Post-stroke cognitive impairment is common even after successful clinical recovery

ORIGINAL ARTICLE Post-stroke cognitive impairment is common even after successful clinical recovery H. Jokinen, S. Melkas, R. Ylikoski, T. Pohjasvaara, M. Kaste, T. Erkinjuntti and M. Hietanen Clinical Neurosciences, Neurology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland EUROPEAN JOURNAL OF NEUROLOGY Keywords: cognition, cognitive impairment, neuropsychology, stroke Received 20 January 2015 Accepted 6 April 2015 European Journal of Neurology 2015, 22: 1288 1294 doi:10.1111/ene.12743 Background and purpose: Cognitive impairment is common after stroke, but the prevalence and long-term significance of the diverse neuropsychological deficits on functional outcome are still not well known. The frequency and prognostic value of domain-specific cognitive impairments were investigated in a large cohort of ischaemic stroke patients. Methods: Consecutive patients (n = 409), aged 55 85 years, from the acute stroke unit of the Helsinki University Hospital, Finland, were evaluated with extensive clinical and neuropsychological assessments 3 months post-stroke. Impairments within nine cognitive domains were determined according to ageappropriate normative data from a random healthy population. Functional disability was evaluated with the modified Rankin scale (mrs) 3 and 15 months post-stroke. Results: In all, 83% patients showed impairment in at least one cognitive domain, whereas 50% patients were impaired in multiple ( 3) domains. In cases with excellent clinical recovery at 3 months (mrs = 0 1, no disability), the occurrence of any cognitive impairment was 71%. Memory, visuoconstructional and executive functions were most commonly impaired. A substantially smaller proportion of patients scored below the conventional or more stringent cut-offs in the Mini-Mental State Examination (MMSE). Domain-specific cognitive impairments were associated with functional dependence at 15 months regardless of stroke severity and other confounders. Conclusions: Cognitive impairment as evaluated with a comprehensive neuropsychological assessment is prevalent in stroke survivors even with successful clinical recovery. Typically multiple domains and complex cognitive abilities are affected. MMSE is not sensitive in detecting these symptoms. Post-stroke cognitive impairment is strongly related to poor functional outcome. Introduction Cognitive impairment is a common consequence of stroke and one of the major determinants of poor long-term outcome [1,2]. Even milder cognitive deficits resulting from stroke can affect patients quality of life, independent functioning and occupational abilities. Earlier studies have reported varying prevalence estimates for post-stroke cognitive impairment due to Correspondence: H. Jokinen-Salmela, Unit of Neuropsychology, Department of Neurology, Helsinki University Hospital, PO Box 302, 00029 HUS, Helsinki, Finland (tel.: +358947173861; fax: +358947174088; e-mail: hanna.jokinen@helsinki.fi). differences in the sample characteristics, assessment methods, definitions of impairment and time interval since stroke onset. Studies using short screening tests of global cognitive dysfunction have typically yielded relatively low prevalence rates, below 25% [3,4], whilst more detailed neuropsychological assessments of domain-specific cognitive impairments have suggested higher occurrence, ranging from 35% to 92% [5 8]. Post-stroke cognitive impairment is not a unitary syndrome but incorporates a variety of deficits in multiple domains such as attention, executive functions, memory, language and visuoperceptual abilities. Traditionally, clinical outcomes in stroke studies have 1288

POST-STROKE COGNITIVE IMPAIRMENT 1289 been evaluated with measures of physical recovery or focal cortical syndromes such as aphasia and neglect, whilst the broad spectrum of cognitive changes has been overlooked. Prospective follow-up studies of the long-term functional impact of domain-specific cognitive impairments are sparse [1,9,10]. Moreover, little is known about the occurrence of cognitive impairment in stroke patients with seemingly good clinical recovery and no physical handicap. The aim of this study was to investigate the prevalence of domain-specific cognitive impairments 3 months after ischaemic stroke in a large representative cohort and in a subgroup of patients with excellent clinical recovery (no significant disability). Cognitive impairment detected in an extensive neuropsychological assessment was compared to that identified by the Mini-Mental State Examination (MMSE). Furthermore, the predictive value of the domain-specific cognitive deficits on functional disability was tested after 15 month follow-up. Methods Subjects and study protocol In the Helsinki Stroke Aging Memory Study, 486 patients consecutively admitted to the acute stroke unit of the Helsinki University Hospital, Finland, were examined 3 months after ischaemic stroke. The baseline recruitment was conducted between December 1993 and March 1995, and the follow-up extended to 15 years. The study procedure and the characteristics of the patients have been described in full previously [11,12]. In summary, the examinations included comprehensive clinical neurological and neuropsychological assessments, and brain magnetic resonance imaging. Of the vascular risk factors, hypertension was defined at the time of study inclusion as systolic blood pressure 160 mmhg and diastolic blood pressure 95 mmhg. Diabetes was defined as previously documented diagnosis, current use of insulin or oral hypoglycaemic medication or fasting blood glucose >7.0 mmol/l. The clinical severity of stroke was evaluated with the National Institutes of Health Stroke Scale (NIHSS) [13], functional abilities with the modified Rankin Scale (mrs) [14] and post-stroke depressive symptoms with the Beck Depression Inventory (BDI) [15]. Excellent clinical recovery was defined as mrs = 0 1 indicating no symptoms at all or no significant disability despite symptoms (able to perform all usual duties and activities). Further, global cognitive status was evaluated with the MMSE using either <24/30 or <27/30 as the cut-off for cognitive impairment [16]. Dementia was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV) criteria [17] and pre-stroke cognitive decline on the basis of all available history from the patient and from the knowledgeable informant. In total, 409 patients participated in the neuropsychological assessment described below after excluding cases with severe hearing or sight disabilities, reduced level of consciousness, and those unwilling to participate. Compared to the included cases, the excluded cases (n = 77) were older (73.7 vs. 70.7 years, P = 0.002) and they had higher NIHSS (5.5 vs. 3.0, P = 0.004) and lower MMSE scores (23.6 vs. 25.3, P = 0.035), but there were no significant differences in education or sex. The majority (n = 332, 81.2%) of the included patients had experienced a first-ever stroke, 69 patients had a history of one previous stroke, and eight had more than one previous stroke. The characteristics of the patients are shown in Table 1. After 15 months, the evaluation of functional abilities with mrs was repeated for 364 of 409 patients either at a follow-up visit (n = 287) or by phone (n = 77). At follow-up, 38 patients had died, six patients refused to participate and one could not be reached. At this stage, functional dependence was defined as mrs = 3 5 indicating moderate to severe disability and need of assistance in daily activities. The study was approved by the Ethics Committee of the Department of Clinical Neurosciences, Helsinki Table 1 Characteristics of the patients participating in the neuropsychological assessment in the Helsinki Stroke Aging Memory Study (n = 409) Age, mean (SD) 70.7 (7.7) Sex, female, n (%) 207 (50.6) Education, years, mean (SD) 9.2 (4.1) Hypertension, n (%) 199 (48.7) Diabetes, n (%) 96 (23.5) Prior ischaemic stroke, n (%) 77 (18.8) NIHSS score, n (%) 0 185 (45.2) 1 4 139 (34.0) 5 15 67 (16.4) 16 20 13 (3.4) 21 42 5 (1.2) mrs score, n (%) 0 1 152 (37.2) 2 3 184 (45.0) 4 5 73 (17.8) MMSE, mean (SD) 25.4 (4.4) BDI, mean (SD) 9.5 (7.2) Post-stroke dementia, DSM-IV, n (%) 68 (17.6) Pre-stroke cognitive decline (%) 48 (11.7) BDI, Beck Depression Inventory; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders IV; MMSE, Mini-Mental State Examination; mrs, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale.

1290 H. JOKINEN ET AL. Table 2 Neuropsychological test battery of the Helsinki Stroke Aging Memory Study Executive functions and attention Trail making test, parts A and B Stroop test, colour naming and interference Wisconsin card sorting test Verbal fluency, phonemic and semantic Digit span, forward and backward Memory WMS-R logical memory, story A, immediate and delayed recall WMS-R visual reproduction, immediate and delayed recall Fuld object memory evaluation, 10-item learning test, 5 trials and delayed recall Visuoconstructional and spatial functions WAIS-R block design Figure copying test Clock drawing test Abstract thinking WAIS-R similarities Aphasia Token test Boston naming test Boston diagnostic aphasia examination, overall speech evaluation Reading and writing Samples Arithmetic Arithmetic operations Neglect Bells test Agnosia Poppelreuter See Appendix S1 for details and references. WAIS-R, Wechsler Adult Intelligence Scale revised; WMS-R, Wechsler Memory Scale revised. University Hospital, Finland. All subjects received and signed an informed written consent. Neuropsychological assessment The neuropsychological assessment was administered by a trained neuropsychologist at a single session an average of 105 (SD 14.8) days after the index stroke [18]. The battery included 19 established cognitive tests grouped into nine cognitive domains (Table 2). Standard instructions and scoring were used. Details and the original references for the individual tests are given in Appendix S1. Impairment within each domain was judged in three age groups using normative data from a random Finnish population without any neurological diseases [19]. Abnormality was defined as a test performance 2 SD below the level of the norm in domains assessed with a single test, or 1 SD below the normative level in several tests in domains assessed with two or more tests [12,19]. Data on impairment were not available for 39 (10%) cases in executive functions/attention, 30 (7%) in memory, 43 (11%) in abstract thinking, 14 (3%) in Figure 1 Number of cognitive domains impaired 3 months poststroke in the Helsinki Stroke Aging Memory Study; total n = 409 visuoconstructional functions, 14 (3%) in aphasia, 50 (12%) in arithmetic, 37 (9%) in reading and writing, 26 (6%) in agnosia and 43 (11%) in neglect. Missing data occurred mainly due to patients inability to complete the entire test battery. Data analysis Comparisons between groups were made with the chisquared test, the independent samples t test or the Mann Whitney U test, where appropriate. Factors associated with cognitive impairments were studied with multiple logistic regression analyses entering demographic characteristics, key risk factors and clinical findings to the models simultaneously. The predictive value of each domain-specific cognitive impairment on functional dependence (mrs = 3 5) in the 15-month follow-up was tested with logistic regression analyses adjusting for confounders. Results Cognitive impairment in the whole cohort In all, 339 of 409 (83%) patients showed impairment in at least one cognitive domain. Eighty (20%) patients had impairment in one domain, 56 (14%) patients in two and 203 (50%) patients in multiple ( 3) domains (Fig. 1). Most frequently, impairments were found in memory, and visuoconstructional and executive functions, following deficits in language and other verbal functions with smaller proportions (Table 3). Neglect and agnosia were relatively uncommon in this sample. The most frequent combinations of deficits were observed between memory and visuo-

POST-STROKE COGNITIVE IMPAIRMENT 1291 Table 3 Frequencies of impairment in each cognitive domain in the whole cohort and in a subgroup of patients with excellent clinical recovery (mrs = 0 1) 3 months after ischaemic stroke in the Helsinki Stroke Aging Memory Study Cognitive domain n (%) All patients n = 409 Memory functions 227 (60) 77 (52) Visuoconstructional 216 (55) 54 (36) and spatial functions Executive functions and 181 (49) 52 (34) attention Aphasia 114 (29) 23 (15) Reading and writing 112 (30) 16 (11) Abstract reasoning 106 (29) 26 (17) Arithmetic 71 (20) 12 (8) Neglect 29 (8) 2 (1) Agnosia 13 (3) 1 (1) constructional functions (147 cases, 39%), executive and visuoconstructional functions (140 cases, 38%), and executive functions and memory (133 patients, 37%). Relative to these, the other combinations remained less frequent (<90 cases). Of the total sample, 94 (24%) patients scored <24 and 192 (50%) patients scored <27 in the MMSE. Cognitive impairment in a subgroup of patients with excellent clinical recovery In patients with excellent clinical recovery 3 months post-stroke (mrs = 0 1), any cognitive impairment was found in 108 of 152 cases (71%). One domain was affected in 38 (25%) cases, two in 22 (15%) cases and multiple domains in 48 (32%) cases. Similarly to the whole cohort, the most common impairments and combinations were found in memory, and visuoconstructional and executive functions (Table 3). However, only nine (6%) patients of this subgroup scored <24 and 47 (31%) <27 in the MMSE. Compared to the rest of the cohort (mrs >1), the patients with good clinical outcome were significantly younger (68.6 vs. 72.0 years, P < 0.001), had higher education (10.1 vs. 8.6 years, P < 0.001), were more often men (89 vs. 63, P = 0.004) and had lower NIHSS (0.4 vs. 4.5, P < 0.001). After excluding cases with a history of previous clinical stroke (n = 21), the frequency of any cognitive deficits was still 69%. Factors associated with cognitive impairment Patients with mrs = 0 1 n = 152 Unequal numbers of cases are presented for each domain due to missing data (see text for details). mrs, modified Rankin Scale. The demographic and clinical correlates of cognitive impairment were investigated by focusing on the three most frequent cognitive impairments. In bivariate analysis, memory impairment was significantly associated with lower education (8.8 vs. 10.0 years, P = 0.005) and higher BDI score (10.5 vs. 8.0, P = 0.001), but not with age, sex, hypertension, diabetes, prior stroke or NIHSS score. Visuoconstructional deficits were related to higher age (72.1 vs. 68.7 years, P < 0.001), female sex (56.9% vs. 42.5%, P = 0.004), lower education (8.6 vs. 10.0 years, P < 0.001) and higher NIHSS score (3.5 vs.1.6, P < 0.001). Executive dysfunction was associated with higher age (71.8 vs. 69.1 years, P = 0.001), lower education (8.8 vs. 9.8 years, P = 0.021) and prior stroke (22.1% vs. 13.8%, P = 0.036), as well as higher NIHSS (3.2 vs. 1.5, P < 0.001) and BDI scores (10.3 vs. 8.6, P = 0.025). As analysed with multiple logistic regression entering all variables together, education and BDI score remained as independent predictors of memory impairment [odds ratio (OR) 0.94, 95% confidence interval (CI) 0.89 0.99, P = 0.027; OR 1.05, 95% CI 1.02 1.09, P = 0.003, respectively]. Age, education and NIHSS score predicted visuoconstructional deficits (OR 1.05, 95% CI 1.02 1.08, P = 0.001; OR 0.93, 95% CI 0.88 0.98, P = 0.010; OR 1.14, 95% CI 1.06 1.22, P < 0.001, respectively), whereas age, prior stroke and NIHSS score predicted executive dysfunction (OR 1.04, 95% CI 1.01 1.07, P = 0.005; OR 1.81, 95% CI 1.01 3.22, P = 0.046; OR 1.12, 95% CI 1.04 1.20, P = 0.002). Cognitive impairments associated with functional dependence at 15-month follow-up In the whole cohort, all domain-specific cognitive impairments analysed one by one, except arithmetic impairment, were significantly associated with functional dependence (mrs = 3 5, n = 192, 44%) at the 15-month follow-up independently of age, sex, years of education and NIHSS (Table 4). After adding BDI score as another covariate, the results remained unchanged. In addition to the cognitive impairments, poor functional outcome was significantly related to age, BDI score and NIHSS score (P < 0.05). These analyses were not run for neglect and agnosia because of small numbers of cases. Discussion The prevalence and long-term functional significance of domain-specific cognitive impairments were investigated in a consecutive cohort of 409 middle-aged and older patients 3 months after ischaemic stroke. In an extensive neuropsychological assessment, cognitive impairment was observed to be highly frequent both

1292 H. JOKINEN ET AL. Table 4 Cognitive impairments 3 months post-stroke as predictors of functional dependence (modified Rankin Scale >2) at 15-month follow-up in the Helsinki Stroke Aging Memory Study Impairment Logistic regression OR (P value) 95% CI for OR Memory functions 2.2 (0.008) 1.2 3.9 Visuoconstructional and spatial functions 5.1 (<0.001) 2.7 9.1 Executive functions and attention 3.2 (<0.001) 1.8 5.7 Aphasia 2.1 (0.017) 1.1 3.9 Reading and writing 2.3 (0.011) 1.2 4.3 Abstract reasoning 2.3 (0.006) 1.3 4.2 Arithmetic 1.9 (0.063) 1.0 3.9 Statistical analyses are adjusted for age, sex, years of education and stroke severity (National Institutes of Health Stroke Scale score). OR, odds ratio; CI, confidence interval. in the whole cohort (83%) and, strikingly so, also in a subgroup of patients with excellent clinical recovery and no functional disability at 3 months (71%). The widely used screening test MMSE, with either the conventional or more stringent cut-off, was only able to detect much lower frequencies of cognitive dysfunction (24% 50% and 6% 31%, respectively). The majority of the patients in both samples had impairments in more than one cognitive domain. These deficits were selectively associated with age, years of education, depression, prior stroke and stroke severity. Independently of the confounding factors, the domain-specific cognitive impairments were related to functional disability at 15-month follow-up. Our results support previous studies reporting a high overall occurrence of cognitive impairment in detailed evaluation of specific domains in a subacute stage of stroke, i.e. up to 3 months after stroke onset [5 8,10]. However, direct comparison of studies is difficult due to many methodological differences such as variation in timing of assessment, evaluated cognitive domains and test batteries, use of control group or normative data, definition of impairment as well as dissimilar settings and patient cohorts. In a pioneering study, Tatemichi and co-workers found 78% of older stroke patients failing at least one cognitive test, whilst 35% were impaired in four or more tests [5]. Recent studies have reported even 76% 92% of patients demonstrating impairment in one or more cognitive domains [7,8,10]. In addition to the overall prevalence of cognitive impairment, there is also some variability in previous studies in the proportions of the impaired domains. Executive functions, attention and processing speed are included in the most commonly impaired domains in different studies [5 8,10,20,21]. As in our study, impairments in long-term (episodic) memory [5,7] and visuoconstructional and spatial functions [6,8] have also been commonly reported. Taken together, cognitive impairment seems to occur most frequently in multiple cognitive domains, typically affecting complex cognitive abilities, in which executive control functions and focused attention have a major role. These processes are also determining factors in complex visuoconstructional functions and in effective memory encoding and retrieval strategies. Impairment in long-term memory, visuoconstruction and executive functions, as well as combinations of these three deficits, were the most common findings in our study. Compared to these, impairments in abstract reasoning, language, reading, writing and arithmetic were less common, although they still occurred in 20% 30% of the patients. Neglect was relatively uncommon occurring only in 8% of all patients, which contrasts some of the earlier findings [6]. A simple cancellation task was used in our assessment, which is widely used in an acute setting but may not be sensitive in detecting residual neglect in the post-acute stage. An important result of the present study is the finding that even the patients with the most favourable clinical outcome (mrs = 0 1), thus having no apparent functional disability, demonstrated a wide spectrum of cognitive deficits similar to the whole cohort. Thus far, there have been few studies investigating cognitive symptoms with detailed neuropsychological assessments amongst clinically well-recovered patients. These patients are typically discharged after short acute care and are expected to return to their previous lives without major difficulties. Planton and co-workers have reported a 40% overall rate of cognitive impairment 3 months post-stroke in 60 patients with good clinical outcome (defined by mrs 2 and NI- HSS 3) [22]. Furthermore, in a subgroup of stroke patients with no apparent neurological symptoms (NI- HSS = 0), Kauranen and co-workers observed cognitive impairment in 41% of patients at 1 2 weeks evaluation and 32% of patients at 6 months followup [23]. The patients of both these studies were younger compared to our study, which could explain their lower figures. The likelihood of pre-stroke cognitive decline is increased in the older age groups. Moreover, these studies only enrolled patients who had experienced their first symptomatic ischaemic stroke, whilst all consecutive patients below the mrs cut-off ( 1) were included here. Excluding patients with previous stroke did not, however, essentially alter our results. As far as is known, this is the largest neuropsychological study investigating post-stroke cognitive impairment. A representative cohort of consecutive patients from a unit responsible for acute stroke management in the city of Helsinki was collected prospectively. In a stable post-acute stage of stroke, extensive

POST-STROKE COGNITIVE IMPAIRMENT 1293 neuropsychological assessments were carried out with established clinical tests and performance was compared with a single source age-matching normative data. The relationship between cross-sectional neuropsychological data and functional outcome was confirmed at 15 month follow-up. A minority of patients were not able to complete all neuropsychological tests, which may cause some selection bias in the results and possible under-representation of cases with the poorest cognitive outcome. Furthermore, a significant proportion of patients in our cohort had low NIHSS scores indicating that patients with mild to moderate ischaemic stroke were over-represented compared to those with severe stroke. In conclusion, cognitive impairment was found in the vast majority of late middle-aged and older patients 3 months after ischaemic stroke, and it was very frequent even in patients with seemingly successful clinical recovery and no functional disability. Typically, complex cognitive abilities, including long-term memory encoding and retrieval, focused attention and executive functions as well as visuoconstructional abilities, were compromised. These impairments were significantly related to poor functional outcome during long-term follow-up. MMSE proved to be insensitive in detecting post-stroke cognitive impairment. The results emphasize the importance of detailed cognitive assessment as part of routine clinical evaluation of stroke patients. Acknowledgements The study was supported by grants from the Clinical Research Institute and the Medical Research Fund of the Helsinki University Central Hospital. Disclosure of conflicts of interest Dr Kaste reports grants from Helsinki University Central Hospital and other support from Lundbeck A/S, Siemens AG, Boehringer Ingelheim, Mitsubishi Pharma Europe Ltd, outside the submitted work. Supporting Information Additional Supporting Information may be found in the online version of this article: Appendix S1. Neuropsychological test battery of the Helsinki Stroke Aging Memory Study. References 1. 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. 2. Oksala NK, Jokinen H, Melkas S, et al. Cognitive impairment predicts poststroke death in long-term follow-up. J Neurol Neurosurg Psychiatry 2009; 80: 1230 1235. 3. Liman TG, Heuschmann PU, Endres M, Floel A, Schwab S, Kolominsky-Rabas PL. Changes in cognitive function over 3 years after first-ever stroke and predictors of cognitive impairment and long-term cognitive stability: the Erlangen Stroke Project. Dement Geriatr Cogn Disord 2011; 31: 291 299. 4. Douiri A, Rudd AG, Wolfe CD. Prevalence of poststroke cognitive impairment: South London Stroke Register 1995 2010. Stroke 2013; 44: 138 145. 5. Tatemichi TK, Desmond DW, Stern Y, Paik M, Sano M, Bagiella E. Cognitive impairment after stroke: frequency, patterns, and relationship to functional abilities. J Neurol Neurosurg Psychiatry 1994; 57: 202 207. 6. Nys GM, van Zandvoort MJ, de Kort PL, Jansen BP, de Haan EH, Kappelle LJ. Cognitive disorders in acute stroke: prevalence and clinical determinants. Cerebrovasc Dis 2007; 23: 408 416. 7. Jaillard A, Naegele B, Trabucco-Miguel S, LeBas JF, Hommel M. Hidden dysfunctioning in subacute stroke. Stroke 2009; 40: 2473 2479. 8. Middleton LE, Lam B, Fahmi H, et al. Frequency of domain-specific cognitive impairment in sub-acute and chronic stroke. NeuroRehabilitation 2014; 34: 305 312. 9. van Zandvoort MJ, Kessels RP, Nys GM, de Haan EH, Kappelle LJ. Early neuropsychological evaluation in patients with ischaemic stroke provides valid information. Clin Neurol Neurosurg 2005; 107: 385 392. 10. Lesniak M, Bak T, Czepiel W, Seniow J, Czlonkowska A. Frequency and prognostic value of cognitive disorders in stroke patients. Dement Geriatr Cogn Disord 2008; 26: 356 363. 11. Pohjasvaara T, Erkinjuntti T, Vataja R, Kaste M. Dementia three months after stroke. Baseline frequency and effect of different definitions of dementia in the Helsinki Stroke Aging Memory Study (SAM) cohort. Stroke 1997; 28: 785 792. 12. Pohjasvaara T, Erkinjuntti T, Ylikoski R, Hietanen M, Vataja R, Kaste M. Clinical determinants of poststroke dementia. Stroke 1998; 29: 75 81. 13. Goldstein LB, Bertels C, Davis JN. Interrater reliability of the NIH stroke scale. Arch Neurol 1989; 46: 660 662. 14. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1988; 19: 604 607. 15. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4: 561 571. 16. Folstein MF, Folstein SE, McHugh PR. Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189 198. 17. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), 4th edn. Washington, DC: American Psychiatric Association, 1994. 18. Jokinen H, Kalska H, Mantyla R, et al. Cognitive profile of subcortical ischaemic vascular disease. J Neurol Neurosurg Psychiatry 2006; 77: 28 33.

1294 H. JOKINEN ET AL. 19. Ylikoski R, Ylikoski A, Keskivaara P, Tilvis R, Sulkava R, Erkinjuntti T. Heterogeneity of cognitive profiles in aging: successful aging, normal aging, and individuals at risk for cognitive decline. Eur J Neurol 1999; 6: 645 652. 20. Hurford R, Charidimou A, Fox Z, Cipolotti L, Werring DJ. Domain-specific trends in cognitive impairment after acute ischaemic stroke. J Neurol 2013; 260: 237 241. 21. Rasquin SM, Verhey FR, Lousberg R, Winkens I, Lodder J. Vascular cognitive disorders: memory, mental speed and cognitive flexibility after stroke. J Neurol Sci 2002; 203 204: 115 119. 22. Planton M, Peiffer S, Albucher JF, et al. Neuropsychological outcome after a first symptomatic ischaemic stroke with good recovery. Eur J Neurol 2012; 19: 212 219. 23. Kauranen T, Laari S, Turunen K, Mustanoja S, Baumann P, Poutiainen E. The cognitive burden of stroke emerges even with an intact NIH Stroke Scale Score: a cohort study. J Neurol Neurosurg Psychiatry 2014; 85: 295 299.