Conclusion: Various tDCS methods have been used in stroke rehabilitation. The evidence so far is not sta-
tistically significant, but is suggestive of, at best, a small beneficial effect on upper extremity impairment.
Significance: Future research should focus on which patients and rehabilitation programmes are likely to
respond to different tDCS regimes.
Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights
reserved.

1. Introduction Clarke, 1999). In Europe, stroke costs around 64.1 billion euros and
in the United Kingdom, around £8.9 billion per annum is spent on
Stroke is a health concern worldwide and one of the main community care and rehabilitation of people with stroke
causes of disability (Kolominsky-Rabas et al., 2001; Albert and (Saka et al., 2009; Gustavsson et al., 2011). At 6 months, 33–66%
Kesselring, 2012). Motor impairment is the main cause of disability of people with upper extremity (UE) impairments do not present
after stroke, leading to major health problems (Boggio et al., 2007; with functional upper limb function and only 5–20% achieve full
recovery (Kwakkel et al., 2003; Kwakkel and Kollen, 2013). Thus
⇑ Corresponding author at: Building 45, Faculty of Health Sciences, University of a number of approaches are now being investigated in an attempt
Southampton, Southampton SO17 1BJ, UK. Tel.: +44 7791383674. to increase the effectiveness of stroke rehabilitation techniques for
E-mail address: l.tedesco-triccas@uea.ac.uk (L. Tedesco Triccas). the UE.

http://dx.doi.org/10.1016/j.clinph.2015.04.067
1388-2457/Ó 2015 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
2 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx

Non-invasive methods of brain stimulation such as transcranial recovery. A Cochrane review showed that tDCS has a small effect
direct current stimulation (tDCS) and repetitive transcranial mag- on UE motor impairments but not on activities of daily living at
netic stimulation (rTMS) are extensively researched and are begin- post-intervention (Elsner et al., 2013). However, at follow-up they
ning to be used clinically to modulate brain activity (Paulus, 2003; showed an effect of tDCS on activities of daily living but not on UE
Pascual-Leone et al., 2000; Hummel et al., 2005). Although these motor impairments. No effect of tDCS in sub-groups involving
two methods have very different modes of action (rTMS stimulates people with acute, sub-acute and chronic stroke was found. The
axons in the brain and initiates new action potentials; tDCS analyses in these reviews combined studies including one or mul-
polarises the neurones, and modulates their ongoing firing pattern) tiple sessions of tDCS, and the pooled effects of only multiple tDCS
both of them, when applied over the motor cortex, produce sessions plus therapy remains uninvestigated.
changes in cortical excitability which, in the case of tDCS can last The aim of the current study was to systematically review the
up to 90 min (Nitsche and Paulus, 2000, 2001; Fitzgerald et al., methodology adopted in various studies of tDCS. In particular
2006). They also enhance motor performance and can change reac- treatment options, outcomes reported, and the effect of multiple
tion times, movement accuracy and speed (Nitsche et al., 2003b; sessions of tDCS with rehabilitation programmes for UE recovery
Kobayashi et al., 2004). More importantly, in the context of possi- post stroke. We included trials in which anodal, cathodal or
ble therapeutic application, they can improve motor skill learning bi-hemispheric tDCS was applied in conjunction with UE rehabili-
(Reis and Fritsch, 2011; Teo et al., 2011) or adapt already learned tation programmes at any stage post stroke.
skills to new conditions (Galea et al., 2011). There has therefore
been considerable interest in examining the potential of these
2. Methods
interventions to augment recovery of motor function after stroke.
Initial investigations with non-invasive brain stimulation con-
2.1. Search
centrated on using methods of rTMS to improve recovery in acute
and chronic stroke (Khedr et al., 2005; Kim et al., 2006). However
A systematic search was conducted for articles written in
in recent years there has been increased interest in using tDCS
English and published between 1990 and July 2014. Full text arti-
because of two main advantages: firstly it is far less expensive than
cles in electronic databases MEDLINE, EMBASE (Excerpta Medica
rTMS, and secondly, stimulation can potentially be applied during
Database), CINAHL (Cumulated Index of Nursing and Allied
rehabilitation whereas rTMS (because the equipment is bulky and
Health Literature), AMED (Allied and Complementary Medicine
the head needs to remain still), it can only be given before (or after)
Database), PubMed, PEDro (Physiotherapy Evidence Database)
a training session (Brunoni et al., 2012). From a practical view-
were systematically searched by the first author (LTT).
point, anodal tDCS is usually assumed to increase excitability
Combination of key words with the use of ‘AND’ and ‘OR’ were
whereas cathodal tDCS reduces excitability (Nitsche and Paulus,
used for the searches (Table 1). Duplicates were removed and
2000). In stroke rehabilitation this means that researchers will
papers were then selected for analysis according to the eligibility
employ anodal tDCS over the stroke hemisphere to improve the
criteria (Fig. 1).
response of that hemisphere to training protocols (Hummel et al.,
2005). Alternatively, employing the logic of inter-hemispheric
imbalance, cathodal stimulation of the non-stroke hemisphere will 2.2. Eligibility criteria
inhibit that hemisphere to reduce its trans-hemispheric inhibition
of the affected hemisphere or bihemispheric stimulation by simul- 2.2.1. Participants
taneously modulating the unaffected and affected motor cortex The type of participants included in the study needed to: (a)
(Nitsche et al., 2003a; Lindenberg et al., 2010). have a confirmed clinical diagnosis of a haemorrhagic or an ischae-
In healthy volunteers, the effects of tDCS on cortical excitability mic stroke as defined by WHO (Monica, 1988); (b) have experi-
and performance are short-lasting and variable (Nitsche and enced a single stroke or multiple strokes; (c) be in the acute
Paulus, 2000; López-Alonso et al., 2014; Wiethoff et al., 2014). (starting intervention during 2-weeks post-stroke), sub-acute
However, it is usually assumed that multiple daily applications in (2 weeks to 3 months post-stroke) or chronic phase (after three
stroke may lead to a build-up of effects that are larger and more months post-stroke); (d) have a subcortical or cortical stroke; (e)
persistent. The main evidence in favour of this comes from studies be male or female over the age of 18 years; (f) have any type of
of rTMS to treat depression: a single session, or even 2 weeks daily UE impairment; and (g) received multiple sessions of tDCS, defined
treatment with rTMS has little effect on symptoms over and above as two or more sessions, plus UE therapy.
placebo, whereas longer (>4 weeks) treatments can improve symp-
toms for several months (Dell’Osso et al., 2011; Galletly et al., 2.2.2. Study designs
2012). Thus most recent clinical trials of tDCS have employed sev- All randomised controlled trials (RCTs) utilising multiple ses-
eral days or weeks of repeated treatment with rehabilitation pro- sions of tDCS in combination with rehabilitation were included
grammes in an attempt to maximise outcome (Lee and Chun, regardless of blinding. Outcome measures were classified accord-
2014; Viana et al., 2014). Interestingly it is still unclear whether ing to the International Classification of Functioning, Disability
repeated daily session of tDCS has cumulative effects in the healthy and Health (ICF) (WHO, 2001). Where possible we selected one
and stroke population (Alonzo et al., 2012; Monte-Silva et al.,
2013).
Table 1
Recent meta-analyses have explored the effect of tDCS in addi- An example of a search strategy on MEDLINE.
tion to rehabilitation on UE activity in stroke (Bastani and
Database MEDLINE
Jaberzadeh, 2012; Adeyemo et al., 2012; Butler et al., 2013;
Elsner et al., 2013). Adeyemo et al. (2012) demonstrated a signifi- Date 26th July 2014
cant effect size (0.58) of non-invasive brain stimulation on motor Strategy #1 and #2 and #3
#1 ‘transcranial direct current stimulation’
function. Bastani and Jaberzadeh, 2012 showed that anodal tDCS #2 ‘stroke/or exp brain stem infarctions/or exp cerebral infarction/’,
had a small non-significant effect size of 0.39 on hand function ‘cerebrovascular accident’
in stroke, but a moderate significant effect size of 0.59 on motor #3 ‘upper extremity/or exp arm/or exp axilla/or exp elbow/or exp
evoked potential amplitude. Butler et al. (2013) also demonstrated forearm/upper extremity/or exp shoulder/’, ‘upper extremity/or
exp fingers/or exp metacarpus/or exp wrist’,
a significant small effect size of 0.40 of anodal tDCS on UE motor

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx 3

Fig. 1. Study selection PRISMA flow diagram.

outcome from each paper in each ICF category: impairment, activ- stimulation as reported in the study of Lee and Chun (2014), were
ity and participation. The chosen primary outcome measure of UE inputted in the program and meta-analysed. Studies reporting only
impairment was Fugl-Meyer Assessment (FMA) and the outcomes the medians of the selected outcomes were not included in the
of activity and participation were classed as secondary. meta-analyses (Wu et al., 2013). All outcome measures were anal-
ysed as continuous variables, using the mean, standard deviation
and number of participants at immediate post-intervention,
2.3. Process
short-term (1–12 weeks) or long-term follow-up (>12 weeks).
Data relating to a different activity outcome measure abstracted
Three review authors (LTT, AMH, GV) independently assessed
from each study were also combined in the activity
the methodological quality of the included studies using a vali-
meta-analysis. The standardised mean differences (using Hedges’
dated, reliable tool, the Modified Downs and Black (Eng et al.,
adjusted g) and 95% confidence intervals were calculated. All out-
2007). Disagreement between the reviewers was resolved through
comes were standardised so that positive differences between
discussion between the three review authors. If no agreement was
tDCS and control groups favoured tDCS. Hedges’ adjusted g, similar
achieved, a fourth review author (JHB) was used to gain consensus.
to Cohen’s (d), was used to estimate the effect size and includes an
Papers with scores lower than 16 out of 27 points were excluded
adjustment for small sample bias of RCTs. The effect size was inter-
from the analysis due to poor quality (Eng et al., 2007).
preted according to Cohen’s convention of small (>0.2), moderate
(>0.5), and large (>0.8) effects (Cohen, 1992). A 95% confidence
2.4. Data synthesis and analyses interval around the estimated effect size was calculated for each
analysis. Meta-analysis of the neurophysiological data was not
Different methodologies including study population, selection conducted due to limited data.
criteria and outcome measures, timing of assessment, intervention
programmes (dosage/intensity, mode of delivery, frequency, dura-
3. Results
tion and timing of tDCS delivery) and adverse reactions were
extracted from the selected publications. The main characteristics
3.1. Search results
of the population were calculated as percentages from the total
sample. The studies were then assessed for heterogeneity to
The PRISMA flow chart (Fig. 1) shows the search and selection
determine the appropriateness of a pooled analysis. Information
results. Nine papers scored 16 points or above using the Downs
from the selected papers was analysed using the Cochrane
and Black scoring were included in the qualitative analysis and
Collaboration’s Review Manager software, RevMan (Version 5.3).
eight were (Eng et al., 2007) (Table 2).
Statistical heterogeneity was assessed using the I-squared and
Chi-squared statistic. Due to the small number of studies included,
a fixed effects model was adopted for all analyses (Borenstein et al., 3.2. Main characteristics of the RCT’s
2010). Paper authors were contacted with respect to any missing
data of the included studies. Data of the FMA impairment outcome In total, 371 participants with stroke (243 males) meeting the
measure in relation to the effect of UE rehabilitation with real tDCS eligibility criteria for our review study were included in the nine
(anodal and/or cathodal or bihemispheric) versus sham tDCS or no selected studies. All participants had a single stroke apart from

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
 Table 2

4
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in

Characteristics of studies included in the review.

Study Objective Design Groups N Mean Mean Time tDCS Training Period Outcomes according to the ICF Modified
To investigate: Age since stroke Stimulation (weeks) [I = Impairment, A = Activity, Downs and
(years) Intensity/Duration/ P = Participation] Black Score
Hemisphere (score out of
27)
Kim et al. tDCS and OT on UL motor Single- Anodal and 6 55.3 34.0 days (i) 2 mA 10 sessions over 2 weeks and FMA*(I) 21
(2010) recovery blinded, OT (ii) 20 min 30 min OT* MBI* (A)
RCT (iii) tDCS during
rehabilitation
(iv) Anodal: ipsi-le-
sional Cathodal:
contra-lesional
Cathodal and 5 53.6 19.4 days
OT
Sham and OT 7 62.9 22.9 days
Lindenberg tDCS and PT and OT on UL Double- Bi- 10 61.7 30.5 months (i) 1.5 mA 5 daily sessions of 60 min OT and PT FMA (I) 17

L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx

et al. motor recovery blinded hemispheric (ii) 30 min fMRI* (I)
(2010) RCT and OT (iii) tDCS during WMFT*(A)
rehabilitation
(iv) Anodal: ipsi-le-
sional Cathodal:
contra-lesional
Sham and OT 10 55.8 40.3 months
Bolognini tDCS and CIMT on UL Double- Bi- 7 42.6 44.4 months (i) 2 mA 14 daily sessions of four hours FMA (I) JTT* (I), HG*(I), Resting Motor 16
et al. motor recovery blinded hemispheric (ii) 40 min CIMT* Threshold and Trans-colossal
(2011) RCT and CIMT (iii) tDCS during inhibition (I), MAL*(A), BI (A)
rehabilitation
(iv) Anodal: ipsi-le-
sional Cathodal:
contra-lesional
Sham and 7 50.9 26.0 months
CIMT
Hesse et al. tDCS and RT on UL motor Double- Anodal and 32 63.9 3.4 weeks (i) 2 mA 30 sessions over 6 weeks involving FMA (I), 22
(2011) recovery blinded RT (ii) 20 min 20 min RT MRC* (I),
RCT (iii) tDCS during MAS*(I),
rehabilitation BI*(A),
(iv) Anodal: ipsi-le- BBT* (I)
sional, Cathode:
contra-lesional
Cathodal and 32 65.4 3.8 weeks
RT
Sham and RT 32 65.6 3.8 weeks
Nair et al. Cathodal/sham tDCS and Double- Cathodal and 7 61.0 33 months (i) 30 min 5 daily sessions of 1 hour OT ROM* (I); 17
(2011) OT on UL motor recovery blinded OT (ii) 1 mA FMA* (I)
RCT (iii) tDCS during fMRI (I)
rehabilitation
(iv) Cathodal:
contra-lesional
Sham and OT 7 56.0 28 months
Khedr et al. Anodal/cathodal/sham Double- Anodal and 14 58.1 13.8 days (i) 25 min 6 daily sessions of one hour National Institute of Health Stroke 25
(2013) tDCS and Rehabilitation UL blinded Therapy (ii) 2 mA rehabilitation (passive movement Scale (I),
motor recovery RCT (iii) tDCS before and range of motion exercises) Orgogozo MCA scale (I),
rehabilitation MRC (I)
Resting and Active Motor Threshold
 Table 2 (continued)
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in

Study Objective Design Groups N Mean Mean Time tDCS Training Period Outcomes according to the ICF Modified
To investigate: Age since stroke Stimulation (weeks) [I = Impairment, A = Activity, Downs and
(years) Intensity/Duration/ P = Participation] Black Score
Hemisphere (score out of
27)
(iv) Anodal: ipsi-le-
sional Cathodal:
contra-lesional
Cathodal and 13 60.0 12.3 days
Therapy
Sham and 13 57.0 12.6 days
Therapy
Wu et al. Cathodal tDCS and Double- Cathodal and 45 45.9 4.9 months (i) 20 min 5 sessions per week for 4 weeks of FMA(I), MAS (I), BI (A) 24
(2013) rehabilitation on UL motor blinded PT (ii) 1.2 mA 30 min (twice daily) PT
recovery and spasticity RCT (iii) ? tDCS delivery
(iv) ? cathode

L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx

placed on
contra-lesional
or ipsi-lesional
M1
Sham and PT 45 49.3 4.9 months
Lee and Cathodal tDCS and virtual Double- Cathodal and 21 60.3 17.4 days (i) 20 min 5 sessions per week for 3 weeks of MAS (I), Manual Muscle Test (I), 17
Chun reality Programme on UL blinded OT (ii) 2 mA 30 min each session of virtual reality Manual Function Test (I). FMA (I). BBT
(2014) impairments Pilot RCT (iii) tDCS during OT (I). Korean MBI (A)
and virtual
reality
(iv) Cathode over
contra-lesional
M1
No tDCS and 22 60.6 16.9 days
virtual
reality
Cathodal and 21 63.1 17.8 days
virtual
reality
Viana et al. Anodal tDCS and virtual Double- Anodal tDCS 10 56.0 31.9 months (i) 13 min 3 sessions per week for 5 weeks of FMA (I), WMFT (I), MAS (I), 18
(2014) reality on UL impairments blinded and virtual (ii) 2 mA one hour each session Dynamometry (I), Stroke Specific
RCT reality (iii) ?before/during Quality of Life Scale (P)
and after
rehabilitation
(iv) Anode over
ipsi-lesional
M1
Sham tDCS 10 55.0 35.0 months
and Virtual
Reality
*
BI = Barthel Index, BBT = Box and Block Test, CIMT = Constraint Induced Movement Therapy, fMRI = functional Magnetic Resonance Imaging, FMA = Fugl-Meyer Assessment, HG = Hand Grip, MAS = Modified Ashworth Scale,
MAL = Motor Activity Log, MBI = Modified Barthel Index, MEP = motor evoked potential, MRC = Medical Research Council Strength, ROM = Range of Motion, MT = Motor Threshold, OT =Occupational Therapy, PT = Physiotherapy,
WMFT = Wolf Motor Function Test.

5
6 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx

the ones in the study by Viana et al. (2014). Limited detail was All studies included a baseline and post-intervention assess-
reported regarding stroke location. Combined over studies ment. Follow-up assessments varied between seven days
included in the meta-analysis, 164 (44%) participants had (Lindenberg et al., 2010); 2 weeks (Bolognini et al., 2011); 4 weeks
cortico-subcortical or cortical strokes and 49 (11%) participants (Wu et al., 2013; Bolognini et al., 2011); 3 months (Hesse et al.,
had subcortical stroke. Three studies did not describe the location 2011; Khedr et al., 2013); and 6 months (Kim et al., 2010)
of the strokes (Lindenberg et al., 2010; Wu et al., 2013; Viana et al., post-intervention. Three studies reported additional assessments:
2014). Four studies involved 65 (18%) participants with chronic at 1 day (Kim et al., 2010); 3 days (Lindenberg et al., 2010); and
stroke (Lindenberg et al., 2010; Bolognini et al., 2011; Nair et al., 6 days (Nair et al., 2011) after the intervention was ended as a
2011; Viana et al., 2014), four studies involved 263 (71%) partici- follow-up assessment.
pants in the sub-acute stage (Kim et al., 2010; Hesse et al., 2011;
Wu et al., 2013; Lee and Chun, 2014) and one study involved 40 3.2.2. tDCS parameters
(11%) people with acute stroke (Khedr et al., 2013). Three studies randomised participants to anodal, cathodal and
One hundred and forty-five (39%) participants had moderate UE sham group (Kim et al., 2010; Hesse et al., 2011; Khedr et al.,
impairments (Kim et al., 2010; Lindenberg et al., 2010; Bolognini 2013); one study involved an anodal and sham group (Viana
et al., 2011; Nair et al., 2011; Lee and Chun, 2014; Viana et al., et al., 2014); two studies randomised participants to cathodal
2014) and 226 (61%) participants had severe UE impairments and sham group (Nair et al., 2011; Wu et al., 2013); one study
(Hesse et al., 2011; Khedr et al., 2013; Wu et al., 2013) at baseline had a cathodal group and no stimulation as a placebo (Lee and
based on the FMA or MRC Scale (shoulder) (Lum et al., 2002). Chun, 2014); and two studies randomised to bihemispheric tDCS
and sham groups (Lindenberg et al., 2010; Bolognini et al., 2011).
3.2.1. Outcome measures and timing of assessments The tDCS parameters deployed in the studies are displayed in
Eight studies used the FMA as an outcome measure of UE motor Fig. 2.
impairments (Kim et al., 2010; Lindenberg et al., 2010; Bolognini The timing of tDCS delivery in relation to the rehabilitation was:
et al., 2011; Hesse et al., 2011; Nair et al., 2011; Wu et al., 2013; during rehabilitation (Kim et al., 2010; Lindenberg et al., 2010;
Lee and Chun, 2014; Viana et al., 2014) and one study used the Bolognini et al., 2011; Hesse et al., 2011; Nair et al., 2011; Lee
Medical Research Council scale (Khedr et al., 2013). Only two stud- and Chun, 2014); before rehabilitation (Khedr et al., 2013; Viana
ies explored the neurophysiological effects of tDCS, using measures et al., 2014); or not reported (Wu et al., 2013).
of Resting and Active Motor Threshold (Khedr et al., 2013), motor
evoked potential amplitude (Bolognini et al., 2011) and transcal- 3.2.3. Rehabilitation programmes
losal inhibition to transcranial magnetic stimulation (Bolognini The rehabilitation programme in one study was constraint
et al., 2011). Two studies employed functional MRI of brain activity induced movement therapy daily for 14 days (Bolognini et al.,
as an outcome measure (Lindenberg et al., 2010; Nair et al., 2011). 2011). In three studies, it was conventional therapy (positioning,
The Barthel Index or Modified Barthel Index and Motor Activity Log passive movements, stretching and movement/functional training
were utilised as measures of activities of daily living (Kim et al., and, goal-directed activities of practical) for 30 min; 5 days per
2010; Hesse et al., 2011; Bolognini et al., 2011; Khedr et al., week, for 4 weeks (Wu et al., 2013); daily for 5 weeks
2013; Wu et al., 2013; Lee and Chun, 2014). (Lindenberg et al., 2010; Nair et al., 2011); or for 6 days (Khedr

Fig. 2. A flow diagram displaying the different tDCS parameters selected for the included studies.

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx 7

Fig. 3. Effect of real tDCS versus sham tDCS for UE global motor impairments measured by FMA at immediate post-intervention.

Fig. 4. Effect of real tDCS versus sham tDCS for UE global motor impairments measured by FMA at short-term follow-up.

Fig. 5. Effect of real tDCS versus sham tDCS for UE global motor impairments measured by FMA at long-term follow-up.

et al., 2013). In the remaining studies it was an occupational ther- obtained favouring anodal over sham tDCS combined with rehabil-
apy programme comprising of task practice as part of the WMFT itation on UE impairments was observed based on three studies
for 10 sessions over 2 weeks (Kim et al., 2010); focused bilateral (Kim et al., 2010; Hesse et al., 2011; Viana et al., 2014) (Fig. 6). A
wrist robot therapy for 20 min for 30 sessions (Hesse et al., small non-significant effect of 0.10 [ 0.26, 0.47] (p = 0.59)
2011); or virtual reality 5 sessions per week for 3 weeks of favoured cathodal stimulation and rehabilitation at
30 min each session (Lee and Chun, 2014) or 3 sessions per week post-intervention, based on three studies (Kim et al., 2010; Hesse
for 5 weeks of 1 h each session (Viana et al., 2014). et al., 2011; Lee and Chun, 2014) (Fig. 7). Bihemispheric stimula-
tion and UE rehabilitation showed a larger non-significant effect
3.2.4. Sensations and adverse reactions of +0.17 [ 0.50, 0.84] (p = 0.62) over sham stimulation, based on
Five studies reported sensations and adverse reactions from two studies involving chronic stroke participants (Lindenberg
tDCS. Participants reported tingling or slight itching under the et al., 2010; Bolognini et al., 2011) (Fig. 8).
tDCS electrodes (Wu et al., 2013; Hesse et al., 2011; Lindenberg
et al., 2010). One participant discontinued anodal tDCS due to a 3.3.2. Sub-analyses
headache and one participant receiving cathodal tDCS reported Five studies (Kim et al., 2010; Bolognini et al., 2011; Hesse et al.,
dizziness (Kim et al., 2010; Hesse et al., 2011). 2011; Khedr et al., 2013; Lee and Chun, 2014) using different out-
come measures of generic activities of daily living showed a
3.3. Quantitative analysis non-significant small effect size of 0.24 [ 0.06,0.54] (p = 0.11)
favouring real tDCS in combination with rehabilitation at
3.3.1. The effect of real versus sham tDCS and rehabilitation on UE post-intervention (Fig. 9). Three studies (Kim et al., 2010; Hesse
motor impairments et al., 2011; Khedr et al., 2013) explored the effect of anodal
Seven studies (Kim et al., 2010; Lindenberg et al., 2010; tDCS and rehabilitation on activities of daily living with a small
Bolognini et al., 2011; Hesse et al., 2011; Nair et al., 2011; Lee non-significant effect of 0.19 [ 0.12, 0.50] (p = 0.23) favouring ano-
and Chun, 2014; Viana et al., 2014) explored the effect of real (ano- dal tDCS and rehabilitation (Fig. 10). Three studies (Hesse et al.,
dal, cathodal or bi-hemispheric) versus sham tDCS combined with 2011; Khedr et al., 2013; Lee and Chun, 2014) explored the effect
rehabilitation programmes on UE motor impairments measured by of cathodal tDCS and rehabilitation on activities of daily living with
FMA at immediate post-intervention (Fig. 3). An overall small a small significant effect of 0.38 [0.03,0.73] (p = 0.03) favouring
non-significant effect size of +0.11[ 0.17, 0.38] (p = 0.44) favoured anodal tDCS and rehabilitation (Fig. 11).
real tDCS and rehabilitation compared to sham stimulation at
post-intervention. A non-significant effect size +0.27 [ 0.40, 4. Discussion
0.95] (p = 0.43) was noted between 7 days and 2 weeks follow-up
(Lindenberg et al., 2010; Bolognini et al., 2011) (Fig. 4). A small Driven by the positive benefits of repeated sessions of rTMS to
non-significant effect size +0.23 [ 0.17, 0.62) (p = 0.26) was noted treat depression, many current trials apply the same logic when
at a long-term follow-up ranging between 3 and 6 months for UE using tDCS to improve recovery of arm function after stroke. This
global motor impairments in participants with sub-acute stroke is the first review to explore the effect of just multiple sessions
(Kim et al., 2010; Hesse et al., 2011) (Fig. 5). A small of tDCS and UE rehabilitation programmes on outcome measures
non-significant effect size of +0.01 [ 0.39, 0.41] (p = 0.96) was based on the ICF; impairment and activity. Therefore, it examines

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
8 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx

Fig. 6. Effect of anodal tDCS versus sham tDCS for UE global motor impairments measured by FMA at immediate post-intervention.

Fig. 7. Effect of cathodal tDCS versus sham tDCS for UE global motor impairments measured by FMA at immediate post-intervention.

Fig. 8. Effect of bihemispheric tDCS versus sham tDCS for UE global motor impairments measured by FMA at immediate post-intervention.

Fig. 9. Effect of real tDCS versus sham tDCS for activities of daily living at immediate post-intervention.

Fig. 10. Effect of anodal tDCS versus sham tDCS for activities of daily living measured at immediate post-intervention.

Fig. 11. Effect of cathodal tDCS versus sham tDCS for activities of daily living measured at immediate post-intervention.

the potential clinical use of tDCS. We performed a meta-analysis of favoured tDCS, and possibly larger sample sizes are needed to
the data, in which this approach was used to provide a confirm such small effects. Unfortunately, although the analyses
current overview of the expected effect size for subsequent trials. included data from studies in acute, sub-acute and chronic stroke
Nine studies met the inclusion criteria. The results of the they were insufficient to analyse specific effects dependent on
meta-analysis showed that when tDCS is applied in conjunction stage of recovery.
with multiple sessions of rehabilitation it has no significant imme- There are a number of important limitations which may affect
diate effect, over and above therapy delivered alone, on UE impair- the validity of these conclusions. A limitation is that the ICF
ment and activity after stroke. The two larger (greater than 0.10) emphasises that in addition to measuring impairments and func-
immediate post-intervention effect sizes on UE impairments both tion, clinicians should also assess participation (Perenboom and

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx 9

Chorus, 2003; WHO, 2001). Most of the studies included only participants had severe UE impairments and showed minimal
reported UE impairment outcome measures. Apart from the effect from real tDCS and rehabilitation. However, it is unclear
Motor Activity Log, the selected outcome measures do not specifi- whether adding tDCS is more important for people with moderate
cally assess UE activities of daily living (Uswatte et al., 2006). In than severe UE impairments. For instance there is evidence that
addition, including outcome measures such as the Stroke Impact robot therapy is more beneficial for people with moderate UE
Scale in clinical trials of tDCS plus rehabilitation (Duncan et al., impairments compared to severely affected people with stroke
2003) would give an insight into the effect of intervention on par- (Ferraro et al., 2003). There is also a lack of evidence whether the
ticipants’ quality of life. different types of tDCS might be more beneficial for people in the
The wide variability in the methodology of both the tDCS acute or the chronic phase of stroke. Thus, stratifying participants
method as well as the rehabilitation therapy used severely limit according to their stage after stroke and their UE impairments in
the conclusions we can draw from these studies. As far as the addition with tDCS application should be the next focus for future
tDCS was concerned, the exact timing of the post-intervention research.
assessment with respect to the last or first treatment session was The review has limitations: only a small number of eligible
often not reported. In addition, the duration, frequency, and timing studies were identified; only absolute group summary statistics
of the intervention were different in every study. Hesse et al. taken immediately post-intervention or at follow-up were
(2011) used a treatment programme of 30 sessions whilst included in the meta-analyses. A more powerful analysis would
Lindenberg et al. (2010) used five sessions. Some studies evaluated be possible if baseline data could be included in order to compare
tDCS applied before rehabilitation whereas in others it was applied this with the post-intervention data. Finally, due to the statistical
during rehabilitation. In healthy individuals, anodal tDCS applied program used, studies reporting medians could not be included
during motor learning can result in faster learning than if applied in the meta-analyses.
prior to motor learning (Stagg et al., 2011). Anodal tDCS can also
result in prolonged skill acquisition through an effect on offline
5. Conclusions
consolidation but whether the same advantage occurs in stroke
is unknown (Reis et al., 2009). Finally, the parameters of stimula-
Systematic reviews and meta-analyses have been a popular
tion differed between studies, some used 1 mA and others 2 mA
method of exploring the effect of tDCS on UE recovery in stroke.
applied for durations varying from 10 to 40 min. Whether different
The results from this review showed that multiple sessions of
chosen currents are equally effective is unknown. In healthy indi-
tDCS regimes combined with UE rehabilitation had a small and
viduals, the effects of cathodal tDCS on motor cortex excitability
non-significant effect on upper limb impairments and activities
reverse from suppression to facilitation when the stimulus inten-
of daily living post-intervention. However, from this review there
sity increases from 1 mA to 2 mA (Batsikadze et al., 2013).
is wide variation in tDCS parameters adopted in the constituent tri-
However, it is not possible to extrapolate with confidence from
als and different UE therapy. Factors such as selection criteria, tDCS
effects on motor excitability in healthy individuals to effects on
parameters, type of rehabilitation programmes and type and stage
motor performance in disease.
of stroke have to be further explored. Future research should focus
The review has highlighted wide differences in protocols, and
on stratifying participants with stroke according to their UE
given these variations coupled with the small numbers of partici-
impairments and evaluate which participants are more likely to
pants, it is not possible to form any firm conclusions about the effi-
benefit from tDCS in addition to stroke rehabilitation. This will
cacy of tDCS in improving the effect of UE rehabilitation after
increase the knowledge about the effect of tDCS and rehabilitation
stroke. A further difficulty is that the efficacy of the tDCS may vary
for UE motor recovery in stroke and eventually will contribute to
depending on the details, and efficacy of the associated UE rehabil-
further high quality meta-analyses.
itation programme. Variations in the rehabilitation therapy applied
to patients were as widespread as those used in tDCS. Some
patients only received ‘‘conventional’’ therapy for 6 days, whereas Acknowledgments
others were employed in robot training protocols or CIMT. It is pos-
sible for example that a ‘‘ceiling effect’’ occurred in some cases. We would to thank the Faculty of Health Sciences of the 502
That is, the therapy was sufficient on its own to produce an University of Southampton and also the Strategic Educational
improvement in outcome that could not be boosted further by add- Pathways Scholarship Scheme (Malta) – The scholarship is
ing tDCS. part-financed by the European Union – European Social Fund.
Greater insight into the effects might be obtained if sensitive Conflicts of interest statement: None of the authors have poten-
cortical activity or neurophysiological outcome measures are uti- tial conflicts of interest to be disclosed.
lised in all randomised controlled trials. Until this is addressed
through large definitive trials the problem will persist. Outcome References
measures, participants’ characteristics and protocols need to be
standardised. This occurred early in treatment trials of rTMS in Adeyemo BO, Simis M, Macea D, Fregni F. Systematic review of parameters of
depression, where a relatively ‘‘standard’’ treatment is 10 Hz stimulation: clinical trial design characteristics and motor outcomes in
noninvasive brain stimulation in stroke. Front Psychiatry 2012;3:88. http://
rTMS of the left dorsolateral prefrontal cortex (O’Reardon et al., dx.doi.org/10.3389/fpsyt.2012.00088.
2007). This standardisation allowed for more effective Albert S, Kesselring J. Neurorehabilitation of stroke. J Neurol 2012;259:817–32.
meta-analyses to be undertaken than is currently possible with Alonzo A, Brassil J, Taylor JL, Martin D, Loo CK. Daily transcranial direct current
stimulation (tDCS) leads to greater increases in cortical excitability than second
tDCS interventions in stroke. daily transcranial direct current stimulation. Brain Stimul 2012;5:208–13.
Recent studies on healthy participants have emphasised that Bastani A, Jaberzadeh S. Does anodal transcranial direct current stimulation
the response to tDCS is highly variable between individuals enhance excitability of the motor cortex and motor function in healthy
individuals and subjects with stroke: a systematic review and meta-analysis.
(López-Alonso et al., 2014; Wiethoff et al., 2014). Several clinical
Clin Neurophysiol 2012;123(4):644–57.
trials involving participants with stroke have also reported highly Batsikadze G, Moliadze V, Paulus W, Kuo MF, Nitsche MA. Partially non-linear
variable responses (Hummel et al., 2005; Bolognini et al., 2011; stimulation intensity-dependent effects of direct current stimulation on motor
Hesse et al., 2011; Khedr et al., 2013). Thus one important future cortex excitability in humans. J Physiol 2013;591:1987–2000.
Boggio PS, Nunes A, Rigonatti SP, Nitsche MA, Pascual-Leone A, Fregni F. Repeated
consideration may be to determine which patients with stroke sessions of noninvasive brain DC stimulation is associated with motor function
are most likely to benefit. In the study by Hesse et al. (2011), the improvement in stroke patients. Restor Neurol Neurosci 2007;25:123–9.

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067
10 L. Tedesco Triccas et al. / Clinical Neurophysiology xxx (2015) xxx–xxx

Bolognini N, Vallar G, Casati C, Latif LA, El-Nazer R, Williams J, et al. Lee SJ, Chun MH. Combination transcranial direct current stimulation and virtual
Neurophysiological and behavioral effects of tDCS combined with constraint- reality therapy for upper extremity training in patients with subacute stroke.
induced movement therapy in poststroke patients. Neurorehabil Neural Repair Arch Phys Med Rehabil 2014;95:431–8.
2011;25:819–29. Lindenberg R, Renga V, Zhu L, Nair D, Schlaug G. Bihemispheric brain stimulation
Borenstein M, Hedges LV, Higgins J, Rothstein HR. A basic introduction to fixed- facilitates motor recovery in chronic stroke patients. Neurology
effect and random-effects models for meta-analysis. Res Synth Methods 2010;75:2176–84.
2010;1:97–111. Lopez-Alonso V, Cheeran B, Río-Rodríguez D, Fernández-Del-Olmo M. Inter-
Brunoni AR, Nitsche MA, Bolognini N, Bikson M, Wagner T, Merabet L, et al. Clinical individual variability in response to non-invasive brain stimulation
research with transcranial direct current stimulation (tDCS): challenges and paradigms. Brain Stimul 2014;7(3):372–80.
future directions. Brain Stimul 2012;5:175–95. Lum PS, Burgar CG, Shor PC, Majmundar M, Van Der Loos M. Obot-assisted
Butler AJ, Shuster M, O’Hara E, Hurley K, Middlebrooks D, Guilkey K. A meta- movement training compared with conventional therapy techniques for the
analysis of the efficacy of anodal transcranial direct current stimulation for rehabilitation of upper-limb motor function after stroke. Arch Phys Med Rehabil
upper limb motor recovery in stroke survivors. J Hand Ther 2013;26:162–71. 2002;83:952–9.
Clarke PJ. Handicap in stroke survivors. Disabil Rehabil 1999;21:116–23. Monica W. The World Health Organization MONICA Project (monitoring trends and
Cohen J. A power primer. Psychol Bull 1992;112(1):155–9. determinants in cardiovascular disease): a major international collaboration:
Dell’Osso B, Camuri G, Castellano F, Vecchi V, Benedetti M, Bortolussi S, et al. Meta- WHO MONICA Project Principal Investigators. J Clin Epidemiol 1988;41:105–14.
review of metanalytic studies with repetitive transcranial magnetic stimulation Monte-Silva K, Kuo MF, Hessenthaler S, Fresnoza S, Liebetanz D, Paulus W, et al.
(rTMS) for the treatment of major depression. Clin Pract Epidemiol Ment Health Induction of late LTP-like plasticity in the human motor cortex by repeated non-
2011;7:167–77. invasive brain stimulation. Brain stimul 2013;6:424–32.
Duncan PW, Bode RK, Min Lai S, Perera S. Rasch analysis of a new stroke-specific Nair DG, Renga V, Lindenberg R, Zhu L, Schlaug G. Optimizing recovery potential
outcome scale: the stroke impact scale. Arch Phys Med Rehabil through simultaneous occupational therapy and non-invasive brain-
2003;84:950–63. stimulation using tDCS. Restor Neurol Neurosci 2011;29:411–20.
Elsner B, Kugler J, Pohl M, Mehrholz J. Transcranial direct current stimulation (tDCS) Nitsche M, Paulus W. Excitability changes induced in the human motor cortex by
for improving function and activities of daily living in patients after stroke. weak transcranial direct current stimulation. J Physiol 2000;527:633–9.
Cochrane Database Syst Rev 2013;11:CD009645. Nitsche MA, Paulus W. Sustained excitability elevations induced by transcranial DC
Eng JJ, Teasell RW, Miller WC, Wolfe DL, Townson AF, Aubut JA, et al. Spinal cord motor cortex stimulation in humans. Neurology 2001;57:1899–901.
injury rehabilitation evidence: method of the SCIRE systematic review. Top Nitsche MA, Nitsche MS, Klein CC, Tergai F, Rothwell JC, Paulus W. Level of action of
Spinal Cord Inj Rehabil 2007;13:1–10. cathodal DC polarisation induced inhibition of the human motor cortex. Clin
Ferraro M, Palazzolo J, Krol J, Krebs H, Hogan N, Volpe B. Robot-aided sensorimotor Neurophysiol 2003a;114:600–4.
arm training improves outcome in patients with chronic stroke. Neurology Nitsche MA, Schauenburg A, Lang N, Liebetanz D, Exner C, Paulus W, et al.
2003;61:1604–7. Facilitation of implicit motor learning by weak transcranial direct current
Fitzgerald PB, Fountain S, Daskalakis ZJ. A comprehensive review of the effects of stimulation of the primary motor cortex in the human. J Cogn Neurosci
rTMS on motor cortical excitability and inhibition. Clin Neurophysiol 2003b;15:619–26.
2006;117:2584–96. O’Reardon JP, Solvason HB, Janicak PG, Sampson S, Isenberg KE, Nahas Z, et al.
Galea JM, Vazquez A, Pasricha N, De Xivry JJO, Celnik P. Dissociating the roles of the Efficacy and safety of transcranial magnetic stimulation in the acute treatment
cerebellum and motor cortex during adaptive learning: the motor cortex retains of major depression: a multisite randomized controlled trial. Biol Psychiatry
what the cerebellum learns. Cereb Cortex 2011;21:1761–70. 2007;62:1208–16.
Galletly C, Gill S, Clarke P, Burton C, Fitzgerald P. A randomized trial comparing Pascual-Leone A, Walsh V, Rothwell J. Transcranial magnetic stimulation in
repetitive transcranial magnetic stimulation given 3 days/week and 5 days/ cognitive neuroscience–virtual lesion, chronometry, and functional
week for the treatment of major depression: is efficacy related to the duration connectivity. Curr Opin Neurobiol 2000;10:232–7.
of treatment or the number of treatments? Psychol Med 2012;42:981–8. Paulus W. Transcranial direct current stimulation (tDCS). Suppl Clin Neurophysiol
Gustavsson A, Svensson M, Jacobi F, Allgulander C, Alonso J, Beghi E, et al. Cost of 2003;56:249–54.
disorders of the brain in Europe 2010. Eur Neuropsychopharmacol Perenboom RJ, Chorus AM. Measuring participation according to the International
2011;21:718–79. Classification of Functioning, Disability and Health (ICF). Disabil Rehabil
Hesse S, Waldner A, Mehrholz J, Tomelleri C, Pohl M, Wener C. Combined 2003;25:577–87.
transcranial direct current stimulation and robot-assisted arm training in Reis J, Fritsch B. Modulation of motor performance and motor learning by
subacute stroke patients an exploratory, randomized multicenter trial. transcranial direct current stimulation. Curr Opin Neurol 2011;24:590–6.
Neurorehabil Neural Repair 2011;25:838–46. Reis J, Schambra HM, Cohen LG, Buch ER, Fritsch B, Zarahn E, et al. Noninvasive
Hummel F, Celnik P, Giraux P, Floel A, Wu WH, Gerloff C, et al. Effects of non- cortical stimulation enhances motor skill acquisition over multiple days
invasive cortical stimulation on skilled motor function in chronic stroke. Brain through an effect on consolidation. Proc Natl Acad Sci U S A 2009;106:1590–5.
2005;128:490–9. Saka Ö, Mcguire A, Wolfe C. Cost of stroke in the United Kingdom. Age Ageing
Khedr EM, Ahmed MA, Fathy M, Rothwell JC. Therapeutic trial of repetitive 2009;38:27–32.
transcranial magnetic stimulation after acute ischemic stroke. Neurology Stagg C, Jayaram G, Pastor D, Kincses Z, Matthews P, Johansen-Berg H. Polarity and
2005;65:466–8. timing-dependent effects of transcranial direct current stimulation in explicit
Khedr EM, Shawky OA, El-Hammady DH, Rothwell JC, Darwish ES, Mostafa OM, motor learning. Neuropsychologia 2011;49:800–4.
et al. Effect of anodal versus cathodal transcranial direct current stimulation on Teo JT, Swayne OB, Cheeran B, Greenwood RJ, Rothwell JC. Human theta burst
stroke rehabilitation a pilot randomized controlled trial. Neurorehabil Neural stimulation enhances subsequent motor learning and increases performance
Repair 2013;27:592–601. variability. Cereb Cortex 2011;21:1627–38.
Kim DY, Lim JY, Kang EK, You DS, Oh MK, Oh BM, et al. Effect of transcranial direct Uswatte G, Taub E, Morris D, Light K, Thompson P. The motor activity log-28
current stimulation on motor recovery in patients with subacute stroke. Am J assessing daily use of the hemiparetic arm after stroke. Neurology
Phys Med Rehabil 2010;89(11):879–86. 2006;67:1189–94.
Kim YH, You SH, Ko MH, Park JW, Lee KH, Jang SH, et al. Repetitive transcranial Viana R, Laurentino G, Souza R, Fonseca J, Silva Filho E, Dias S, et al. Effects of the
magnetic stimulation-induced corticomotor excitability and associated motor addition of transcranial direct current stimulation to virtual reality therapy
skill acquisition in chronic stroke. Stroke 2006;37:1471–6. after stroke: a pilot randomized controlled trial. NeuroRehabilitation
Kobayashi M, Hutchinson S, Theoret H, Schlaug G, Pascual-Leone A. Repetitive TMS 2014;34(3):437–46.
of the motor cortex improves ipsilateral sequential simple finger movements. WHO 2001. International Classification of Functioning, Disability and Health (ICF)
Neurology 2004;62:91–8. [Online]. WHO. Available: http://www.who.int/classifications/icf/en/ [Accessed
Kolominsky-Rabas PL, Weber M, Gefeller O, Neundoerfer B, Heuschmann PU. 15/07/2012 2012].
Epidemiology of ischemic stroke subtypes according to TOAST criteria: Wiethoff S, Hamada M, Rothwell JC. Variability in response to transcranial direct
incidence, recurrence, and long-term survival in ischemic stroke subtypes: a current stimulation of the motor cortex. Brain Stimul 2014;7(3):468–75.
population-based study. Stroke 2001;32:2735–40. Wu D, Qian L, Zorowitz RD, Zhang L, Qu Y, Yuan Y. Effects on decreasing upper-limb
Kwakkel G, Kollen B. Predicting activities after stroke: what is clinically relevant? poststroke muscle tone using transcranial direct current stimulation: a
Int J Stroke 2013;8:25–32. randomized sham-controlled study. Arch Phys Med Rehabil 2013;94:1–8.
Kwakkel G, Kollen BJ, Van Der Grong J, Prevo AJH. Probability of regaining dexterity
in the flaccid upper limb: impact of severity of paresis and time since onset in
acute stroke. Stroke 2003;34:2181–6.

Please cite this article in press as: Tedesco Triccas L et al. Multiple sessions of transcranial direct current stimulation and upper extremity rehabilitation in
stroke: A review and meta-analysis. Clin Neurophysiol (2015), http://dx.doi.org/10.1016/j.clinph.2015.04.067