Guidelines

International Journal of Stroke

2019, Vol. 14(8) 774–782
Cognition in stroke rehabilitation and ! 2019 World Stroke Organization
Article reuse guidelines:
recovery research: Consensus-based sagepub.com/journals-permissions
DOI: 10.1177/1747493019873600

core recommendations from the second journals.sagepub.com/home/wso

Stroke Recovery and Rehabilitation

Roundtable

Matthew W McDonald1,3 , Sandra E Black2,3, David A Copland4,

Dale Corbett1,3, Rick M Dijkhuizen5, Tracy D Farr6 ,
Matthew S Jeffers1,3 , Rajesh N Kalaria7, Frini Karayanidis8,
Alexander P Leff9, Jess Nithianantharajah10, Sarah Pendlebury11,
Terence J Quinn12, Andrew N Clarkson13,* and
Michael J O’Sullivan14,*

Abstract
Cognitive impairment is an important target for rehabilitation as it is common following stroke, is associated with
reduced quality of life and interferes with motor and other types of recovery interventions. Cognitive function following
stroke was identified as an important, but relatively neglected area during the first Stroke Recovery and Rehabilitation
Roundtable (SRRR I), leading to a Cognition Working Group being convened as part of SRRR II. There is currently
insufficient evidence to build consensus on specific approaches to cognitive rehabilitation. However, we present recom-
mendations on the integration of cognitive assessments into stroke recovery studies generally and define priorities for
ongoing and future research for stroke recovery and rehabilitation. A number of promising interventions are ready to be
taken forward to trials to tackle the gap in evidence for cognitive rehabilitation. However, to accelerate progress requires
that we coordinate efforts to tackle multiple gaps along the whole translational pathway.

Keywords
Cognitive function, practice guideline, stroke, rehabilitation, recovery, consensus

Received: 20 May 2019; accepted: 17 July 2019

6
School of Life Science, University of Nottingham, Nottingham, UK
7
Background Institute of Neuroscience, Newcastle University, Newcastle upon Tyne,
UK
Epidemiology and importance 8
Priority Research Centre for Stroke & Brain Injury, The University of
Newcastle, Callaghan, Australia
The first Stroke Recovery and Rehabilitation 9
Department of Brain Repair and Rehabilitation, UCL Queen Square
Roundtable (SRRR I)1 focused primarily on motor Institute of Neurology, London, UK
10
recovery, as it was a more developed area in terms of Florey Institute of Neuroscience and Mental Health, Florey Department
mechanistic understanding and readiness for clinical of Neuroscience, University of Melbourne, Parkville, Australia
11
Centre for Prevention of Stroke and Dementia, Nuffield Department of
trials. Cognitive function has since been identified Clinical Neurosciences, University of Oxford, Oxford, UK
12
Institute of Cardiovascular and Medical Sciences, University of Glasgow,
1 Glasgow, UK
Department of Cellular and Molecular Medicine, University of Ottawa, 13
The Department of Anatomy, Brain Health Research Centre and Brain
Ottawa, Canada
2 Research New Zealand, University of Otago, Dunedin, New Zealand
Department of Medicine, Sunnybrook Health Sciences Centre, 14
University of Queensland Centre for Clinical Research, Faculty of
University of Toronto, Toronto, Canada
3 Medicine, University of Queensland, Brisbane, Australia
Canadian Partnership for Stroke Recovery, Ottawa, Canada
4
University of Queensland Centre for Clinical Research, School of Health *Joint senior authors.
& Rehabilitation Sciences, University of Queensland, Brisbane, Australia Corresponding author:
5
Biomedical MR Imaging and Spectroscopy Group, Center for Image Michael J O’Sullivan, University of Queensland, Building 71/918, RBWH
Sciences, University Medical Center Utrecht and Utrecht University, Campus, Brisbane, Queensland 4029, Australia.
Utrecht, Netherlands Email: m.osullivan1@uq.edu.au

International Journal of Stroke, 14(8)

McDonald et al.
as an area of unmet need requiring discussion.
The definition of post-stroke cognitive impairment,
adopted by the cognition working group in SRRR II,
is a new cognitive deficit that develops in the first three
months following stroke and persists for a minimum of
six months, which is not explained by any other condi-
tion or disease.2 Such deficits occur in 30–40% of indi-
viduals,3 in one or more cognitive domains, including,
language, executive function, visuospatial cognition,
episodic and working memory.4 Furthermore, cogni-
tive, affective and behavioral consequences of stroke
are more strongly associated with poor quality of life
(QoL) than measures of physical disability.5 The risk of
dementia after stroke is high, with a post-event inci-
dence of 34% one year after severe stroke
(NIHSS>10), with lower rates after TIA and minor
stroke.6 International guidelines highlight the lack of
evidence on specific approaches for rehabilitation of
cognitive function as a significant gap.7 Therefore, a
major goal of SRRR II was to define current consensus
and research priorities to advance our understanding
and maximize research alignment in post-stroke
cognition.
Mechanisms of impairment and recovery
Cognitive function emerges from complex interactions
between cortical and subcortical sites across distributed
brain networks. Lesions, such as focal stroke, may dis-
rupt networks either directly, or indirectly, through sec-
ondary mechanisms of injury. A proper account of the
consequences of damage to specific domains therefore
requires an understanding of the distributed neural net-
works that underpin these neurocognitive domains and
their interactions8 (Figure 1). Typically following
stroke, multiple networks are affected to varying
degrees. This heterogeneity, along with a vast range
of approaches for testing cognition (replicated in
animal behavioral paradigms) creates difficulties in
defining consistent measurement approaches for use
in rehabilitation trials. Acutely, impairments can be
exaggerated by systemic factors, such as infection,
metabolic disorders and drugs causing delirium.
Moreover, many cognitive impairments can be subtle
and often delayed, and manifest when patients return
fully to the complex demands of daily life. Another
complicating factor is the lack of information on pre-
morbid cognitive status.9 On the positive side, alter-
ations in structural and functional connectivity can
now be evaluated non-invasively by imaging and elec-
trophysiological techniques. However, longitudinal stu-
dies with clinical outcomes are needed to fully
understand how brain connectivity changes underpin
the evolution of post-stroke cognitive impairment and
impact on QoL.
775
Context and scope
There have been a number of initiatives to develop con-
sensus in relation to clinical entities that overlap or can
co-exist with acute stroke, such as cerebral small vessel
disease and vascular cognitive impairment (VCI).10,11
These include agreed guidelines for diagnosis and stan-
dardized approaches to neuropsychological testing and
imaging.12,13 A current UK initiative is seeking to build
consensus around animal models of VCI.14 These
efforts are relevant, notably in the emphasis on trans-
lation and new therapies. However, the SRRR II
Cognitive working group remains focused primarily
on the setting of stroke rehabilitation (process of
care) and recovery trials. Trials to modify the course
of VCI, and aspects such as silent infarction or insidi-
ous small vessel disease, were not within the group’s
scope.
Animal models provide a means to explore basic
mechanisms of plasticity and repair and test interven-
tions to promote repair and ameliorate secondary
injury. The two-way interaction between preclinical
and clinical research was therefore viewed as central
to cognitive recovery research and a core component
of the working group’s mission.
Methods
The cognition working group gathered experts from a
diverse range of fields. A core group met in Saint-
Sauveur, Canada in October 2018 and a wider advisory
group was established to provide additional expertise.
Expertise in clinical stroke, rodent models of stroke,
human and animal neuroimaging, neuropsychology,
the neurobiology of language and cognitive rehabilita-
tion were represented.
In advance of SRRR II, a structured survey was sent
to participants and from this, a list of major challenges
was defined, and an agenda formed for the working
group meeting. In a number of areas, it was recognized
that there was inadequate evidence to support formal
consensus. In these areas, the methodology shifted to
definition of major priorities for research in post-stroke
cognition.
Cognition in rehabilitation research:
Generic recommendations
An essential aspect of outcome to be measured in a
stroke recovery study might be defined as one that: is
likely affected by stroke; is sensitive to therapy; and has
importance relative to patients’ overall outcome.
Therapeutic approaches are often not specific to
motor function (e.g. exercise) so that cognitive
improvement may be part of a wider therapeutic
International Journal of Stroke, 14(8)
776 International Journal of Stroke 14(8)

Figure 1. Neural networks that underpin the neurocognitive domains. Each figure sketches the major regions recognized as part of
the network supporting each domain. Key points are that all networks are widely distributed across the brain frequently intersecting
and overlapping so that multiple networks may be injured by a single stroke.

effect. Furthermore, because cognitive impairment
determines QoL after stroke, the collective view was
that cognitive function meets the predefined criteria
for evaluation in all trials and observational studies of
stroke recovery (see Table 1). No consensus was
reached on a single approach to assessment, but the
need for tests that are easily implemented, validated
in different cultural settings and sensitive to executive
function was recognized.
Selection bias and selective attrition—for example,
the exclusion of patients with aphasia or other barriers
to standard cognitive assessments (e.g. visual

International Journal of Stroke, 14(8)

McDonald et al.
Table 1. Consensus recommendations
Recommendations: Observational studies and trials
All clinical intervention studies and trials should include
evaluation of cognition across multiple domains.
Cognitive function should be evaluated at study enrolment
and as an outcome measure (secondary if not primary).
Wherever possible, studies should include evaluation of other
behavioral aspects that are associated with cognition and
important for quality of life: e.g. mood, apathy, fatigue,
anxiety, sleep.
Strategies to limit selection bias and selective attrition should
be standard in clinical studies. Selection based on language
deficits should require formal language assessment, and
adjust test administration for aphasic patients when pos-
sible (e.g. using Supported Conversation16). Reports
should state who were excluded and why.
Preclinical research should utilize models that reproduce
common, clinically relevant cognitive deficits, using a bat-
tery of tests sensitive to multiple domains.
Developmental priorities
Measures of cognitive functioning better adapted to the needs
of trials
Parallel studies of cognitive functioning across multiple
domains, with long follow up periods, in clinical and pre-
clinical research to facilitate translation
Identification of biomarkers for processes and epochs of
recovery (identification of targets for intervention).
Greater use of cognitive paradigms that translate between
clinical and preclinical research (supported by standards for
selection, execution and reporting of tests).
Preclinical models should incorporate age, sex, cardiovascular
and metabolic comorbidities.
impairment)15 was identified as a major concern.
Inclusion of explicit strategies to minimize bias was
another area of agreed consensus (Table 1).
Research priorities: Enhancing the
translational potential of preclinical
cognitive recovery research
Preclinical stroke recovery research focused on motor
systems has identified critical periods of sensorimotor
recovery and cellular mechanisms that govern neural
repair following stroke.17,18 Previous consensus recom-
mendations for aligning preclinical and clinical stroke
recovery research from SRRR I focused on sensori-
motor recovery, although many of these guidelines
could be applied to cognitive recovery.19 The preclinical
recommendations emphasized the importance of using
sensitive outcome measures that are analogous to meas-
ures used in humans.
Traditionally, preclinical research has utilized the
middle cerebral artery occlusion (MCAo) model.
777
Going forward, greater priority should be given to
models that replicate behavioral impairments observed
clinically using photothrombosis, endothelin-1, and
microvascular emboli models that produce targeted
damage to brain regions without major motor deficits.
Additionally, models should assess cognitive deficits in
a variety of domains, particularly higher-order pro-
cesses commonly affected in humans. Most of these
domains can be examined in rodents (Table 2); how-
ever, for some higher order functions non-human pri-
mate models may have greater translational potential.
While traditional paradigms that assess cognition still
have merit (e.g. Morris water maze, etc.), greater
emphasis should be placed on cognitive tasks that dir-
ectly translate across species. This can be achieved
using rodent touch-screen tablets that allows testing
paradigms that mirror those employed in humans.20
A significant component of stroke rehabilitation is
relearning; therefore, preclinical studies should consider
including paradigms that measure both relearning of
old information as well as learning new task informa-
tion. We also know that the severity of cognitive deficits
can be modulated by underlying cognitive risk factors.
Therefore, preclinical models should incorporate fea-
tures such as age, sex, cardiovascular and metabolic
comorbidities.19
Effort should be made to monitor long-term behav-
ioral changes to reflect chronicity and progressive
decline in cognitive function, in combination with
structural and functional changes in neural networks
(histology, neuroanatomical tracing, MRI, electro-
physiology), to identify important epochs and markers
of cognitive recovery. In addition, the preclinical envir-
onment can readily validate novel therapeutic strate-
gies. To maximize translation, adoption of a level of
rigor equivalent to clinical trials in humans is required,
including randomization, blinding, consistent training
in outcome evaluation, and standards for reporting of
experimental procedures.20
Research priorities: Translational and
clinical research
Recovery epochs, therapeutic windows and
biomarkers
The cellular and biochemical changes triggered by
stroke include both early and late events that occur
both proximal and distal to the site of injury. The
notion of distinct recovery epochs dominated by one
or several cellular or biochemical mechanisms empha-
sizes the challenge of correct timing of interventions in
trials. One consensus conclusion was that epochs need
to be defined mechanistically because mechanisms
define candidate therapies. Major gaps in existing
International Journal of Stroke, 14(8)
 Table 2. Cognitive paradigms and translation from models to humans
Neurocognitive
domain (DSM-5) Subdomain Human paradigm
Executive function Cognitive flexibility Wisconsin card Sorting Test
Intra-extradimensional
set-shifting
Digit symbol substitution33
Inhibition/impulsivity Go no-go tasks
Continuous performance
test
International Journal of Stroke, 14(8)
Working memory Digit span
Spatial span
Spatial working memory
Complex attention Sustained attention Choice reaction time35,36
Continuous performance
test
Speed of processing Reaction time tasks
Divided attention Walking while counting
backward
Neglect Cancellation tasks38
Line bisection
Perceptual-motor Object recognition Visual discrimination
function (agnosia) Pattern recognition
Learning and Recognition Delayed non-matching
memory to sample
Scene recognition39,40
778
Preclinical paradigms Comments
Attention set-shift
Reversal learning34
Operant extinction learning
Continuous performance test
T-maze (delayed alternation task) Tasks of spatial working memory
Y-maze translate more easily between
Radial 8-arm maze humans and animal models than
Morris water maze more familiar digit or letter span
Trial-unique delayed non- tasks.
matching-to-location (TUNL)
5-choice serial reaction time37
5-choice continuous performance test
Signal detection task
Cross-modal stimulus presentations
Continuous performance test
5-choice serial reaction time
Unclear whether tested in
preclinical models
Adhesive strip removal The adhesive removal test has a motor
component. Cancellation tasks that
replace letters with other simple
objects more easily translate
between animal models and humans.
Pairwise visual discrimination
Delayed non-matching
to sample
(continued)
International Journal of Stroke 14(8)
McDonald et al. 779

data include a lack of detailed longitudinal studies

human studies, but less well-suited

Tests of verbal recall are common in
using imaging (structural and functional) and other

than spatial, object-based or per-

biomarkers in humans, and a relative paucity of long-

ceptual tasks to translation

term follow up data in animals. Biomarkers provide
promising avenues for the definition of epochs, with
the potential to span clinical and preclinical models.
For example, positron emission tomography (PET) lig-
ands can track microglial activation to provide insights
into inflammatory mechanisms, and label biochemical
Comments

hallmarks of late neurodegeneration, such as amyloid

and tau deposition. Translation is simply one aspect of
the role that biomarkers can play in optimizing
approaches to rehabilitation and recovery research.21
The previous consensus position—that no cognitive
biomarkers are yet ready for use in phase III trials,
with the possible exception of lesion anatomy-based
prediction of language outcome—remains intact.
However, identifying and evaluating sensitive cognitive
Object-in-scene memory
Paired associate learning

biomarkers, to support translational research, is a

Not tested in preclinical

developmental priority. One promising area to begin

Morris water maze
Preclinical paradigms

biomarker development is chronic reactive astrogliosis

Fear conditioning

stroke models

within white matter tracts, which is linked to delayed

impairment in memory retrieval in humans and
rodents.22

Premorbid function and functional reserve

Pre-stroke cognitive performance is thought to influ-
ence cognitive outcome after stroke, including the risk
Object-in-scene memory
Paired associate learning

of developing future dementia. However, accurate

(human adaptation)

ascertainment of premorbid ability is challenging.9

Morris water maze

Little studied in post

Further, demographic factors (e.g. education) may

Examples of existing successful cross-species translation are highlighted in bold type.

also contribute to altered cognitive outcomes following

Human paradigm

stroke setting

stroke. There is no validated method to assess cognitive

reserve although multivariate approaches to neuroima-
ging data are beginning to reveal information about
overall brain status, or ‘‘brain age.’’ Application in a
stroke rehabilitation setting is, however, complicated
by the structural and functional alterations induced
by injury, which are known to extend well beyond
Associative learning

sites of visible infarction. The potential of multivariate

Emotional memory

approaches to address the difficult questions in stroke

Spatial memory

rehabilitation21 is an exciting area of future research.

Subdomain

Integration of patient- and carer-reported

outcome measures
Much attention has focused on defining optimal object-
ive testing approaches for post-stroke cognitive impair-
Table 2. Continued

ment. However, more work is required to link this

domain (DSM-5)

Social cognition
Neurocognitive

approach to stroke outcome as defined by patients,

relatives and carers. Understanding the associations
between cognitive function and QoL after stroke is
essential both in setting research priorities—across the
translational spectrum—and defining the health

International Journal of Stroke, 14(8)

780
economics benefits of new interventions. Technology
provides new opportunities for integration of objective,
patient- and carer-reported outcome measures. For
example, touchscreen tablets and smartphones can deli-
ver cognitive tests and prompt reporting by patients
and carers. In addition, the development of virtual real-
ity provides a new way to test cognition in a more eco-
logically valid manner.23 Wearable devices can provide
information on natural behavior (locomotion) and
information about factors that modulate cognitive per-
formance, such as sleep.24 The integration of patient-
and carer-reported and technology-derived information
with more traditional evaluation of cognition presents
an opportunity for recovery research.
Candidate therapies for cognitive
rehabilitation
The approaches we are interested in directly target cog-
nitive impairments themselves (e.g. executive functions)
and not aids (e.g. pagers) to improve daily real-world
functioning without changing cognitive processing.
There are several detailed reviews relating to this
topic,25,26 so here we confine ourselves to outlining
some key issues. Firstly, it is difficult to isolate individual
cognitive functions in terms of measuring outcomes (e.g.
impairments in working memory and attention can both
impact performance on tests of executive function).
Secondly, interventions need to be delivered in high
enough doses to maximize the likelihood of clinically
meaningful gains. A way to do this is to augment thera-
pist-delivered, face-to-face training with digital carer-
delivered therapies. This has shown promise in studies
designed to improve: working memory27; goal process-
ing and sustained attention.28 While not all cognitive
interventions and tests currently have direct equivalents
in animal models (Table 2), there are synergies in cellular
and genetic mechanisms that mediate higher cognitive
functions across species. For example, stroke induces
an elevation in tonic GABA signaling and compounds
that dampen this response have shown promise in
animal models for motor recovery29 and are currently
being tested in a Phase II trial (ClinicalTrials.gov ID;
Servier RESTORE BRAIN Study—NCT02877615).
These compounds have also been tested in a preclinical
model of VCI and shown to improve working memory.
Similarly, Brain Derived Neurotrophic Factor (BDNF),
which plays an important role in regulating plasticity,
decreases with age and negatively impact on recovery.
Aerobic exercise training can elevate BDNF levels and
this has been implicated as mediating improved spatial
memory in healthy older adults; training increased hip-
pocampal volume, effectively reversing age-related loss
by one to two years.30 A recent phase II clinical trial in
patients with post-stroke cognitive impairment showed
International Journal of Stroke, 14(8)
International Journal of Stroke 14(8)
that exercise paired with cognitive training improved
fluid intelligence, but the relationship to BDNF was
less clear.31 It is here that animal models provide a
much more fine-grained approach to understanding the
intricacies of the cellular and genetic substrates that
underpin human cognition,32 allowing us to be better
positioned to test interventions for cognitive recovery
studies in patients.
Conclusions
Research on cognitive recovery after stroke is at an ear-
lier stage of evolution than research in motor recovery.
Nevertheless, international consensus is possible in a
number of areas. All clinical stroke recovery studies
should integrate cognitive evaluation and outcome into
their design. Preclinical basic neuroscience studies are
essential for developing new interventions to enhance
recovery. In order to achieve this, greater alignment
between preclinical and clinical research—and the devel-
opment of an agenda of shared priorities—is required to
accelerate progress towards novel therapies. This is best
achieved using a bedside to bench to bedside approach.
Acknowledgement
The authors disclose receipt of the following financial support
to conduct this meeting: Canadian Institutes of Health
Research (CIHR) CaSTOR (Canadian Stroke Trials for
Optimized Results) Group (note that CaSTOR is a joint ini-
tiative of the Canadian Stroke Consortium and the Canadian
Partnership for Stroke Recovery), Heart and Stroke
Canadian Partnership for Stroke Recovery, and NHMRC
Centre of Research Excellence in Stroke Rehabilitation and
Brain Recovery. An unrestricted educational grant was pro-
vided by Ipsen Pharma. We would also like to acknowledge,
Julie Bernhardt for convening the second Stroke
Rehabilitation and Recovery Roundtable (SRRR II), Dale
Corbett and Karen Borschmann for organizing the meeting
and Farrell Leibovitch for moderating discussions.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.
ORCID iD
Matthew W McDonald https://orcid.org/0000-0002-0171-
6102
Tracy D Farr https://orcid.org/0000-0002-6781-5226
Matthew S Jeffers https://orcid.org/0000-0002-4148-2638
Michael J O’Sullivan https://orcid.org/0000-0002-2869-
4580
McDonald et al.
Notes
This contribution, first published in International Journal of
Stroke, is being co-published in the following journals:
Neurorehabilitation and Neural Repair.
References
1. Bernhardt J, Hayward KS, Kwakkel G, et al. Agreed def-
initions and a shared vision for new standards in stroke
recovery research: the stroke recovery and rehabilitation
roundtable taskforce. Int J Stroke 2017; 12: 444–450.
2. Hachinski V. Vascular dementia: a radical redefinition.
Dementia 1994; 5: 130–132.
3. Nys GM, van Zandvoort MJ, de Kort PL, Jansen BP, de
Haan EH and Kappelle LJ. Cognitive disorders in acute
stroke: prevalence and clinical determinants. Cerebrovasc
Dis 2007; 23: 408–416.
4. Nys GM, van Zandvoort MJ, de Kort PL, et al. The
prognostic value of domain-specific cognitive abilities in
acute first-ever stroke. Neurology 2005; 64: 821–827.
5. Brookes RL, Willis TA, Patel B, Morris RG and Markus
HS. Depressive symptoms as a predictor of quality of life
in cerebral small vessel disease, acting independently of
disability: a study in both sporadic small vessel disease
and CADASIL. Int J Stroke 2013; 8: 510–517.
6. Pendlebury ST, Rothwell PM and Oxford Vascular S.
Incidence and prevalence of dementia associated with
transient ischaemic attack and stroke: analysis of the
population-based oxford vascular study. Lancet Neurol
2019; 18: 248–258.
7. Quinn TJ, Paolucci S, Sunnerhagen KS, et al. Evidence-
based stroke r-ehabilitation: an expanded guidance docu-
ment from the European Stroke Organisation (ESO)
guidelines for management of ischaemic stroke and tran-
sient ischaemic attack 2008. J Rehabil Med 2009; 41:
99–111.
8. Sachdev PS, Blacker D, Blazer DG, et al. Classifying
neurocognitive disorders: the DSM-5 approach. Nat
Rev Neurol 2014; 10: 634–642.
9. McGovern A, Pendlebury ST, Mishra NK, Fan Y and
Quinn TJ. Test accuracy of informant-based cognitive
screening tests for diagnosis of dementia and multido-
main cognitive impairment in stroke. Stroke 2016; 47:
329–335.
10. Sachdev P, Kalaria R, O’Brien J, et al. Diagnostic criteria
for vascular cognitive disorders: a VASCOG statement.
Alzheimer Dis Assoc Disord 2014; 28: 206–218.
11. Mijajlovic MD, Pavlovic A, Brainin M, et al. Post-
stroke dementia – a comprehensive review. BMC Med
2017; 15: 11.
12. Skrobot OA, Black SE, Chen C, et al. Progress toward
standardized diagnosis of vascular cognitive impairment:
guidelines from the vascular impairment of cognition
classification consensus study. Alzheimers Dement 2018;
14: 280–292.
13. Hachinski V, Iadecola C, Petersen RC, et al. National
Institute of Neurological Disorders and Stroke-
Canadian Stroke Network vascular cognitive impairment
harmonization standards. Stroke 2006; 37: 2220–2241.
781
14. Horsburgh K, Wardlaw JM, van Agtmael T, et al. Small
vessels, dementia and chronic diseases – molecular mech-
anisms and pathophysiology. Clin Sci (Lond) 2018; 132:
851–868.
15. Pendlebury ST, Klaus SP, Thomson RJ, et al.
Methodological factors in determining risk of dementia
after transient ischemic attack and stroke: (iii) applicabil-
ity of cognitive tests. Stroke 2015; 46: 3067–3073.
16. Kagan A, Black SE, Duchan FJ, Simmons-Mackie N and
Square P. Training volunteers as conversation partners
using ‘‘supported conversation for adults with aphasia’’
(SCA): a controlled trial. J Speech Lang Hear Res 2001;
44: 624–638.
17. Carmichael ST. Cellular and molecular mechanisms of
neural repair after stroke: making waves. Ann Neurol
2006; 59: 735–742.
18. Murphy TH and Corbett D. Plasticity during stroke
recovery: from synapse to behaviour. Nat Rev Neurosci
2009; 10: 861–872.
19. Corbett D, Carmichael ST, Murphy TH, et al. Enhancing
the alignment of the preclinical and clinical stroke recov-
ery research pipeline: consensus-based core recommenda-
tions from the stroke recovery and rehabilitation
roundtable translational working group. Int J Stroke
2017; 12: 462–471.
20. Tanila H. Testing cognitive functions in rodent disease
models: present pitfalls and future perspectives. Behav
Brain Res 2018; 352: 23–27.
21. Boyd LA, Hayward KS, Ward NS, et al. Biomarkers of
stroke recovery: consensus-based core recommendations
from the stroke recovery and rehabilitation roundtable.
Int J Stroke 2017; 12: 480–493.
22. Chen A, Akinyemi RO, Hase Y, et al. Frontal white
matter hyperintensities, clasmatodendrosis and gliovascu-
lar abnormalities in ageing and post-stroke dementia.
Brain 2016; 139: 242–258.
23. Plancher G, Tirard A, Gyselinck V, Nicolas S and Piolino
P. Using virtual reality to characterize episodic memory
profiles in amnestic mild cognitive impairment and alz-
heimer’s disease: influence of active and passive encoding.
Neuropsychologia 2012; 50: 592–602.
24. Wams EJ, Wilcock GK, Foster RG and Wulff K. Sleep-
wake patterns and cognition of older adults with amnestic
mild cognitive impairment (aMCI): a comparison with
cognitively healthy adults and moderate alzheimer’s dis-
ease patients. Curr Alzheimer Res 2017; 14: 1030–1041.
25. Weicker J, Villringer A and Thone-Otto A. Can impaired
working memory functioning be improved by training? A
meta-analysis with a special focus on brain injured
patients. Neuropsychology 2016; 30: 190–212.
26. Bogdanova Y, Yee MK, Ho VT and Cicerone KD.
Computerized cognitive rehabilitation of attention and
executive function in acquired brain injury: a systematic
review. J Head Trauma Rehabil 2016; 31: 419–433.
27. Westerberg H, Jacobaeus H, Hirvikoski T, et al.
Computerized working memory training after stroke – a
pilot study. Brain Inj 2007; 21: 21–29.
28. Levine B, Schweizer TA, O’Connor C, et al.
Rehabilitation of executive functioning in patients with
International Journal of Stroke, 14(8)
782
frontal lobe brain damage with goal management train-
ing. Front Hum Neurosci 2011; 5: 9.
29. Clarkson AN, Huang BS, Macisaac SE, Mody I and
Carmichael ST. Reducing excessive GABA-mediated
tonic inhibition promotes functional recovery after
stroke. Nature 2010; 468: 305–309.
30. Erickson KI, Voss MW, Prakash RS, et al. Exercise train-
ing increases size of hippocampus and improves memory.
Proc Natl Acad Sci USA 2011; 108: 3017–3022.
31. Ploughman M, Eskes GA, Kelly LP, et al. Synergistic
benefits of combined aerobic and cognitive training on
fluid intelligence and the role of IGF-1 in chronic
stroke. Neurorehabil Neural Repair 2019; 33: 199–212.
32. Schmidt-Wilcke T, Fuchs E, Funke K, et al. Gaba-from
inhibition to cognition: emerging concepts. Neuroscientist
2018; 24: 501–515.
33. O’Sullivan M, Morris RG and Markus HS. Brief cogni-
tive assessment for patients with cerebral small vessel dis-
ease. J Neurol Neurosurg Psychiatry 2005; 76: 1140–1145.
34. Houlton J, Zhou LYY, Barwick D, Gowing EK and
Clarkson AN. Stroke-induces a BDNF-dependent
improvement in cognitive flexibility in aged mice.
Neural Plasticity 2019; 1460890.
International Journal of Stroke, 14(8)
International Journal of Stroke 14(8)
35. Bonnelle V, Leech R, Kinnunen KM, et al. Default mode
network connectivity predicts sustained attention deficits
after traumatic brain injury. J Neurosci 2011; 31:
13442–13451.
36. Langner R, Willmes K, Chatterjee A, Eickhoff SB and
Sturm W. Energetic effects of stimulus intensity on pro-
longed simple reaction-time performance. Psychol Res
2010; 74: 499–512.
37. Grottick AJ and Higgins GA. Assessing a vigilance dec-
rement in aged rats: effects of pre-feeding, task manipu-
lation, and psychostimulants. Psychopharmacology
(Berl) 2002; 164: 33–41.
38. Li K and Malhotra PA. Spatial neglect. Pract Neurol
2015; 15: 333–339.
39. Mundy ME, Downing PE, Dwyer DM, Honey RC and
Graham KS. A critical role for the hippocampus
and perirhinal cortex in perceptual learning of scenes
and faces: complementary findings from amnesia and
FMRI. J Neurosci 2013; 33: 10490–10502.
40. Moodley K, Minati L, Contarino V, et al. Diagnostic
differentiation of mild cognitive impairment due to alz-
heimer’s disease using a hippocampus-dependent test of
spatial memory. Hippocampus 2015; 25: 939–951.