Prevention of post-stroke cognitive decline: a review focusing on lifestyle interventions

Abstract

Background and purpose

Despite a high prevalence of post-stroke cognitive impairment, therapeutic possibilities are still limited. Stroke and dementia share the same cluster of modifiable risk factors. Thus, lifestyle interventions and strict adherence to medication may not only decrease the risk of recurrent stroke but also the risk of post-stroke cognitive decline.

Methods

We performed a systematic literature search for randomized clinical trials (RCTs) targeting modifiable risk factors for the prevention of cognitive decline following stroke.

Results

We identified 25 non-pharmacological interventions and eight multiple risk factor interventions in stroke patients using cognition as outcome measure. None of the published trials investigated interventions aimed at the prevention of post-stroke cognitive decline. However, a number of ongoing trials aim at risk factor reduction and include measures on cognition.

Conclusion

Evidence for risk factor modification for the prevention of cognitive decline after stroke is scarce and comes mainly from observational studies. There is a need for more RCTs targeting the prevention of post-stroke dementia using lifestyle interventions and a multiple risk factor approach.

Introduction

Following the first year post-stroke, the incidence of new onset dementia ranges from 7% to 29% [1-3]. After stroke, the risk of developing dementia doubles, and the progression rate from mild cognitive impairment to dementia increases [2-4]. Depending on stroke subtype and the criteria used, 17–92% of stroke patients have mild cognitive impairment at 3 months post-stroke [5-7]. A cognitive decline of 0.8–2 points on the MMSE scale was reported in the first year after stroke [8, 9]. Despite the high prevalence of post-stroke cognitive impairment and the massive burden associated with it, no treatment of post-stroke dementia has been established so far. However, the treatment of vascular risk factors offers possibilities for prevention. Obesity, metabolic syndrome, physical inactivity, hypertension, unhealthy nutrition hypercholesterolaemia, heavy alcohol consumption and smoking have been found to be associated with cognitive decline and dementia including both the vascular and the Alzheimer type [10-13]. These factors are a well-known cluster of modifiable risk factors for stroke and other vascular diseases. Whereas a recent review [14] focused on pharmacological intervention for the prevention and the treatment of post-stroke cognitive impairment, this review will focus on lifestyle and multiple risk factor interventions. We performed a systematic literature search for published or registered randomized clinical trials (RCTs) aiming at the prevention of cognitive decline following stroke by targeting modifiable vascular risk factors.

Methods

We searched PubMed and Cochrane Central Register of Controlled Trials (CENTRAL) for studies published between 1966 and March 2011. Additionally, we searched ClinicalTrials.gov, International Standard Randomised Controlled Trial Number Register (ISRCTN), Australian New Zealand Clinical Trials Registry (ANZCTR) and International Clinical Trial Registry Platform (ICTRP) of the WHO for ongoing trials. We used the following search terms: stroke, cognition, cognitive, dementia, cognitive impairment, cognitive decline, prevention, lifestyle, risk factors, nutrition, diet, vitamins, physical activity, exercise, smoking, education, attention, memory, executive, problem solving and reasoning. We included RCTs and randomized crossover studies that had outcome variables measuring cognition, cognitive impairment or dementia and that included at least 10 adults and more than 50% of participants with a diagnosis of stroke. We excluded cognitive impairment related to other diseases such as traumatic brain injury, brain tumour, Parkinson's disease, multiple sclerosis. We included studies using non-pharmacological interventions (i.e. physical activity, diet, cognitive training and smoking cessation) and excluded studies restricted to pharmacological interventions. Finally, we performed a hand search of the references of the included studies.

Results

We screened 5986 protocols of registered clinical trials and 1680 abstracts of published studies (Fig. 1). Finally, 16 published and 17 ongoing studies were included in this review; 25 were non-pharmacological (Table 1) and eight multiple risk factor interventions (Table 2). Except for two studies testing the effects of vitamins, all published studies were small (Table 1). Half of the studies were subject to possible bias owing to inadequate or unclear randomization (four studies) and to unblinded outcome assessment (four studies), or because statistics were not corrected for testing several cognitive measures (eight studies). Studies aimed mainly at stroke rehabilitation or for secondary stroke prevention. Cognitive test instruments were highly variable; however, the majority of studies used extended neuropsychological test batteries. The effect of the intervention was measured after a minimum of 20 min and a maximum of 2 years.

Figure 1. Flow chart of selected studies.

Table 1. Non-pharmacological interventions

Trial	Intervention	Control	n	Measures related to cognition	Main inclusion criteria	Results	Country	Agea	Sexb	Time since strokea	Ischaemic stroke	Length of Follow-up	Risk of bias
PO, primary outcome variable; SO, secondary outcome variable; TMT, trail making test; MMSE, Mini Mental State Examination; ADAS-cog, Alzheimer's Disease Assessment Scale-cognitive subscale. ^a Mean; ^b Women/men.
15	20 min on treatmill	20 min review of home exercise	21	PO: TMT-A, TMT-B, symbol digit test, paced auditory addition test after intervention	Age: ≥16 6 months – 5 years post-stroke; first stroke, at least grade 2–7 of motor control on the Chedoke-McMaster Impairment Inventory for Arm and Hand; MMSE > 24	Treadmill exercise did not enhance change in cognitive performance compared with control	Canada	61	8/13	20  months	19	7–10 days	High
NCT00228306 16	8-week stationary bicylce training	8-week stretching exercise	38	PO: Executive function at 8 weeks, 16 weeks (Wisconsin Card Sorting Test, Stroop, TMT-A, TMT-B)	Single ischaemic stroke occuring 6–72 months prior Fugl-Meyer score (upper + lower extremity) ≥45; MMSE > 23	Improvement in cognitive domains related to motor learning but not in executive functions	USA	62	21/17	4.9 years	NA	16 weeks	Low
PROMOTE NCT01027858 17	6-month aerobic-based exercise training	Standard care	70	PO: ADAS-Cog, EXIT-25 and ADCS-ADL at 6 and 12 months SO: specific executive processes at 6 and 12 months	Age: ≥45 Both: cognitive syndrome and small vessel ischaemic disease; MoCA < 26; MMSE > 20	Ongoing	USA					12 months
ACTRN-12606000479505	12-month lower limb group exercise class	12-month upper limb and cognitive exercise	350	SO: MoCA at 12 months	One or more strokes; able to walk 10 m independently with or without aid; MMSE ≥ 24	Not published	Australia					12 months
NCT00827827	3-month lower-extremity exercise – strength training	3-month upper and lower body stretching	52	SO: Cognitive function at 3 months	Age: 40–85 Stroke greater than 6 months prior with residual hemiparetic gait	Ongoing	USA					3 months
ICARE NCT00871715	10-week accelerated skill acquisition	10-week usual arm therapy	360	SO: Short blessed memory test, D- KEFS verbal fluency test, Hopkins verbal learning test – revised, digits span backward, Color TMT 1 and 2 at 1 year	Age: ≥21; stroke within the last 106 days Hemiparesis in an upper extremity	Ongoing	USA					12 months
VISP 27	High-dose vitamin B6 + vitamin B12 + folic acid for 2 years	Low-dose vitamin B6 + vitamin B12 + folic acid for 2 years	3680	SO: MMSE ≥ 2 points decrease at 1 year and 2 years	Age: ≥35; non- disabling ischaemic stroke; modified Rankin Stroke Scale ≤3; onset ≤120 days before randomization	No differences in MMSE scores or in ≥2 points decrease between the groups	USA, UK, Canada	66	1379/2301	71 days	3680	24 months	Low
THRIVED 28	(i) Folic acid +  vitamin B12 (ii) Vitamin B6 (iii) Riboflavin (iv) Folic acid + vitamin B12 + riboflavin (v) Folic acid + vitamin B12 + vitamin B6 (vi) Riboflavin +  vitamin B6 (vii) Folic acid + vitamin B12 + vitamin B6 + riboflavin for 12 weeks	(viii) Placebo for 12 weeks	185	Telephone interview for cognitive status at 6 and 12 months Letter Digit Coding Test at 12 months	Age: >65 Ischaemic vascular disease (history of angina pectoris, acute myocardial infarction, evidence of major ischaemia or previous infarction, ischaemic stroke, TIA, intermittent claudication, previous surgery for ischaemic vascular disease)	Oral folic acid + vitamin B12 decreased homocysteine concentrations but no beneficial effects on cognition	UK	74	97/88	NA	112	12 months	Low
VITATOPS ISRCTN74743444	Folate + Vitamin B6 + B12	Placebo	8164	SO: Dementia	Within 7 months post-stroke (ischaemic, haemorrhagic) or TIA	No data published on cognition	20 countries; 4 continents	63	2946/5218	NA	5.703	Median: 3.4 years	Low
ChiCTR-TRC- 00000440	Butylphthalide	Placebo	240	PO: ADAS-Cog SO: MMSE	Age: ≥50 Ischaemic cerebrovascular disease confirmed by imaging (stroke lesions, white matter ischaemic lesions); cognitive dysfunction, dementia standard CDR score ≤0.5	Not published	China					NA
43	4-week memory strategy training	4-week non- specific memory training	12	Name-face paired associated memory test, Stylus maze test, 15 words test, Oxford recurring faces Test, memory questionnaire after intervention	3–5 months since stroke	Training programme group better than control on trained tasks but not on control tasks; no differences on everyday memory	Netherlands	51	NA	NA	NA	4 weeks	Unclear
44	4-week training (i) Process- oriented memory training group (ii) Compensation strategies	4-week low- dose memory training	62	General intelligence, attention, word fluency, verbal memory, short term and working memory, visual spatial memory, logical memory after intervention	Age: 30–81 Organic memory disorder of acute and recent onset	High-intensity memory training groups superior to low intensity. Process- oriented memory focus superior to teaching compensation strategies	Germany	60	23/39	5.8 months	Stroke: 41	4 weeks	High
45	Active virtual environment exploration	Passive virtual environment observation	48	Spatial layout and object recognition test after intervention	Age: 25–85 Vascular brain injury	Active participation in virtual environment enhances memory for spatial layout. No group differences for object recognition	UK	61	NA	NA	48	<1 day	Unclear
ReMIND ISRCTN92582254	(i) Restitution group (ii) Compensation group	(iii) Attention placebo group	180	PO: Everyday memory questionnaire SO: Rivermead Behavioural memory Test – extended version (RBMTE), Doors & people, memory aids questionnaire	Age >18 Stroke, TBI, or MS: memory deficit, defined as overall profile score of 2 to 0 on the RBMTE >1 month post- injury and or diagnosis	Not published	UK					NA
46	Pager for 7 weeks	Crossover waiting list	36	Patients and carers defined memory- related goals. Percentage tasks achieved during the last 2 weeks of intervention	Memory and/or planning impairment	Pager use was associated with significant improvements in everyday function compared to control. After withdrawl of pager return to pre-treatment levels	UK	47	12/24	3.3 years	14	16 weeks	Unclear
47	3-week computer-assisted reaction training	Cross over waiting list	27	Several attention tests, as well as more general cognitive tasks after intervention and after another 6 weeks	4–36 week post- stroke, left hemisphere	Benefits on perceptual speed and selective attention, no effects on vigilance or general cognitive functioning	Germany	51	8/19	16 weeks	NA	12 weeks	Unclear
START 48	4-week attention process training	Standard care	66	PO: Computerised Integrated visual & auditory continuous performance test at 5 weeks and 6 months SO: Bells test; TMT A/B; Stroop test; paced auditory serial addition test (PASAT); neuropsychologi cal assessments of memory and language	Age ≥ 18 First-ever- stroke; within 2 weeks of stroke; attention deficit	Greater improvement in some attention measures compared to control	New Zealand	69	31/47	19 days	68	6 months	Low
49	5-week computerized training on working memory tasks	No training	18	WAIS revised-NI – Span board; Stroop; Claeson- Dahl (word list recall test); Digit span from WAIS R; Raven's progressive matrices; Word list delayed recall; PASAT version A; RUFF 2&7 (serial cancellation test) and self- rating on cognitive functioning in daily life (CFQ) after 5 weeks	Age: 30–65 Stroke 12–36  months ago self-reported deficits in attention	Training effects on the non-trained tests for working memory and attention	Sweden	55	6/12	20 months	13	5 weeks	High
50	7-week goal management training	7-week brain health workshop	19 (11 stroke)	Battery of standardized and experimental tests of executive functioning and attention after intervention and at 4 months	Age: 23–70 Focal brain injury stabilized in the chronic phase of recovery (minimum 6 months post-injury) and executive functioning problems	Outcome data indicated specific effects of GMT on the Sustained Attention to Response Task as well as the Tower Test, a visuospatial problem-solving measure that reflected far transfer of training effects	Canada	49	5/14	NA	NA	4 months	High
51	3-week mental imagery programme	Conventional rehabilitation	46	SO: Color TMT after therapy and at 1 month	Age ≥ 60 First unilateral cerebral infarction; independent in daily activities before admission	Mental imagery improve ability to retain trained tasks and transfer the skills to other untrained tasks	Hong Kong	72	24/22	14 days	46	1 month	Low
52	3-week mental imagery programme	Conventional rehabilitation	33	Performances on the trained and untrained tasks in a ‘novel simulated environment’ after 3 weeks	First stroke; unilateral infarction within the middle cerebral artery system; ≥7 days post-stroke	Mental imaging group better in trained and untrained tasks in ‘novel environment’ than control	Hong Kong	70	13/20	12 days	33	3 weeks	Low
NCT01304017	Virtual reality therapy	Traditional therapy	70	SO: Executive functions route finding test at 6 months	Age: 18–85 At least 6 months post-stroke; experiencing difficulty using their legs and/or arms; able to walk 10 meters (with or without assistance)	Ongoing						6 months
ISRCTN84064492	Conductive education intervention for 10 weeks	Cross over, waiting list	120	SO: Test of everyday attention; Wisconsin card sorting test at 3 and 6 months	Age ≥ 18; stroke and in the post-acute stage of recovery	Ongoing						6 months
53	(i) Music group (ii) Language group for 2 months	(iii) No listening material	55	Extensive neuropsychological test battery (verbal memory, short-term and working memory, language, visuospatial cognition, music cognition, executive functions, focused attention and sustained attention) at 3 and 6 months	Age ≤75; acute ischaemic MCA stroke in the left or right temporal, frontal, parietal or subcortical regions, right-handed	Larger improvement in verbal memory and focused attention in music group versus other groups	Finland	59	26/29	6.2 days	55	6 months	Low
ISRCTN03493594	3-month music therapy	Crossover, waiting list	45	SO: extensive neuropsychological test battery covering verbal and visual memory functions, working memory, visual perception, attention, executive functions and basic verbal functions at 3, 4 and 7 months	Age ≤75; middle cerebral artery stroke in the right temporal, parietal frontal and/or subcortical brain region; no prior clinically diagnosed cerebral stroke; <21 days post-stroke; right-handed	Ongoing	Finland					7 months

Table 2. Multiple risk factor interventions

Trial	Intervention	Control	n	Measures related to cognition	Main inclusion criteria
PO, primary outcome variable; SO, secondary outcome variable; TMT, trail making test; MMSE, Mini Mental State Examination; ADAS-cog, Alzheimer's Disease Assessment Scale-cognitive subscale.
Prevent
NCT00885456	Exercise and education to induce physiological and behavioural changes	Standard care	200	SO: MoCA at 12 months	Age ≥ 17 first TIA or non-disabling stroke; NIHSS < 6; post-event <30 days; orientation to time, place and person and ability to follow simple three-step commands
NCT00536562	Cardiac rehabilitation program (multi-disciplinary intervention targeting risk factors and exercise training)	Usual Care	200	Improvement in cognition at 6 months	TIA or mild non-disabling stroke within the previous 3 months
NCT00929994	Individualized cardiac rehabilitation (health education and aerobic and resistance training)	Usual Care	40	SO: MoCA and MRI analysis at 6 and 12 months	TIA within the last 3 months
NCT00495248	Lifestyle programme including physical activity with focus on social participation	Physical activity at the senior centre	140	SO: TM-A, TM-B at 6 and 9 months	Age ≥ 65 Clinical stroke; MMSE > 23 of 30; Barthel ADL-index >14 before stroke
NCT00506818	Multifactorial vascular-risk-factor-intervention	Usual care	250	PO: Changes in TMT-A and 10-words-test at 12 months SO: dementia at 12 months	First ever stroke or TIA; no cognitive decline (IQCODE < 3.7)
ASPIS
NCT01109836	Multifactorial intervention targeting vascular risk factors	Usual care	200	PO: Cognitive decline; ADAS-cog at 24 months SO:Cognitive decline; ADAS-cog at 12 months, MMSE; change in each of five cognitive domains at 12 and 24 months	Age 40–80 ischaemic stroke within the last 3 months NIHSS on admission 1–14 mRS before stroke 0–2
Partners
ISRCTN07607027	Volunteer-facilitated risk factor modification program	Usual care	844	SO: Cognitive tests scores at 12 and 24 months	Age ≥ 18 year TIA or non-disabling stroke within 90 days of randomization; hypertension
ISRCTN55406009	Individually tailored process of case management, liaison, guidance and counselling, coaching, mentoring and occupational therapy	Usual care	52	SO: Cognition; postal questionnaire at 3 and 6 months	People of working age having suffered a stroke, working or in full time education at stroke onset

Physical activity

Two small trials [15, 16] investigated the effect of physical activity on cognition in stroke patients (Table 1). Both studies did not investigate the prevention of post-stroke cognitive decline but whether exercise can enhance executive control and thereby motor learning in patients with motor deficits; they found no improvement in executive functions. Four ongoing studies were identified (Table 1). Two of these aim at the rehabilitation of motor deficits and one at the prevention of falls; all three include cognition as secondary outcome variable. The fourth study is a small proof-of-concept study and will test whether 6 months of aerobic exercise can improve cognition in people with both cognitive syndrome and small vessel ischaemic disease [17]. In a non-controlled pilot study, this group found improvement in memory and executive function in 11 participants with chronic stroke after 6 month of physical exercise [18].

Other evidence for the effect of physical activity on cognition

A large number of RCTs investigated the effects of physical activity for the prevention of cognitive decline and dementia. Usually, these studies exclude patients with previous stroke. A meta-analysis reviewing the effect of exercise in people with cognitive impairment or dementia included 10 studies and found a moderate effect of physical exercise on cognition [19]. A Cochrane review including older people without known cognitive impairment analysed 11 studies and found significant positive effects of aerobic exercise on cognitive speed and visual attention compared to other interventions, and positive effects for auditory attention and motor function when compared to no intervention [20]. However, 11 cognitive domains were compared and no effect was found in any of the other categories. A recent meta-analysis included 29 studies on non-demented adults and found moderate positive effects of aerobic exercise for attention and processing speed, executive function and memory, but not for working memory [21].

Diet

No published or ongoing study tested the effect of diet for the prevention of cognitive decline in stroke patients.

Other evidence for the effect of diet on cognition

Evidence for a positive effect of diet on cognition derives mainly from observational studies; RCTs are scarce with inconsistent results possibly owing to their shorter follow-up times [22]. Beneficial effects of the Mediterranean diet on health were found in observational studies, and only in one of a total of 41 RCTs cognition was measured [23]. This study found no effect of diet on cognition [24]. In the ENCORE study testing the effect of the Dietary Approaches to Stop Hypertension (DASH) diet in hypertensive overweight individuals, participants underwent a neurocognitive test battery [25]. After 4 months, participants on DASH diet combined with a behavioural weight management programme showed greater improvement in executive functions and psychomotor speed. None of the 124 participants had a history of stroke. A review on the effect of diet on cognition identified three small RCTs on caloric restriction in overweight adults [22]. One of these trials found improvement in verbal memory after 3 months but not in attention or working memory.

Nutritional supplements

Three studies tested the effect of supplementation with folic acid, vitamins B6 and B12 on cognition of stroke survivors [26-28]. However, the main aim of these studies was secondary stroke prevention, and so far, only two studies have published the results on cognition. No effect of vitamin supplementation on cognitive function was found [27, 28]. Additionally, an unpublished Chinese RCT tested the effect of NPD (Butylphthalide) in patients with cognitive impairment and a history of ischaemic cerebrovascular disease.

Other evidence for the effect of nutritional supplements on cognition

Reviews on the effect of vitamins B6, B12 and folate on cognition found no consistent evidence [29-32]. A Cochrane review found no evidence for the effect of vitamin E for the treatment for Alzheimer's disease (AD) and MCI in two trials [33]. No evidence for omega fatty acids for the prevention of cognitive decline was found in five RCTs – three in mild to moderate AD [34-36] and two in older, cognitive healthy subjects [37, 38]. The effect of flavonoids on cognitive function was mainly studied in small RCTs, investigating the effect of soy isoflavones in post-menopausal women [39]. The methods used differ greatly; however, there are indications for a positive effect of flavonoids on executive functions and working memory. Finally, a Cochrane review found no consistent evidence for Ginkgo bilboa in dementia and cognitive impairment [40].

Cognitive training

The majority of trials investigating cognitive interventions in stroke patients are small and aim at the rehabilitation of cognitive deficits and not at the prevention of a future cognitive decline [41, 42]. Post-stroke cognitive decline is mainly found in executive functions, speed of mental processing, working memory and general memory; therefore, we will focus on these domains.

Four small published studies aimed at improving memory in stroke patients [43-46]. Only one of these studies was not restricted to patients with memory deficits [45]. Despite the variety of interventions (memory training, compensatory strategies, external memory aid, virtual training), all four studies found an effect on memory when compared to controls. Additionally, one unpublished study aims to test two types of cognitive rehabilitation techniques for memory deficits (ISRCTN92582254).

Three small RCTs aimed at the rehabilitation of deficits in attention or working memory [47-49]. All three studies found training effects for attention and working memory compared to no-training groups.

One RCT tested the effect of goal management in patients with deficits in executive functions [50]. Improvement in the goal management group compared to control was found in two of eight measurements. One ongoing study is testing virtual reality intervention for stroke rehabilitation (NCT01304017); however, it aims at increasing physical activity, which includes also one measure for executive functions. Two trials from the same working group tested mental imaging for relearning daily tasks. Compared to conventional occupational therapy, mental imaging seems better in improving task learning, and the transfer to untrained tasks and new environments [51, 52]. Another ongoing trial intends to test conductive education for the improvement of movements and the performance of daily tasks (ISRCTN84064492); outcome measures will include attention und executive functions.

Finally, music listening was found to enhance verbal memory and focused attention at 3 and 6 months post-stroke [53]. An ongoing study (ISRCTN03493594) will test whether active music therapy over 3 months can improve post-stroke recovery, including cognitive recovery.

Other evidence for the effect of cognitive training on cognition

A Cochrane review on cognitive training in healthy older people and in people with MCI considered only data on memory training sufficiently available [54]. Cognitive training was found to be effective to improve memory compared to no-treatment controls.

Smoking cessation

Despite the increased risk of dementia and cognitive decline associated with smoking [55], no RCT investigated the long-term effect of smoking cessation on cognition – neither in stroke patients nor in other populations. However, one non-randomized trial in older people participating in a smoking cessation programme found a significant larger decline in ADAS-cog score over 24 months for unsuccessful quitters compared to successful quitters and never-smokers [56].

Multiple risk factor intervention

Current research emphasizes that risk factors occur often together and that treatment should be multifactorial and individualized. Guidelines on ischaemic stroke treatment recommend such a multifactorial approach for secondary prevention [57]. And indeed, we identified 11 published and 27 ongoing RCTs investigating multiple risk factor intervention after stroke. Interventions were mainly education and motivation to enhance lifestyle changes and compliance with prescribed medications. Unfortunately, none of the published trials has included cognition as outcome variable. Eight of the ongoing trials include cognition as outcome; however, only in two studies, it is designated as primary outcome variable (ASPIS and NCT00495248, Table 2).

Furthermore, no RCT aiming at the prevention of dementia by multiple risk factor intervention and lifestyle changes has been published. However, four large ongoing trials are testing the effect of multiple lifestyle intervention for the prevention of dementia or cognitive decline in people at risk ([10], NCT01012947).

Discussion

Despite its high prevalence, so far, no RCT has tested the effect of lifestyle interventions or multiple risk factor modification for the prevention of cognitive decline after stroke. Similarly, a recent review on factors preventing cognitive decline in older adults found mainly evidence from observational studies, and the only factor with strong positive evidence was cognitive training [58]. The majority of clinical trials measuring the effect of any intervention on cognition in stroke patients include either cognition as secondary outcome variable and aim at secondary stroke prevention or address rehabilitation of evident cognitive deficits after stroke.

Data on prevalence of cognitive deficits after stroke and on the rate of cognitive decline – especially in mild, non-disabling strokes – are still needed. However, the systematic use of more sensitive test instruments indicates that the number of affected individuals is larger than previously thought [6, 7, 59]. The mechanisms leading to post-stroke dementia are not sufficiently understood, but current evidence suggests that underlying vascular processes are ongoing after stroke and increase the risk for cognitive decline [2, 7]. Stroke, vascular dementia and AD share the same cluster of risk factors and occur often together. In the absence of known direct connections, the vascular risk profile is a candidate for a common underlying mechanism [13]. Thus, the evidence gained from dementia prevention trials may apply to stroke patients as well.

Beside its effect for the prevention of dementia and cognitive decline in elderly people [19-21], physical activity seems to enhance cognitive plasticity [60]. In stroke rehabilitation, these effects have only been studied for motor learning. Similarly, cognitive training has been shown to have some effect on cognition in healthy and cognitively impaired persons [54]. In stroke, cognitive training is only used in people with cognitive deficits. Physical exercise and cognitive training might be promising candidates not only for rehabilitation but also for the prevention of cognitive decline in people with mild, non-disabling strokes. Further research should also include the effect of smoking cessation and diet on cognition. In the absence of other therapeutics, multiple risk factor interventions including lifestyle changes and help with adherence to drugs for blood pressure and lipid lowering may offer the possibility to stop the process of post-stroke cognitive decline before reaching a point of no return. These interventions can be implemented in the early subacute phase of stroke, thus triggering long-term behaviour changes as early as possible.

Performing clinical trials in the area of long-term post-stroke behaviour modification should consider not only an adequate sample size but also an extended follow-up period. According to our experience, a 12-month follow-up should be considered a minimum. The high variability in stroke-specific manifestations makes this population very heterogeneous and the choice of adequate cognitive test instruments difficult. Screening tests such as the MOCA are sensitive to detect cognitive deficits but may not be sensitive to detect small changes over time and do not include information on speed of processing [61]. If practicable, primary outcome variables should involve extended neuropsychological test batteries comprising all cognitive domains possibly affected by vascular cognitive decline such as executive functions, speed of processing, attention short-term memory and memory, similar to the National Institute of Neurological disorders and Stroke-Canadian Stroke Network Harmonization Standards Neuropsychological Battery [59].

Conclusion

Preventive measures that aim at maintaining cognitive functions at normal levels in patients who have suffered from stroke are not well established, and clinical trials testing such measures are needed. The development of such measures should be seen at least as important as developing measures aiming at repairing mental functions in post-stroke cognitively impaired patients. There is a need for well-designed large trials testing the effect of multiple interventions intending to adapt and improve lifestyle and medication adherence upon neuropsychological variables over a wide range of cognitive domains. Whilst a few first trials are under way that test selective components with a potential for cognitive improvement or preservation of cognitive function, additional issues such as the effect of multiple interventions, their timing, duration or intensity still have to be determined.

Disclosure of conflict of interest

The authors declare no financial or other conflict of interests.