Journal of the Neurological Sciences 346 (2014) 241–249

Contents lists available at ScienceDirect

Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns

Profile and determinants of vascular cognitive impairment in African stroke survivors: The CogFAST Nigeria Study Rufus O. Akinyemi a,b,⁎,1, Louise Allan b, Mayowa O. Owolabi c, Joshua O. Akinyemi d, Godwin Ogbole e, Akinlolu Ajani a, Michael Firbank b, Adesola Ogunniyi c, Raj N. Kalaria b,⁎,1 a

Division of Neurology, Department of Medicine, Federal Medical Center Abeokuta, Nigeria Institute for Ageing and Health, Campus for Ageing and Vitality, Newcastle University, Newcastle upon Tyne, United Kingdom Department of Medicine, University College Hospital/College of Medicine, University of Ibadan, Ibadan, Nigeria d Department of Epidemiology and Medical Statistics, University College Hospital/College of Medicine, University of Ibadan, Ibadan, Nigeria e Department of Radiology, University College Hospital/College of Medicine, University of Ibadan, Ibadan, Nigeria b c

a r t i c l e

i n f o

Article history: Received 24 April 2014 Received in revised form 25 August 2014 Accepted 28 August 2014 Available online 1 September 2014 Keywords: Africa Neuropsychology Neuroimaging Stroke Vascular cognitive impairment Vascular dementia

a b s t r a c t Objective: Sub-Saharan Africa faces a potential epidemic of non-communicable diseases including stroke and dementia but little is known about the burden of stroke-related cognitive dysfunction. We assessed the baseline profile and factors associated with vascular cognitive impairment (VCI) in stroke survivors participating in the Cognitive Function After STroke (CogFAST) Nigeria Study. Methods: We recruited 217 subjects (N45 years old) comprising 143 stroke survivors and 74 demographically matched stroke-free healthy controls. We obtained demographic, clinical and lifestyle information and assessed the cognitive status of the subjects at baseline three months after stroke. Standard neuropsychological tests included the Vascular Neuropsychological Battery, which assessed executive function/mental speed, memory, language, and visuospatial/visuoconstructive functioning. Cognitive impairment and dementia were defined based on the AHA/ASA VCI guidelines and the DSM IV criteria. Results: Among the stroke survivors (mean age = 60.4 + 9.5 years, 43.4% female, mean number of years of education = 9.4 + 5.6 years, median modified Rankin score = 2), 57 (39.9%) had cognitive impairment no dementia while 12 (8.4%) were demented at baseline. Multivariate analysis revealed that older age [OR = 1.05 (1.00–1.09)], low education [OR = 5.09 (2.17–11.95)], pre-stroke cognitive decline [OR = 4.51 (1.20–16.88)] and medial temporal lobe atrophy [OR = 2.25 (1.16–4.35)] were independently associated with cognitive dysfunction whereas pre-stroke daily intake of fish [p = 0.022, OR = 0.39 (0.15–0.89)] was inversely associated. Conclusions: These results suggest a high frequency of early VCI in older Nigerian stroke survivors. Apart from aging, associated neurodegeneration and cognitive decline, educational level and pre-stroke diet particularly fish consumption were identified as modifiable factors. This emphasizes the vital role of education and healthy nutrition in building reserves to ameliorate cognitive dysfunction after stroke. © 2014 Elsevier B.V. All rights reserved.

1. Introduction It is estimated that one in three persons worldwide will develop a stroke, dementia or both during their lifetime [1,2]. Sub-Saharan Africa (SSA) is in epidemiologic transition and faces a growing burden of non-communicable diseases in addition to the persistent burden of infections, conflicts and poverty [3,4]. Aging populations, rapid urbanization and lifestyle changes appear to be the main drivers of the

⁎ Corresponding authors at: Institute for Ageing and Health, Campus for Ageing and Vitality, Newcastle upon Tyne, NE4 5PL, United Kingdom. Tel.: + 44 191 248 1352; fax: + 44 191 248 1301. E-mail address: [email protected] (R.N. Kalaria). 1 AO and RNK were equal contributors.

http://dx.doi.org/10.1016/j.jns.2014.08.042 0022-510X/© 2014 Elsevier B.V. All rights reserved.

burgeoning epidemic of vascular disorders (hypertension, diabetes, metabolic syndrome) which frequently culminate in stroke. Although physical disability is most commonly associated with stroke, cognitive changes and other non-motor consequences are quite frequent in survivors [5]. Post-stroke cognitive dysfunction encompasses a multi-domain impairment of attention and concentration, executive function, language, memory and visuospatial function with executive dysfunction being the earliest and predominantly affected domain [2,6,7]. In a lifetime, up to 64% of stroke survivors suffer some degree of cognitive impairment and about 30% develop dementia [8]. In a recent meta-analysis, the pooled prevalence estimates of poststroke dementia (PSD) within one year of stroke ranged from 7.4% (4.8–10.0) in population-based studies of first-ever stroke excluding pre-stroke dementia to 41.3% (29.6–53.1) in hospital-based studies of all strokes including pre-stroke dementia [9]. However, there is a wide

242

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249

variation in the definition and estimates of vascular cognitive impairment (VCI) owing to differences in diagnostic criteria, study populations and methodologies. These have led to recent harmonization efforts and criteria development [8]. The cognitive trajectory that ensues after stroke may be also influenced by patient-related variables such as age at stroke occurrence, aging-related brain shrinkage and cognitive decline, surrogates of cognitive reserve (educational attainment, occupational attainment, environmental enrichment), physical activity, and dietary and nutritional lifestyles. Cardiovascular risk factor load and stroke-related variables are also important determinants [5,10,11] while neuroimaging substrates include the number, size and sites of lesion, white matter changes, lacunar infarcts, strategic infarcts, cerebral microbleeds, medial temporal lobe atrophy and global cerebral atrophy [12–14]. Little is known, however, about the burden or risk factors of poststroke cognitive dysfunction in Africa. Patterned after our study in Newcastle [10], the aim in this first comprehensive prospective African project was to report the baseline frequency, and pattern, and identify factors associated with early VCI in Nigerian African stroke survivors participating in the Cognitive Function After STroke (CogFAST) Nigeria Study. We hypothesized that a high frequency, multidomain vascular cognitive impairment would be seen in a cohort of Nigerian African stroke survivors three months post-stroke. 2. Methods 2.1. Stroke participants Consecutively presenting stroke patients diagnosed by the most senior physician/attending neurologist were admitted to the medical wards of two specialist hospitals (Federal Medical Center Abeokuta and University College Hospital Ibadan, southwestern Nigeria) between July 2010 and June 2012. Subjects and their family/caregivers were approached regarding participation in the study at discharge from hospital or during the initial outpatient visit. Three months after the ictus, 220 stroke survivors were screened for eligibility. Stroke was classified using the WHO definition, the Oxford Community Stroke Project Classification (OCSP) [15] and neuroimaging (CT scan and/or MRI) findings [7]. However, neuroimaging confirmation was possible only in 61.5% of the cohort because of limited access and high cost in the study geographical area. The WHO clinical definition/criteria has been shown to have a sensitivity of 73% for hemorrhage, 69% for infarction and an overall accuracy of 71% in Nigeria [16]. Ischemic stroke was sub-typed based on the Oxfordshire Community Stroke Project (OCSP) classification (based on clinical presentation), neuroimaging (CT and/ or MRI) findings and the Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria [17]. Based on the pattern of clinical presentation, subjects were categorized into total anterior circulation infarct (TACI), partial anterior circulation infarct (PACI), lacunar infarct (LACI) and posterior circulation infarct (POCI) as described in the OCSP classification [15]. Etiological sub-typing was done in accordance with the TOAST criteria [17] into large vessel, small vessel, cardioembolic or stroke of undetermined cause based on the findings of neuroimaging and relevant cardiac investigations. Subjects classified as cardioembolic stroke had electrocardiographic (ECG) evidence of atrial fibrillation and/or echocardiographic evidence of thrombus formation within the cardiac chambers in addition to clinical and radiologic evidence and subtypes of the stroke. Exclusion criteria were: [i] age less than 45 years, [ii] subarachnoid hemorrhage, [iii] significant physical illness and motor impairment that precluded paper and computer-based neuropsychological evaluation (e.g. visual impairment, moderate-severe aphasia, hemiparesis affecting the dexterous hand, MRC power grade b 3), [iv] any co-morbid psychiatric or neurologic illness (background dementia, schizophrenia, major depression, manic-depressive disorder and Parkinson's disease), [v] any systemic disease that could impair cognition e.g. chronic liver

Fig. 1. Flow chart showing numbers screened and recruited and numbers of assessments at each year of the study, together with deaths and withdrawals.

disease, chronic kidney disease, and [vi] failure to give consent. Fig. 1 shows the total number and distribution of stroke patients logged onto the study register and screened for eligibility three months after the index stroke. 2.2. Stroke-free controls For comparison with the neuropsychological data from stroke survivors, apparently healthy controls that were free of clinically-evident stroke were recruited from a representative pool of community-dwelling volunteers participating in a community health literacy program. Controls were also recruited from among the spouses and unrelated caregivers of stroke survivors as previously done in some studies [16]. Individuals with background dementia (DSM IV criteria), psychiatric disorders e.g. schizophrenia, major depression, manic-depressive disorder; background neurological disorders e.g. Parkinson's disease (evidence from case records, informant or self-report), or who were unable to provide consent and/or information were excluded from being controls. Up to June 2012, 74 stroke-free controls of comparable age, gender and education were included in the analysis. The local health research and ethics committees granted approval for the study (Federal Medical Center Abeokuta and University College Hospital, Ibadan) while written informed consent was obtained from each subject. 2.3. Subject evaluation Subject evaluation was performed at three months post-stroke in tandem with the design of Desmond et al. [18] to enable the resolution of acute post-stroke delirium. The evaluation included comprehensive medical history, assessment of neurological impairment, disability and activities of daily living (using the modified Rankin Scale [19], Stroke Levity Score [20] and Barthel Index [19]), depressive symptoms (using the Centre for Epidemiologic Studies Depression Scale [21] and fouritem Geriatric Depression Scale [22]), and blood screens. The Stroke Levity Score is an assessment of stroke severity based on maximum power (0–5) in the dexterous hand + maximum power in the weaker lower limb + mobility score-1 (if aphasia present). It has shown excellent correlation with the National Institutes of Health Stroke Scale (NIHSS) [20]. Confirmation of cardiovascular risk factors (hypertension, diabetes mellitus, atrial fibrillation, dyslipidemia, cigarette smoking, and alcohol use) were based on self-report, use of relevant medications and review of medical notes.

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249

Lifestyle factors including nutritional status and physical activity were assessed according to the design of the INTERSTROKE Study [23]. Dietary patterns were assessed in all subjects with a food frequency questionnaire assessing the frequency of different types of food taken by the subjects in the prior year before the onset of stroke (and separately in the three months period after stroke) with appropriate addition of common local foods and delicacies to relevant categories. Physical activity was assessed at work and during leisure time and subjects were stratified into sedentary, mild, moderate and heavy physical activity categories [23,24]. All participants were interviewed in the presence of a spouse, family member or caregiver, who had lived with the subject for at least six months in order to validate dietary recall and other historical details provided by the subjects.

2.4. Cognitive assessment The cognitive assessment instrument consisted of the Community Screening Instrument for Dementia (CSID) — the cognitive part [24], the mini-mental state examination (MMSE) [25] and the Vascular Neuropsychological Battery (V-NB) [8]. Cognitive assessment was performed by experienced interviewers who received a further two weeks of training on the study instrument and had to achieve an inter-rater reliability of at least 90% in mock assessments done with volunteers from the hospital community before conducting cognitive assessment on the study subjects. The CSID and the MMSE are tests of general cognitive functioning while the V-NB consists of a battery of tests assessing functioning in specific domains of cognition. Pre-stroke cognitive status was assessed using the CSID-informant part [26] by trained interviewers. The CSID is a paper and pencil test of global cognitive performance with established adaptability, validity and utility in populations from different cultural, educational and socio-economic backgrounds [24, 25]. It has a sensitivity of 87% and a specificity of 83% for the diagnosis of dementia and has been used reliably and widely to assess cognition in the Yoruba speaking population of southwestern Nigeria wherein the present study was conducted [27]. The CSID included sub-scores for attention, orientation, calculation, short and long term memory, language comprehension and expression, praxis and abstract thinking. A raw score method was used for scoring resulting in a score range of 0–30 with higher scores indicating better cognitive function [28]. The Vascular Neuropsychological Battery [V-NB] was modeled after the NINDS–CSN Harmonization Standards 60-minute neuropsychological protocol [8] with minor modifications to ensure adaptability to the language and culture of the study population. The V-NB consisted of multiple validated test items examining specific cognitive domains (executive function, memory/learning, language, visuospatial/visuoconstructive skills). Executive function/activation and mental speed were assessed using the category (animal) fluency test [25], phonemic (letter) fluency test, verbal reasoning and visual reasoning tests which were adapted from the Cambridge Cognitive Examination (CAMCOG) battery previously utilized for the CogFAST-Newcastle Study [29,30]. The number of animals listed in the first 15 s of the animal fluency test provided an assessment of mental speed while all the tests differently assessed mental flexibility and divergent thinking [8]. Memory/learning was assessed with the 10-item word list learning test and delayed recall of stick design. The word list learning is a 3-trial, 10-item test with free recall taken after each learning trial and after a brief delay. The total number of words recalled across the three trials makes up the total score (range 0–30) while the delayed recall is scored 0–10, with higher scores indicating better performance [25,31]. Language was assessed through the 15-item Boston Naming Test. In a prior validation study of the CERAD battery among Yoruba Nigerians, subjects were requested to name line drawings of common and uncommon objects. Four of the low frequency items from the standard CERAD-NB were replaced with items felt to be more culturally appropriate (i.e. guitar for harmonica, blacksmith tongs for ice cube

243

tongs, mosquito netting for hammock, and ‘ayo’ (Nigerian board game) for dominoes) [25]. Visuospatial/visuoconstructive functioning was assessed through the Stick Design Test [31] and the Modified Token Test or IU Token Test [32,33]. The Stick Design Test is a non-graphomotor test of visuospatial/visuoconstructive ability. The respondent is requested to use match sticks to reproduce four different graphical shapes with particular attention to the correctness of the relative orientation of the match heads. Thereafter the respondent reproduces the four shapes without any cues to assist. The test is particularly useful in older adults with limited formal education [31]. Test items from the Cognitive Drug Research (CDR) computerized assessment battery were also included in the V-NB for the evaluation of attention, processing speed and executive function. The constituent tests included Simple Reaction Time (SRT), Choice Reaction Time (CRT), Digit Vigilance Reaction Time (VIGRT) and Spatial Working Memory Reaction Time (SPMRT) [34]. 2.5. Neuroimaging assessment (magnetic resonance imaging) Brain magnetic resonance imaging (MRI) was performed on a subset of stroke survivors with two MRI scanners operating between 0.2 T and 0.35 T (Fig. 2). All images were transferred to a computer workstation with a ClearCanvas DICOM viewer and evaluated by two experienced radiologists. All ratings were performed by consensus agreement. Medial temporal lobe atrophy (MTLA) (Fig. 3) and white matter changes were assessed using the Schelten's scales [35,36]. Assessment of brain volumes – total intracranial volume, total brain volume and ventricular volume – was performed by creating a mask using the Brain Extraction Tool (BET) from the FSL software (www.fmrib.ox.ac.uk/fsl/). 2.6. Cognitive diagnosis To make a cognitive diagnosis on a subject, all available datasets including cognitive scores, functionality and disability scores (the Barthel Index and modified Rankin score) coupled with the physician's assessment were assembled and discussed by the research team for consensus diagnosis. Functional impairment was defined as a Barthel Index score b 75 [37]. Final cognitive categorization was based on the VCI criteria proposed by the American Stroke Association/American Heart Association VCI Guidelines [2] and the DSM IV criteria (American Psychiatric Association, 1994). 2.7. Operational definitions of cognitive dysfunction Failure of an individual subject on a test was defined as a mean score that was at least 1.5 standard deviations below the mean score of the control group. Impairment in a domain was defined as failure on at least 50% of tests examining that particular domain [38]. Vascular MCI and PSD were defined according to the American Stroke Association/American Heart Association VCI Guidelines [2]. Vascular MCI or vascular CIND [2] was defined as impairment in at least 1 cognitive domain (executive function, memory/learning, language, visuospatial/visuoconstructive skills) and normal or mild impairment of activities of daily living independent of motor/sensory symptoms. Post stroke dementia (PSD) [2], in accord with the DSM IV criteria, was defined as impairment in N2 cognitive domains that were of sufficient severity to affect the subject's activities of daily living independent of motor/sensory symptoms [2]. 2.8. Statistical analysis Data were analyzed using the Statistical Package for Social Sciences version 19.0 (SPSS Inc., Chicago). Categorical variables were examined, analyzed and compared using the chi square test while continuous variables were described using measures of central tendency and compared using the Student's t-test and analysis of variance (ANOVA). Logistic regression models were used to determine univariate and multivariate

244

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249

Fig. 2. Magnetic resonance imaging (MRI) T1- and T2-weighted (A and B), and fluid-attenuated inversion recovery (C) axial images from a 70-year old female Nigerian stroke survivor showing moderate white matter abnormalities.

relationships between cognitive status and patient-related variables including demographic, lifestyle and vascular risk factors; stroke disability, depression symptoms and neuroimaging parameters. Unadjusted and adjusted ORs with 95% CIs were estimated. For multivariate analysis, variable groups were entered incrementally so that the mediating effect of each could be evaluated. Level of statistical significance was set at p b 0.05. Bonferroni correction was done for multiple pairwise comparisons across cognitive groups resulting in an adjusted level of significance, α (adjusted) = 0.0083 where applicable in order to reduce familywise Type 2 error. 3. Results We recorded 417 patients with a stroke, of whom 101 died in the hospital, 31 were discharged against medical advice and 65 were lost to follow up. The remaining 220 were screened at three months postictus for eligibility in the study, out of whom 145 met the selection criteria. Fig. 1 shows the flow chart of study participation in line with the STARD (Standards for Reporting of Diagnostic Accuracy) guidelines [39]. Seventy-five subjects were excluded due to: [i] severe aphasia and motor impairment precluding computer-assisted cognitive impairment (n = 18), [ii] b 45 years of age (n = 32) and [iii] non-consent (n = 25). The records of 2 stroke survivors were excluded from further analysis because of incomplete assessment details. There were no significant differences between the total cohort of stroke survivors who were screened (n = 220) and those who met the selection criteria and were logged into the final data analysis (n = 143) with respect to age [57.9 (13.1) versus 60.4 (9.5) years] (t = − 1.97, p = 0.05); gender: female [105 (47.7%) versus 62 (43.4%)] (χ2 = 1.25, p = 0.536) and stroke type [cerebral infarction/intracerebral hemorrhage/indeterminate (155/45/20 versus 114/17/12)] (χ2 = 4.78, p = 0.092). 3.1. Demographics and clinical profile of the participants Table 1 shows details of the sociodemographic and clinical profile of stroke survivors and the stroke-free apparently healthy controls. Although most of the subjects (80% of cases and 70% of controls) earned less than 50,000 naira (approximately 300 USD) per month, 93.4% of cases and 91.9% of controls had access to personal or proxy mobile phones. Survivors who had neuroimaging (brain CT scan and/or MRI) (n = 88) did not differ significantly from those who did not with respect to age [60.4 (9.4) vs 60.3 (9.8) years] (t = 0.061; p = 0.952), years of educational attainment [9.3 (5.6) vs 9.5 (5.6)] (t = 0.178, p = 0.859),

gender: female [45.5% vs 42.0%] (χ2 = 0.160, p = 0.689) and mean modified Rankin score [2.32 (1.1) vs 2.29 (1.0), t = 0.150, p = 0.881]. In subjects clinically diagnosed with stroke, who also underwent neuroimaging, the sensitivity was 97.9% and positive predictive value was 92.3% for a stroke diagnosis. Table 1 also shows further details of stroke types, subtypes and etiological classification in the subset that had neuroimaging. Seventy cases (79.5%) were ischemic and 18 (20.5%) were hemorrhagic strokes. Ischemic stroke subtypes based on the OCSP classification and TOAST etiological classification were: 44.4% partial anterior circulation, 41.4% lacunar infarction, 10.0% total anterior circulation infarction and 4.2% posterior circulation infarction. The etiologic classes based on the TOAST criteria were: 30.0% large vessel disease, 58.6% small vessel disease, 7.1% cardioembolic and 4.2% indeterminate respectively. However, in a subset of 42 cases that had CT scan within one week of presentation, 28 cases (66.7%) were ischemic and 12 cases (33.3%) were hemorrhagic showing that a higher relative proportion of hemorrhagic stroke was diagnosed with earlier neuroimaging. Across the evaluated cohort of 143 stroke survivors, most cases had mild disability (median modified Rankin score = 2), median stroke Levity score = 12 and good functional score (median Barthel Index = 80) 3 months after stroke. 3.2. Profile of cognitive performance At three months following the index stroke, 74 (51.7%) stroke survivors had no apparent cognitive impairment (NCI), 57 (39.9%) had VCI no dementia (vCIND) while 12 (8.4%) had PSD. The vCIND group included 19/57 (33.3%) single domain non-amnestic, 11/57 (19.3%) multiple domain non-amnestic, 5/57 (8.8%) single domain amnestic and 22/57 (38.6%) multiple domain amnestic. Pre-stroke cognitive impairment was detected in 16 (11.2%) stroke survivors. Table 2 shows the pattern of performance of the controls and different cognitive sub-groups of stroke survivors on the various test items assessing different domains of cognition while Fig. 4 shows the pattern of performance on Choice Reaction Time (CRT). 3.3. Comparison of stroke survivors with no vCIND and controls Compared to controls, stroke survivors without apparent vCIND were of comparable age (t = −0.76; p = 0.563), number of years of educational attainment (t = 0.63; p = 0.458) and performance on cognitive test items assessing the domains of memory, language and visuospatial/visuoconstructive functioning (Table 2). However, they

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249 Table 1 Socio-demographic and clinical profile of stroke survivors and stroke-free healthy controls. Characteristic

Socio-demographic profile Gender: n (%) female Age: (Mean ± SD) years Educational attainment (years) None Primary education (1–6) Secondary education (7–12) Tertiary education (N12) Marital status Married Widowed Separated Living situation Alone Nuclear family Extended

Stroke survivors

Healthy controls

N = 143

N = 74

62 (43.4) 27 (36.5) χ2 = 0.951; p = 0.329 60.4 ± 9.5 58.8 ± 8.6 0.136⁎ χ2 = 5.24; p = 0.264 21 (14.7) 8 (10.8) 39 (27.3) 13 (17.6) 35 (24.5) 18 (24.3) 48 (33.6) 35 (47.3) χ2 = 2.11; p = 0.348 122 (85.3) 65 (87.8) 17 (11.9) 9 (12.2) 4 (2.8) – χ2 = 4.74; p = 0.093 9 (6.8) 8 (11.1) 109 (82.6) 62 (86.1) 14 (10.6) 2 (2.8)

Clinical profile Cardiovascular risk factors, n (%) Hypertension 132 (93.6) Diabetes mellitus 34 (24.5) 2 Obesity (BMI N 30 kg/m ) 23 (16.1) Atrial fibrillation 5 (3.6) Previous TIA 7 (4.9) Previous stroke 18 (12.6) Ever smoked 28 (19.6) Ever used alcohol 49 (34.3) Moderate–strenuous 25 (17.5) physical activity Fish intake (at least weekly) 58 (40.6) Fruit intake (at least weekly) 59 (41.3) Depressive symptoms CESD score GDS-4 Stroke type by CT scan and/or MRI⁎⁎

Significance

6.3 + 4.8 0.7 + 1.1 n = 88

44 (59.5) 3 (4.1) 14 (18.9) – – – 5 (6.8) 33 (51.4) 5 (6.8)

χ2 = 37.80; p b 0.001 χ2 = 14.17; p b 0.001 χ2 = 3.85; p = 0.146 –

35 (47.3) 40 (54.1)

χ2 = 1.96, p = 0.582 χ2 = 8.88; p = 0.031

4.8 + 3.4 0.4 + 0.7

p = 0.025⁎ p = 0.248+

χ2 = 15.43 p = 0.017 χ2 = 19.34; p = 0.011 χ2 = 5.40; p = 0.067

Ischemic 70 (79.5%) Hemorrhagic 18 (20.5%) Stroke location Anterior circulation 83 (94.3) Posterior circulation 5 (5.7) Ischemic stroke subtype n = 70 (OCSP) OCSP classification Partial anterior circulation 31 (44.4) infarct Lacunar infarct 29 (41.4) Total anterior circulation 7 (10.0) infarct Posterior circulation 3 (4.2) infarct Presumed primary etiology of ischemic stroke (TOAST criteria) Small vessel 41 (58.6) Large vessel 21 (30.0) Cardioembolic 5 (7.1) Undetermined 3 (4.2) Disability score Modified Rankin Scale 2.32 (mean, SD)a (1.04) Stroke Levity Score 11.96 b (mean, SD) (2.23) Abbreviations: CESD = Centre for Epidemiologic Studies Depression Scale; GDS-4 = four item Geriatric Depression Scale. Significant p values are indicated in bold. a Modified Rankin Scale [19] at baseline 3 months after stroke. b Stroke Levity Score [20] at baseline 3 months after stroke. ⁎ Compared by Student's t-test. + Compared by Kruskal Wallis test; others by chi-square test. ⁎⁎ 88 stroke survivors who had CT scan and/or MRI.

performed worse than controls on cognitive measures of mental speed and executive function [animal fluency, t = − 3.17; p = 0.002 and Digit Vigilance Reaction Time (VIGRT), t = −3.58; p b 0.001] (Table 2).

245

3.4. Comparison of stroke survivors with no vCIND and those with vCIND Stroke patients meeting the criteria for vCIND were significantly older (t = 2.54; p = 0.007), less educated (t = − 5.02; p b 0.001) and more impaired in all cognitive domains compared to those without vCIND: executive function and mental speed (animal fluency t = −5.35 p b 0.001; letter fluency, t = −8.18; p b 0.001; CRT mean, t = −5.22; p b 0.001; VIGRT mean, t = − 3.42; p = 0.001); language (BNT, t = − 8.15, p b 0.001), memory (WLL, t = − 3.31; p = 0.001) and visuospatial/visuoconstructive functioning (Stick Design Test, t = − 8.15; p b 0.001; Modified Token Test, t = − 8.79; p b 0.001) (Table 2). 3.5. Comparison of survivors with vascular CIND and those with post-stroke dementia Stroke survivors with vCIND were comparable to those with PSD with respect to age (t = −1.03, p = 0.309), years of formal education (t = 0.671, p = 0.453) and cognitive performance in the domain of language (BNT, t = 1.22; p = 0.207) while the demented sub-group had worse performance in the cognitive domains of mental speed/ executive function (animal fluency, t = − 3.20; p = 0.003; letter fluency, t = −3.08, p = 0.003; SPMSI, t = −2.98; p = 0.004), memory (word list learning, t = − 4.05; p b 0.001) and visuospatial/ visuoconstructive functioning (Modified Token Test, t = − 3.33; p = 0.001). 3.6. Correlation of clinical, cognitive and neuroimaging variables Pre-stroke informant cognitive score showed significant correlation with post-stroke memory score (r = − 0.321, p = 0.022) and poststroke cognitive dysfunction (r = 0.199, p = 0.027). In a subset of stroke survivors with available brain MRI (n = 58), age correlated significantly with total brain volume (r = − 0.393, p = 0.004) and MTLA total score (r = 0.525, p b 0.001) while MTLA (Fig. 3) correlated significantly with total WMH score (r = 0.461, p = 0.002), total CSID score (r = − 0.378, p = 0.019), memory score (r = − 0.702, p b 0.001) and executive function score (r = − 0.369, p = 0.016) but not total brain volume (r = − 0.203, p = 0.157). Deep WMH frontal score correlated significantly with executive function (r = − 0.350, p = 0.013), Choice Reaction Time (r = 0.345, p = 0.015) and memory (r = −0.333, p = 0.021). Deep WMH parietal score correlated with memory (r = −0.502, p b 0.001) and executive function (r = −0.315, p = 0.026), while deep WMH temporal score correlated with executive function (r = −0.303, p = 0.033) but not with memory (r = −0.226, p = 0.123). The presence of hypertension correlated significantly with total WM score (r = 0.361, p = 0.001) and total deep WM score (r = 0.375, p = 0.007). Left parietal infarcts were also significantly associated with cognitive dysfunction as an outcome (r = 0.780, p = 0.002). 3.7. Factors associated with cognitive dysfunction We used various logistic regression models to estimate odds ratio (OR) for cognitive dysfunction (Table 3). Unadjusted odds of cognitive dysfunction were statistically significantly higher for older baseline age at stroke occurrence, less than 6 years of educational attainment, female gender, alcohol use, pre-stroke cognitive decline and medial temporal lobe atrophy but were lower for daily pre-stroke fish intake, moderate to heavy pre-stroke physical activity and total brain volume. Multivariate logistic regression models were fitted for the full cohort (n = 143) and the MRI sub-cohort (n = 58) and comparative odd ratios (ORs) generated are shown in Table 3. For the full cohort, ORs were attenuated and no longer significant for female gender, alcohol use, and moderate to heavy pre-stroke physical activity and total brain volume. However, the ORs remained significant and similar for baseline age, and stronger for less than 6 years of educational attainment, daily

246

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249

Table 2 Demographic and cognitive performance profiles of stroke survivors and control subjects. Variable

Controls (N = 74)

No vCIND (N = 71)

vCIND (N = 57)

PSD (N = 12)

ANOVA

A

B

C

D

Age (years) Education (years) CSID total score MMSE total score AF score AF (0–15 s) FAS total score VR test score WLL total score WL_recall score BNT score SD total score SD_recall total score MTT total score SRT (ms) CRT (ms) VIGRT (ms) SPMRT (ms) VIGACC (%) SPMSI

58.8 (8.6) 11.0 (5.7) 27.8 (2.8) 27.4 (3.0) 13.4 (4.0) 5.4 (1.7) 25.0 (15.1) 3.3 (1.6) 17.2 (3.9) 5.7 (1.9) 12.1 (2.7) 11.5 (1.6) 6.1 (3.0) 20.7 (3.0) 682.6 (617.6) 731.4 (204.5) 470.9 (65.9) 2201.6 (1015.2) 87.2 (14.3) 0.6 (0.4)

58.2 (9.3) 11.6 (4.4) 27.3 (1.8) 27.8 (2.7) 11.3 (3.2) 4.5 (2.9) 20.1 (9.7) 2.6 (1.5) 17.2 (3.1) 5.2 (1.8) 12.0 (1.9) 11.8 (0.6) 5.9 (2.1) 20.5 (2.1) 647.4 (403.3) 826.3 (286.7) 518.6 (83.6) 2412.5 (982.3) 81.5 (19.5) 0.6 (0.4)

62.3 (8.9) 7.1 (5.8) 25.0 (2.7) 23.0 (4.1) 7.7 (3.8) 3.6 (1.6) 6.6 (6.3) 1.8 (1.2) 12.7 (3.9) 2.4 (1.9) 8.2 (3.3) 8.0 (3.8) 3.1 (2.6) 14.8 (4.9) 1268.1 (1017.3) 1360.1 (766.0) 570.7 (76.5) 3522.6 (2090.6) 54.4 (24.2) 0.4 (0.3)

65.4 (10.2) 5.9 (5.3) 15.9 (5.0) 13.1 (4.2) 3.8 (3.7) 2.1 (1.7) 0.9 (2.0) 1.2 (0.7) 7.6 (4.5) 1.2 (1.5) 6.7 (6.6) 6.5 (4.1) 1.9 (2.2) 9.8 (3.8) 2128.6 (1421.8) 2026.9 (945.4) 605.9 (71.4) 3802.4 (2546.0) 40.6 (26.1) 0.1 (0.3)

b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001

0.563 0.458 0.249 0.175 0.004 0.002 0.056 0.982 0.762 0.151 0.836 0.301 0.676 0.556 0.702 0.034 b0.001 0.235 0.065 0.586

0.007 b0.001 b0.001 b0.001 b0.001 0.006 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 0.001 b0.001 b0.001 b0.001

0.309 0.453 b0.001 b0.001 0.003 0.005 0.003 0.001 b0.001 0.038 0.207 0.201 0.168 0.001 0.023 0.016 0.169 0.704 0.097 0.004

0.004 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001

ANOVA = analysis of variance. Between group comparisons were performed by Student's t-test (i.e. A, B, C and D). Due to Bonferroni corrections for multiple pairwise comparisons, adjusted level of significance, α (adjusted) = 0.0083. Between group comparisons — A (control vs no vCIND); B (no vCIND vs vCIND); C (vCIND vs PSD); D (no vCIND vs vCIND + PSD). Abbreviations: vCIND = vascular cognitive impairment no dementia; PSD = post-stroke dementia; CSID = Community Screening Instrument for Dementia; MMSE = Mini-mental State Examination; AF = animal fluency; FAS = letter fluency; VR = Visual Reasoning Test; WLL = World List Learning; BNT = Boston Naming Test; SD = Stick Design Test; MTT = Modified Token Test; SRT = Simple Reaction Time; CRT = Choice Reaction Time; DVRT = Digit Vigilance Reaction Time; SPMRT = Spatial Working Memory Reaction Time; VIGACC = digit vigilance accuracy; SPMOACC = spatial working memory original stimuli accuracy; SPMSI = spatial working memory sensitivity index.

pre-stroke fish intake, pre-stroke cognitive decline and medial temporal lobe atrophy. In the MRI sub-cohort, baseline age, pre-stroke fish intake and pre-stroke cognitive decline were no longer significant whereas educational attainment and medial temporal lobe atrophy rating sustained their statistical significance for independent association with cognitive dysfunction three months after stroke. 4. Discussion Our principal findings show that at three months post-stroke, 39.9% of a cohort of Nigerian African stroke survivors had VCI but no dementia while 8.4% had PSD according to the AHA/ASA criteria [2] and our cognitive battery modeled after the NINDS–CSN Harmonization Standards Neuropsychological Battery [8]. Furthermore, attention, mental speed and executive function deficits were found in stroke patients who were without clinically apparent cognitive impairment compared to stroke-free controls. In patients who met criteria for vCIND, there was further worsening of executive functioning with additional significant impairment in the domains of language, memory and visuospatial/visuoconstructive function. Survivors who were demented at three months had greater worsening of executive function, memory and visuospatial/visuoconstructive function. Significant factors associated with cognitive dysfunction at three months post-stroke included older baseline age, female gender, lower educational attainment, history of alcohol use, pre-stroke cognitive decline and medial temporal lobe atrophy while pre-stroke daily intake of fish and moderate-to-heavy physical activity were protective. The average age of stroke onset of about

60 years in this study population agrees with previous findings from Nigeria [40] and other parts of sub-Saharan Africa [41]. This may be due to the current demographic structure of Nigeria characterized by a very high proportion of young people (only 4% are N 65 years and older), the high prevalence of undiagnosed and uncontrolled hypertension and possible genetic factors. The frequency of vCIND found in our study is comparable to findings from several centers worldwide: Sydney, Australia, 39.4% [42], Chongqing, China, 37.1% [43], and Santiago, Chile, 39.0% [44]; but it is lower than 49.9% from Korea [45], 55.0% from Lisbon, Portugal [46] and 54.8% from Singapore [38]. The rates were, however, lower in Newcastle ranging from 24% [34] to 32% [30], as well as in a previous study from Hong Kong, China (21.8%) [47] while data from the South London Stroke Register (SLSR) cohort using the MMSE revealed a rate of 22% [48]. However, our PSD rate of 8.4% is similar to the findings of 8.6% by Stephens et al. [34] and lies within the range of pooled post-stroke prevalence of 7.4%–41.3% within the first year reported in a recent systematic review [9]. Although the mean age of our cohort is about 60 years similar to that of Dong et al. [38] but younger than in other studies [7] the reported relatively high rates of vCIND and dementia suggest the possibility that negroids may also have an increased susceptibility to the cognitive sequelae of stroke just as they have demonstrated worse cardiovascular outcomes in studies involving multiracial populations [9]. Apart from this, a stroke injury may trigger/accelerate cognitive decline [49], the age of occurrence notwithstanding [40]. This outcome may also be a reflection of the sub-optimal status of acute stroke care, rehabilitation and secondary prevention as seen in many developing countries [50].

Fig. 3. Magnetic resonance imaging (MRI) T1-weighted coronal images showing different degrees of medial temporal lobe atrophy (MTLA) in Nigerian stroke survivors: [A] Grade 4 MTLA in a 70 year old male; [B] Grade 3 MTLA in an 81 year female; [C] Grade 2 MTLA in a 52 year male; [D] Grade 1 MTLA in a 48 year male; and [E] Grade 0 MTLA in a 49 year female.

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249

Fig. 4. The pattern of performance on Choice Reaction Time (CRT) in controls and impaired subjects.

The heterogeneity of the prevalence of post-stroke MCI and dementia from different studies may be due to differences in patient characteristics (sample size, cut-off age, gender, population structure, ethnicity), study setting (hospital vs community-based), study design (first ever and/or recurrent strokes, timing of assessment), differences in study instruments and in the diagnostic criteria for post-stroke cognitive impairment and dementia [8,9,11]. It is important to point out that with the growing understanding of the early and prominent occurrence of executive dysfunction in subjects with VCI, recent studies that have used cognitive tools with robust items sensitive to executive dysfunction have reported more cases of MCI, which hitherto might have been wrongly passed as normal. This may explain the relatively high frequency of vCIND recorded in this and similar recent studies [38,51]. Our previous observations from Newcastle [7,34] showed varying degrees of early cognitive impairment in stroke survivors. Compared to controls, stroke patients without significant cognitive impairment at three months after stroke still showed deficits in attention, mental speed and executive function confirming the early occurrence of executive dysfunction in VCI. Subjects with vCIND and PSD had persistence and progression of executive dysfunction in addition to multi-domain impairments of language, memory and visuospatial/visuoconstructive functioning in tandem with previous findings in literature [7,34,51]. The prominent worsening of memory and visuospatial functioning in patients with PSD may suggest the presence of underlying neurodegenerative pathologies producing an additive/synergistic effect with vascular pathologies [49,52]. Like previous findings [51], we also found non-amnestic (single and multi-domain) impairment more common than amnestic (single and multi-domain) impairment in the vCIND subgroup, a reflection of the sensitivity of the NINDS/CSN Neuropsychological Battery for detecting non-memory domain impairments in subjects with vascular MCI [8]. It is important to note that patients with vCIND had a MMSE score of 23 (compared to MMSE of 27 in patients without vCIND and controls). However, interpretation of MMSE values depends on the specific population. The MMSE dementia cut off score of 24 used in Western populations has been found to be inappropriate in some populations with limited formal education. In fact, previous validation of MMSE among Yoruba-speaking Nigerians suggested an MMSE cut off score of 13 for dementia in persons with no formal education and 16 in those with one or more years of formal education [25]. Our finding of older age, female gender, low educational attainment, pre-stroke cognitive decline and medial temporal lobe atrophy as

247

factors significantly associated with post-stroke cognitive dysfunction are consistent with findings from previous studies [6,9,11,49,53,54] and suggest a role for age-related neurodegeneration synergizing with stroke to cause cognitive impairment and dementia [49]. As populations age and life expectancy grows in the developing regions of the world, the burden of aging-associated disorders including stroke and dementia have been projected to grow [55–57]. In this study, educational attainment significantly stratified stroke survivors into cognitive categories three months after stroke suggesting that education and related cognitive reserve might have provided some compensation against vascular brain injury [58]. Educational attainment has been found to protect against cognitive impairment in numerous disorders including MCI, Alzheimer's disease and vascular dementia [3,59]. Educational attainment together with occupational complexity and social engagement constitute the ‘Cognitive Lifestyle’ paradigm [60–62] which has been associated with a reduced long-term risk of dementia and found to be associated with neurotrophic changes in the prefrontal lobe consistent with a compensatory process. Physical activity promotes brain health by enhancing cerebral blood flow and the production of growth factors [63]. The protective effect of fish intake against cerebral vascular disease risk (including stroke and VCI) is established [64,65]. Although, omega-3 fatty acids (docosahexaenoic acid and eicosapentaenoic acid) in fish have been largely implicated in the protection, other nutritional constituents of fish including vitamins D and B complex, essential amino acids and trace elements in fish (for example, taurine, arginine, selenium, calcium, magnesium, potassium, and iodine) may have potentially favorable

Table 3 Univariate and multivariate predictors of cognitive dysfunction 3 months after stroke. Subject characteristic

Univariate analysis Multivariate analysis OR (95% CI) OR (95% CI) Full cohort MRI sub-cohort (n = 143) (n = 58)

Baseline age (years) Female gender b6 years of education Hypertension Diabetes mellitus Previous stroke Smoking Alcohol use Pre-stroke daily fish intake Pre-stroke moderate to heavy physical activity No of vascular risk factors Pre-stroke cog. decline Stroke type Modified Rankin score Barthel Index score CESD score

1.06 (1.02–1.10)

1.05 (1.00–1.09)

2.27 (1.15–4.45) 4.84 (2.36–9.92)

1.87 (0.80–4.40) 0.80 (0.14–4.71) 5.09 (2.17–11.95) 8.52 (2.58–28.12)

1.18 (0.30–4.58) 1.29 (0.59–2.79) 1.38 (0.51–3.10) 1.25 (0.51–3.10) 2.01 (1.01–4.00) 0.42 (0.20–0.88)

1.19 (0.47–3.00) 0.37 (0.15–0.89)

0.17 (0.04–0.84)

1.00 (0.99–1.02)

0.86 (0.15–5.02)

0.57 (0.05–6.40) 0.93 (0.17–5.07) 5.2 (0.125–220.05)

1.08 (0.68–1.72) 3.61 (1.09–11.9)

4.51 (1.20–16.88) 1.49 (0.11–19.73)

0.72 (0.34–1.54) 1.03 (0.53–1.98) 0.98 (0.90–1.06) 1.04 (0.96–1.12)

MRI sub-cohort (n = 58) Log_ICV 0.01 (0.001–1141.1) Log_TBV 0.04 (0.01–0.20) Log_Ven Vol 18.6 (0.81–429.36) MTLA rating 2.05 (1.28–3.27) Periventricular 1.22 (0.96–1.56) WM score Deep WM score 1.07 (0.99–1.15) Total WM score 1.06 (1.00–1.13)

1.00 (1.00–1.00) 2.18 (1.24–3.83)

Significant results are shown in bold (p b 0.05). Abbreviations: CESD = Centre for Epidemiologic Studies Depression Scale; ICV, intracerebral volume; MTLA, medial temporal lobe atrophy; TBV, total brain volume; Ven Vol, ventricular volume; WM, white matter.

248

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249

effects on inflammation, endothelial function, vascular tone, neuronal functioning and cell death [65,66]. Recently, lower red blood cell omega-3 fatty acid levels were associated with smaller brain volumes, increased white matter hyperintensity and vascular pattern of cognitive impairment on tests of visual memory and executive function [67]. However, this finding should be interpreted with caution, as we did not collect specific details on the different species of edible fish or the methods of preparation, whether fried or boiled. Pre-stroke cognitive decline has been associated with neurodegeneration [68]. The predictive role of MTLA in stratifying cognitive categories of stroke survivors and predicting cognitive decline, dementia and death had been previously shown [13,69,70] while significant correlation between MTLA and WMH strengthens the case for a vascular basis of neurodegeneration, WMH being a surrogate of small vessel disease [52, 71]. Unexpectedly, cardiovascular risk factors did not significantly predict early cognitive dysfunction in this study. Pendlebury and Rothwell [9] had also reported in their meta-analysis the lack of association of hypertension, and ischemic heart disease with PSD. This paradox is explained by Allan et al. [10] wherein cardiovascular risk factor load predicted long term cognitive outcome and mortality in elderly stroke survivors. Our study is not without its limitations. The sample size was modest and adequate characterization of all our stroke cases was limited by the availability and cost of neuroimaging. The limited size of the neuroimaging sub-cohort further affected the strength of statistical associations of cognitive outcome in the cohort. Nonetheless, cases with and without neuroimaging were quite similar statistically, and our cohort shares similar demographic and phenomic characteristics with stroke cases described earlier in Nigeria [72] and sub-Saharan Africa [23]. Follow up of our cohort is currently in progress to explore the early evolution of cognitive dysfunction and the determining factors. Such longitudinal studies engaging population-based samples are needed to tease out the influence of cardiovascular risk factors, genetic architecture, epigenetic factors, social engagement and other cultural influences on the cognitive trajectory after stroke in populations of African ancestry [48]. Contributors The co-authors have no disclosures with regard to this report. The study was not industry-sponsored. The individual contributor statements are as follows: Rufus O. Akinyemi designed the study, was responsible for the recruitment of subjects, performed clinical assessments and data analysis and wrote the first draft of the manuscript. Louise Allan advised on the study design, neuropsychometric assessments and analysis of the results and reviewed the manuscript for intellectual content. Mayowa O. Owolabi assisted in the recruitment, advised on the study design and clinical assessments and reviewed the manuscript for intellectual content. Joshua O. Akinyemi advised on and interpreted the statistical analysis and reviewed the manuscript for intellectual content. Godwin Ogbole (GO) analyzed the images and reviewed the manuscript for intellectual content. Akinlolu Ajani assisted on the clinical and neuropsychometric assessments and collated the data. Michael Firbank (MF) analyzed the images and reviewed the manuscript for intellectual content. Adesola Ogunniyi was the key facilitator of the recruitment, advised on the study design and clinical assessments and reviewed the manuscript for intellectual content. Raj N. Kalaria conceived the study, advised on the assessments, reviewed the manuscript for intellectual content and obtained the funding.

Conflict of interest None declared. Funding Our work is supported by grants from the Newcastle Centre for Brain Ageing and Vitality (BBSRC, EPSRC, ESRC and MRC, LLHW), the UK Medical Research Council (MRC, G0500247), and the Alzheimer's Research UK (UK). ROA is supported by a fellowship from the International Brain Research Organization (IBRO) Paris, France, and by an ORS Award from the Newcastle University, UK. Acknowledgments We are very grateful to the patients, families and clinical staff for their co-operation and help with the execution of this study. References [1] Seshadri S, Wolf PA. Lifetime risk of stroke and dementia: current concepts, and estimates from the Framingham Study. Lancet Neurol 2007;6:1106–14. [2] Gorelick PB, Scuteri A, Black SE, Decarli C, Greenberg SM, Iadecola C, et al. Vascular contributions to cognitive impairment and dementia: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2011;42:2672–713. [3] Kalaria RN, Maestre GE, Arizaga R, Friedland RP, Galasko D, Hall K, et al. Alzheimer's disease and vascular dementia in developing countries: prevalence, management, and risk factors. Lancet Neurol 2008;7:812–26. [4] Akinyemi RO, Ogah OS, Ogundipe RF, Oyesola OA, Oyadoke AA, Ogunlana MO, et al. Knowledge and perception of stroke amongst hospital workers in an African community. Eur J Neurol 2009;16:998–1003. [5] Pendlebury ST. Stroke-related dementia: rates, risk factors and implications for future research. Maturitas 2009;64:165–71. [6] Henon H, Pasquier F, Leys D. Poststroke dementia. Cerebrovasc Dis 2006;22:61–70. [7] Ballard C, Rowan E, Stephens S, Kalaria R, Kenny RA. Prospective follow-up study between 3 and 15 months after stroke: improvements and decline in cognitive function among dementia-free stroke survivors N75 years of age. Stroke 2003;34: 2440–4. [8] Hachinski V, Iadecola C, Petersen RC, Breteler MM, Nyenhuis DL, Black SE, et al. National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards. Stroke 2006;37:2220–41. [9] Pendlebury ST, Rothwell PM. Prevalence, incidence, and factors associated with prestroke and post-stroke dementia: a systematic review and meta-analysis. Lancet Neurol 2009;8:1006–18. [10] Allan LM, Rowan EN, Firbank MJ, Thomas AJ, Parry SW, Polvikoski TM, et al. Long term incidence of dementia, predictors of mortality and pathological diagnosis in older stroke survivors. Brain 2011;134:3716–27. [11] Gottesman R, Hillis AE. Predictors and assessment of cognitive dysfunction resulting from ischaemic stroke. Lancet Neurol 2010;9:895–905. [12] Burton EJ, Kenny RA, O'Brien J, Stephens S, Bradbury M, Rowan E, et al. White matter hyperintensities are associated with impairment of memory, attention, and global cognitive performance in older stroke patients. Stroke 2004;35:1270–5. [13] Firbank MJ, Burton EJ, Barber R, Stephens S, Kenny RA, Ballard C, et al. Medial temporal atrophy rather than white matter hyperintensities predict cognitive decline in stroke survivors. Neurobiol Aging 2007;28:1664–9. [14] Akinyemi RO, Mukaetova-Ladinska EB, Attems J, Ihara M, Kalaria RN. Vascular risk factors and neurodegeneration in ageing related dementias: Alzheimer's disease and vascular dementia. Curr Alzheimer Res 2013;10:642–53. [15] Bamford JSP, Dennis M, Burn J, Warlow C. Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet 1991;337:1521–6. [16] Ogun SA, Oluwole O, Fatade B, Ogunseyinde AO, Ojini FI, Odusote KA. Comparison of Siriraj Stroke Score and the WHO criteria in the clinical classification of stroke subtypes. Afr J Med Med Sci 2002;31:13–6. [17] Adams HPJ, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. Trial of Org 10172 in Acute Stroke Treatment (TOAST). Stroke 1993;24:35–41. [18] Desmond DWMJ, Sano M, Stern Y. Recovery of cognitive function after stroke. Stroke 1996;27:1798–803. [19] Wilkinson PR, Wolfe CD, Warburton FG, Rudd AG, Howard RS, Ross-Russell RW, et al. Longer term quality of life and outcome in stroke patients: is the Barthel Index alone an adequate measure of outcome? Qual Health Care 1997;6:125–30. [20] Owolabi MO, Ogunniyi A. Profile of health-related quality of life in Nigerian stroke survivors. Eur J Neurol 2009;16:54–62. [21] Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med 1994;10:77–84. [22] Almeida OP, Almeida SA. Short versions of the geriatric depression scale: a study of their validity for the diagnosis of a major depressive episode according to ICD-10 and DSM-IV. Int J Geriat Psychiatry 1999;14:858–65.

R.O. Akinyemi et al. / Journal of the Neurological Sciences 346 (2014) 241–249 [23] O'Donnell M, Xavier D, Diener C, Sacco R, Lisheng L, Zhang H, et al. Rationale and design of INTERSTROKE: a global case–control study of risk factors for stroke. Neuroepidemiology 2010;35:36–44. [24] Hall KS, Gao S, Emsley CL, Ogunniyi AO, Morgan O, Hendrie HC. Community screening interview for dementia (CSI ‘D’); performance in five disparate study sites. Int J Geriatr Psychiatry 2000;15:521–31. [25] Guruje O, Unverzargt FW, Osuntokun BO, Hendrie HC, Baiyewu O, Ogunniyi A, et al. The CERAD Neuropsychological Test Battery: norms from a Yoruba-speaking Nigerian sample. West Afr J Med 1995;14:29–33. [26] Shen J, Gao S, Unverzagt FW, Ogunniyi A, Baiyewu O, Gureje O, et al. Validation analysis of informant's ratings of cognitive function in African Americans and Nigerians. Int J Geriatr Psychiatry 2006;21:618–25. [27] Ogunniyi A, Baiyewu O, Gureje O, Hall KS, Unverzagt F, Siu SH, et al. Epidemiology of dementia in Nigeria: results from the Indianapolis–Ibadan study. Eur J Neurol 2000; 7:485–90. [28] Guerchet M, Houinato D, Paraiso MN, von Ahsen N, Nubukpo P, Otto M, et al. Cognitive impairment and dementia in elderly people living in rural Benin, West Africa. Dement Geriatr Cogn Disord 2009;27:34–41. [29] Roth M, Tym E, Mountjoy CQ, Huppert FA, Hendrie H, Verma S, et al. CAMDEX. A standardised instrument for the diagnosis of mental disorder in the elderly with special reference to the early detection of dementia. Br J Psychiatry 1986;149:698–709. [30] Ballard C, Stephens S, McLaren A, Wesnes K, Kenny RA, Burton E, et al. Neuropsychological deficits in older stroke patients. Ann N Y Acad Sci 2002; 977:179–82. [31] Baiyewu O, Unverzagt FW, Lane KA, Gureje O, Ogunniyi A, Musick B, et al. The Stick Design Test: a new measure of visuoconstructional ability. J Int Neuropsychol Soc 2005;11:598–605. [32] Unverzagt FW, Morgan OS, Thesiger CH, Eldemire DA, Luseko J, Pokuri S, et al. Clinical utility of CERAD neuropsychological battery in elderly Jamaicans. J Int Neuropsychol Soc 1999;5:255–9. [33] Akinyemi RO, Okubadejo NN, Akinyemi JO, Owolabi MO, Owolabi LF, Ogunniyi A. Cognitive dysfunction in Nigerians with Parkinson's disease. Mov Disord 2008;23: 1378–83. [34] Stephens S, Kenny RA, Rowan E, Allan L, Kalaria RN, Bradbury M, et al. Neuropsychological characteristics of mild vascular cognitive impairment and dementia after stroke. Int J Geriatr Psychiatry 2004;19:1053–7. [35] Scheltens P, Leys D, Barkhof F, Huglo D, Weinstein HC, Vermersch P, et al. Atrophy of medial temporal lobes on MRI in “probable” Alzheimer's disease and normal ageing: diagnostic value and neuropsychological correlates. J Neurol Neurosurg Psychiatry 1992;55:967–72. [36] Scheltens P, Barkhof F, Leys D, Pruvo JP, Nauta JJ, Vermersch P, et al. A semiquantative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging. J Neurol Sci 1993;114:7–12. [37] Uyttenboogaart M, Stewart RE, Vroomen PC, De Keyser J, Luijckx GJ. Optimizing cutoff scores for the Barthel Index and the modified Rankin Scale for defining outcome in acute stroke trials. Stroke 2005;36:1984–7. [38] Dong Y, Venketasubramanian N, Chan BP, Sharma VK, Slavin MJ, Collinson SL, et al. Brief screening tests during acute admission in patients with mild stroke are predictive of vascular cognitive impairment 3–6 months after stroke. J Neurol Neurosurg Psychiatry 2012;83:580–5. [39] Bossuyt PM, Reitsma JB. The STARD initiative. Lancet 2003;361:71. [40] Owolabi MO. Taming the burgeoning stroke epidemic in Africa: stroke quadrangle to the rescue. West Indian Med J 2011;60:412–21. [41] O'Donnell MJ, Xavier D, Liu L, Zhang H, Chin SL, Rao-Melacini P, et al. Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22 countries (the INTERSTROKE study): a case–control study. Lancet 2010;376:112–23. [42] Sachdev PS, Brodaty H, Valenzuela MJ, Lorentz L, Looi JCL, Wen W, et al. The neuropsychological profile of vascular cognitive impairment in stroke and TIA patients. Neurology 2004;62:912–9. [43] Zhou DH, Wang JY, Li J, Deng J, Gao C, Chen M. Study on frequency and predictors of dementia after ischemic stroke: the Chongqing stroke study. J Neurol 2004;251:421–7. [44] Delgado C, Donoso A, Orellana P, Vasquez C, Diaz V, Behrens MI. Frequency and determinants of poststroke cognitive impairment at three and twelve months in Chile. Dement Geriatr Cogn Disord 2010;29:397–405. [45] Yu KH, Cho SJ, Oh MS, Jung S, Lee JH, Shin JH, et al. Cognitive impairment evaluated with Vascular Cognitive Impairment Harmonization Standards in a multicenter prospective stroke cohort in Korea. Stroke 2013;44:786–8.

249

[46] Madureira S, Guerreiro M, Ferro JM. Dementia and cognitive impairment three months after stroke. Eur J Neurol 2001;8:621–7. [47] Tang WK, Chan SS, Chiu HF, Ungvari GS, Wong KS, Kwok TC, et al. Frequency and determinants of poststroke dementia in Chinese. Stroke 2004;35:930–5. [48] Douiri A, Rudd AG, Wolfe CD. Prevalence of poststroke cognitive impairment: South London Stroke Register 1995–2010. Stroke 2013;44:138–45. [49] Kalaria RN. Risk factors and neurodegenerative mechanisms in stroke related dementia. Panminerva Med 2012;54:139–48. [50] Norrving B, Kissela B. The global burden of stroke and need for a continuum of care. Neurology 2013;80:S5–S12. [51] Pendlebury ST, Mariz J, Bull L, Mehta Z, Rothwell PM. MoCA, ACE-R, and MMSE versus the National Institute of Neurological Disorders and Stroke-Canadian Stroke Network Vascular Cognitive Impairment Harmonization Standards Neuropsychological Battery after TIA and stroke. Stroke 2012;43:464–9. [52] Kalaria RN. Cerebrovascular disease and mechanisms of cognitive impairment: evidence from clinicopathological studies in humans. Stroke 2012;43:2526–34. [53] Patel M, Coshall C, Rudd AG, Wolfe CD. Natural history of cognitive impairment after stroke and factors associated with its recovery. Clin Rehabil 2003;17:158–66. [54] Das S, Paul N, Hazra A, Ghosal M, Ray BK, Banerjee TK, et al. Cognitive dysfunction in stroke survivors: a community-based prospective study from Kolkata, India. J Stroke Cerebrovasc Dis 2013;22:1233–42. [55] Prince M, Acosta D, Ferri CP, Guerra M, Huang Y, Llibre Rodriguez JJ, et al. Dementia incidence and mortality in middle-income countries, and associations with indicators of cognitive reserve: a 10/66 Dementia Research Group population-based cohort study. Lancet 2012;380:50–8. [56] Prince M, Bryce R, Albanese E, Wimo A, Ribeiro W, Ferri CP. The global prevalence of dementia: a systematic review and metaanalysis. Alzheimers Dement 2013;9:63-75.e2. [57] Feigin VL, Forouzanfar MH, Krishnamurthi R, Mensah GA, Connor M, Bennett DA, et al. Global and regional burden of stroke during 1990–2010: findings from the Global Burden of Disease Study 2010. Lancet 2014;383:245–54. [58] Stern Y. Cognitive reserve. Neuropsychologia 2009;47:2015–28. [59] Stern Y. Cognitive reserve in ageing and Alzheimer's disease. Lancet Neurol 2012;11: 1006–12. [60] Valenzuela M, Brayne C, Sachdev P, Wilcock G, Matthews F. Cognitive lifestyle and long-term risk of dementia and survival after diagnosis in a multicenter population-based cohort. Am J Epidemiol 2011;173:1004–12. [61] Valenzuela MJ, Matthews FE, Brayne C, Ince P, Halliday G, Kril JJ, et al. Multiple biological pathways link cognitive lifestyle to protection from dementia. Biol Psychiatry 2012;71:783–91. [62] Valenzuela MJ, Sachdev P. Brain reserve and cognitive decline: a non-parametric systematic review. Psychol Med 2006;36:1065–73. [63] Cotman CW, Berchtold NC. Physical activity and the maintenance of cognition: learning from animal models. Alzheimers Dement 2007;3:S30–7. [64] Perez L, Heim L, Sherzai A, Jaceldo-Siegl K, Sherzai A. Nutrition and vascular dementia. J Nutr Health Aging 2012;16:319–24. [65] Chowdhury R, Stevens S, Gorman D, Pan A, Warnakula S, Chowdhury S, et al. Association between fish consumption, long chain omega 3 fatty acids, and risk of cerebrovascular disease: systematic review and meta-analysis. BMJ 2012;345: e6698. [66] Robinson JG, Ijioma N, Harris W. Omega-3 fatty acids and cognitive function in women. Womens Health (Lond Engl) 2010;6:119–34. [67] Tan ZS, Harris WS, Beiser AS, Au R, Himali JJ, Debette S, et al. Red blood cell omega-3 fatty acid levels and markers of accelerated brain aging. Neurology 2012;78:658–64. [68] Henon H, Pasquier F, Durieu I, Godefroy O, Lucas C, Lebert F, et al. Preexisting dementia in stroke patients. Baseline frequency, associated factors, and outcome. Stroke 1997;28:2429–36. [69] Henon H, Pasquier F, Durieu I, Pruvo JP, Leys D. Medial temporal lobe atrophy in stroke patients: relation to pre-existing dementia. J Neurol Neurosurg Psychiatry 1998;65:641–7. [70] Firbank MJ, Allan LM, Burton EJ, Barber R, O'Brien JT, Kalaria RN. Neuroimaging predictors of death and dementia in a cohort of older stroke survivors. J Neurol Neurosurg Psychiatry 2012;83:263–7. [71] Oosterman JM, Van Harten B, Weinstein HC, Scheltens P, Sergeant JA, Scherder EJ. White matter hyperintensities and working memory: an explorative study. Neuropsychology 2008;15:384–99. [72] Owolabi MO, Agunloye AM. Risk factors for stroke among patients with hypertension: a case–control study. J Neurol Sci 2013;325:51–6.