Abstract

The association between speed of information processing and cognition has been extensively validated in normal aging and other neurocognitive disorders. Our aim was to determine whether such a relationship exists in stroke. Thirty patients and 30 age- and education-matched healthy individuals were administered a comprehensive battery of neuropsychological tests divided into the following six cognitive domains: processing speed (PS), verbal memory, visual memory, visuoperceptual function, language, and cognitive flexibility. The results demonstrate that stroke patients were characterized by cognitive deficits in almost all of these domains, but have the most pronounced deficits in PS. After adjusting for symbol digit modalities test score, all significant group differences in cognitive functioning disappeared. However, group differences remained significant after controlling for the influence of other cognitive factors. These findings suggest that decreased PS appears to underlie post-stroke cognitive dysfunction and may serve as a potential target for intervention.

Introduction

Cognitive impairment is a common sequel of stroke, with estimated rates of 35%–70% in the post-acute and chronic phases after onset (Leśniak, Bak, Czepiel, Seniów, & Członkowska, 2008; Tatemichi et al., 1994). Several factors have been identified which are associated with a greater risk of the development of post-stroke cognitive decline, including older age (Nys et al., 2007), female gender (Nys et al., 2007), low educational level (Madureira, Guerreiro, & Ferro, 2001), depression (Kauhanen et al., 1999), previous history of stroke (Mok et al., 2004), causes of ischemic stroke such as large artery atherosclerosis, small vessel occlusion, or lacunar infarction (Douiri, Rudd, & Wolfe, 2013), prestroke vascular risk factors such as hypertension, diabetes mellitus, atrial fibrillation, and hypercholesterolemia (Gorelick et al., 2011; Kilander, Nyman, Boberg, Hansson, & Lithell, 1998), multiple infarctions (Moroney et al., 1997), middle cerebral artery territory strokes (Jaillard, Grand, Le Bas, & Hommel, 2010; Nys et al., 2007; Tatemichi et al., 1994), comorbid neurodegenerative disease (most commonly Alzheimer's disease) (Allan et al., 2011), and structural abnormalities (presence of silent infarcts, white matter (WM) lesions, and cerebral atrophy) (Tatemichi et al., 1993). The resultant cognitive profiles can vary, depending on the side of lesion (right or left hemisphere) (van der Ham, van Zandvoort, Frijns, Kappelle, & Postma, 2011), vascular territories (anterior or posterior circulation) (Barker-Collo et al., 2012; Hannay, Howieson, Loring, Fischer, & Lezak, 2004), subtypes of hemorrhagic stroke (intracerebral or subarachnoid hemorrhage, SAH) (Barker-Collo, Feigin, & Krishmnamurthi, 2013), and anatomic location (cortical or subcortical) (Jokinen et al., 2006). A stroke occurring in a localized area may lead to selective impairments such as aphasia and hemispatial neglect, while diffuse neuronal damage resulting from underlying subclinical stroke, such as WM disease or silent infarcts (Cumming, Marshall, & Lazar, 2013; Gottesman & Hillis, 2010), produces a more general pattern of cognitive problems, affecting multiple domains, with some cognitive functions appearing more susceptible than others to the effects of stroke.

Considerable evidence points to processing speed (PS) as being the most severe cognitive deficit subsequent to stroke (Ballard et al., 2003; Barker-Collo, Feigin, Parag, Lawes, & Senior, 2010; Brand et al., 2014; Hochstenbach, Mulder, van Limbeek, Donders, & Schoonderwaldt, 1998; Hurford, Charidimou, Fox, Cipolotti, & Werring, 2013; Rasquin, Lodder, & Verhey, 2005). Cumming and colleagues (2013) assert that deficits in PS should be included in the “typical” post-stroke cognitive profile along with deficits in attention and executive function. PS is clinically relevant, as it ranks high on the list of self-reported cognitive complaints (Winkens, Van Heugten, Fasotti, Duits, & Wade, 2006), and has been shown to interfere with long-term functional recovery and quality of life in stroke survivors (Barker-Collo et al., 2010; Narasimhalu et al., 2011). Such findings suggest that PS is a feasible target for early intervention after stroke.

PS has long been conceptualized as an essential element of intellectual abilities (Kail & Salthouse, 1994; Salthouse, 1996) and together with working memory, presumably imposes a constraint on the amount of information that can be activated at any one time (Vernon & Kantor, 1986). Fast and efficient information processing is intellectually advantageous because it makes available limited processing resources that can be used to engage in more higher-order cognitive tasks (Kail & Salthouse, 1994).

Unfortunately, PS is vulnerable to the aging process and brain insults. Measures of PS have consistently been found to be most sensitive and able to differentiate between younger and older groups, and between clinical groups and healthy controls. There is now a large body of evidence suggesting PS underlies the age- and disease-related changes in cognitive abilities. In short, PS is considered a core component that variably mediates the relationship between aging and decrements in other cognitive domains, such as fluid intelligence (Bugg, Zook, DeLosh, Davalos, & Davis, 2006), verbal fluency (Elgamal, Roy, & Sharratt, 2011), and episodic memory (Lee et al., 2012). In a study by Rodríguez-Sánchez, Crespo-Facorro, González-Blanch, Perez-Iglesias, and Vázquez-Barquero (2007), differences between first-episode schizophrenia patients and controls in tests related to verbal memory, attention, working memory, and executive function became non-significant when differences in PS was controlled by using covariate analyses. In the case of late life depression, alterations in PS could explain most of the variance in tests of different cognitive domains (such as visuospatial function, executive function, memory, and language) (Butters et al., 2004). Reduced PS was also implicated in attentional difficulties after traumatic brain injury (Willmott, Ponsford, Hocking, & Schönberger, 2009). In patients with multiple sclerosis, slower processing was correlated with lower verbal fluency and poorer recall of verbal and visual–spatial information (Diamond, Johnson, Kaufman, & Graves, 2008). Along a similar vein, McKinlay, Dalrymple-Alford, Grace, and Roger (2009) reported that pragmatic language deficits in patients with Parkinson's disease might be attributed to their deficits in PS. Finally, PS was the only cognitive parameter with significant relationships to dimensional measures (i.e., reading and inattention) of both reading disability and attention-deficit hyperactivity disorder, even after adjusting for the effects of all other cognitive factors (McGrath et al., 2011).

It seems plausible to speculate that the same results would hold true for patients with stroke in light of high prevalence of subjective (Winkens et al., 2006) and objective (Barker-Collo et al., 2010; Hochstenbach et al., 1998; Hurford et al., 2013) PS deficits in these patients. However, so far there appears to be only limited evidence to support this. In a small-scale qualitative study, the majority of stroke patients reported that their mental slowness causes everyday problems pertaining to attention, memory, and executive function (Winkens et al., 2006). Given the prognostic implications of cognitive impairment after stroke (Barker-Collo et al., 2010; Narasimhalu et al., 2011), determining in a systematic and objective fashion whether stroke-related cognitive decline may be at least in part a consequence of PS deterioration seems warranted.

Hence, the primary objective of this investigation was to evaluate the role of PS in subserving different aspects of cognition among patients with ischemic stroke and intracerebral hemorrhage (ICH). SAH was excluded because it often does not cause direct damage to the brain tissue. PS is typically indexed by timed tests of speeded behavior that usually entail repetitive perceptual-motor and cognitive operations and responses. Considering that assessment of PS in stroke patients is complicated by the existence of a motor response in many tests purported to tap this construct (Lovell & Smith, 1997), we chose the Symbol Digit Modalities Test (SDMT; Smith, 1982), a task involving matching numbers to symbols, for several reasons. First, the SDMT has nonspecific sensitivity to brain dysfunction and has been widely used and repeatedly validated in various clinical populations including stroke (Lezak, Howieson, & Lowing, 2004). Second, the SDMT allows for oral administration (equivalent to the written version) to exclude confounding effects of hand laterality on test performance for those whose dominant limbs are hemiparetic following a stroke. Third, although no neuropsychological measure can be claimed to be process pure, research to date strongly suggests that cognitive speed is the prime determinant of the coding task performance, with other elementary mental processes such as working memory and selective attention playing a subsidiary role (Joy, Kaplan, & Fein, 2004; Lezak et al., 2004; Salthouse, 1996). As such, differences in performance on SDMT reflect the variation in the time taken to execute the relevant cognitive operations needed by the task.

Method

Participants

A convenience sample of 30 stroke patients (19 men, 11 women) was drawn from an outpatient stroke rehabilitation program at an affiliated teaching hospital of the Kaohsiung Medical University (KMU). Eligibility criteria included a first-ever unilateral hemispheric stroke, at least 3 months post onset to minimize confounding by spontaneous recovery, a score of 24 or higher on the mini-mental state examination (MMSE; Folstein, Folstein, & McHugh, 1975) for those aged 60 and over to rule out possible dementia, and ability to communicate intelligibly and follow instructions to ensure the validity of the neuropsychological test scores obtained. Patients who had history of psychiatric or other neurological diseases, alcohol, or drug abuse, subdural hematomas, arteriovenous malformation, SAH, or transient ischemic attack were excluded because these medical conditions may offer an alternative explanation for study outcomes. For a similar reason, presence of aphasia, hemianopia, or unilateral neglect defined as ≥1 point for best language, visual fields, and neglect items on the modified National Institute of Health Stroke Scale (mNIHSS; Lyden, Lu, Levine, Brott, & Broderick, 2001) was excluded. Stroke was diagnosed according to the World Health Organization (WHO) criteria (WHO MONICA Project Principal Investigators, 1988). In all cases, computed tomography (CT) and/or magnetic resonance imaging (MRI) scans were carried out to confirm the diagnosis.

The mean age of the stroke sample was 51 years (range 23–74 years) and the mean number of years of education was 11 (range 6–16) (Table 1). There were equal numbers of hemorrhagic and ischemic strokes. Among them, 67% had right hemisphere lesions. In terms of vascular risk factors, 73.3% had hypertension alone, 10% had diabetes mellitus (DM) alone, 6.7% had heart disease alone, 6.7% had both hypertension and DM, and 3.3% had both hypertension and heart disease before stroke. Twelve (40%) patients were classified as having a mild stroke (mNIHSS score < 5) and 18 (60%) as having a moderate stroke (mNIHSS score, 5–15). The time interval between onset and assessment ranged from 3 months to 10 years. Five (16.7%) patients had a stroke within 6 months, 6 (20%) 6–12 months, 13 (43.3%) 13–36 months, and 6 (20%) over 36 months. These patients were reclassified into three groups: ≤1 year (n = 11), 1–3 year (n = 13), and >3 year (n = 6). Using a cutoff point of ≥5 on the 15-item Geriatric Depression Scale (GDS-15; Sheikh &Yesavage, 1986), 15 patients were categorized as depressed.

Table 1.

Demographic and clinical characteristics of the study participants

 Patients (n = 30) Controls (n = 30) 
Age, M (SD50.6 (12.1) 50.8 (11.4) 
% Women 11 (36.7) 20 (66.7) 
Education, M (SD10.7 (3.5) 12.0 (2.8) 
GDS score, M (SD5.1 (3.3) 3.4 (3.2) 
mNIHSS (Median) 6.0 — 
% Cerebral infarction 50 — 
% Right brain damage 66.7 — 
Months post-stroke (frequency) 
 ≤1 year 11 — 
 1–3 years 13 — 
 >3 years — 
Site of lesion (frequency)a 
 Basal ganglia 12 — 
 Thalamus — 
 Pons — 
 Corona radiate — 
 Internal capsule — 
 Middle cerebral artery — 
Cerebrovascular risk factor (frequency) 
 Hypertension 22 — 
 DM — 
 Heart disease — 
 Both hypertension and DM — 
 Both hypertension and heart disease — 
 Patients (n = 30) Controls (n = 30) 
Age, M (SD50.6 (12.1) 50.8 (11.4) 
% Women 11 (36.7) 20 (66.7) 
Education, M (SD10.7 (3.5) 12.0 (2.8) 
GDS score, M (SD5.1 (3.3) 3.4 (3.2) 
mNIHSS (Median) 6.0 — 
% Cerebral infarction 50 — 
% Right brain damage 66.7 — 
Months post-stroke (frequency) 
 ≤1 year 11 — 
 1–3 years 13 — 
 >3 years — 
Site of lesion (frequency)a 
 Basal ganglia 12 — 
 Thalamus — 
 Pons — 
 Corona radiate — 
 Internal capsule — 
 Middle cerebral artery — 
Cerebrovascular risk factor (frequency) 
 Hypertension 22 — 
 DM — 
 Heart disease — 
 Both hypertension and DM — 
 Both hypertension and heart disease — 

Notes: GDS = Geriatric Depression Scale; mNIHSS = modified National Institute of Health Stroke Scale; DM = Diabetes mellitus.

aIn eight (26.7%) patients, the lesion data were not available.

Although all 30 patients underwent CT scan or MRI, the reports were available for 22 patients, while the remaining eight patients (ICH = 1, ischemia = 7) had CT scanning at other hospitals and reports were not available in the medical records. Of these 22 patients, basal ganglia (BG) (54.5%) was the commonest site of ICH followed by thalamus (9.1%), whereas five of eight ischemic stroke patients had supratentorial lesions which were located in corona radiata, internal capsule, and middle cerebral artery territory, with the remaining patients having infratentorial lesions which involved the pons.

For comparison purposes, a control group (n = 30) from the general community (spouses or family of patients or volunteers who came to our attention by word of mouth), consisting of participants who had no history of neurological or psychiatric problems, was matched group-wise with stroke patients on age and education. Excluded were those aged 60 or older whose MMSE (Folstein et al., 1975) score was 23 points or lower. The control sample aged between 25 and 71, with a mean of 51 years (Table 1). There were 10 men and 20 women, with an average of 12 years of schooling (range 6–16 years). 26.7% were classified as depressed. The current study was approved by the institutional review board of the KMU Hospital, and written informed consent was obtained from all participants prior to study.

Procedure

Clinical data were obtained using the 30 included patients' medical records. Assessment of neuropsychological function and depression was performed by two occupational therapists who were highly trained and closely supervised by the first author (C.Y.S.). These tests were administered in a fixed order according to standardized procedures to ensure consistency in administration and scoring. Neuropsychological testing took ∼2–3 h conducted in two to three sessions on different days in the same week to minimize fatigue.

Instruments

Global cognitive functioning. The MMSE (Folstein et al., 1975). It is an 11-item measure that covers five areas of function, namely orientation, memory, attention, language, and visuoconstructive abilities. The maximum possible total score is 30 points.

Depression. The GDS-15 (Sheikh &Yesavage, 1986). Participants were asked to answer “yes” or “no” to 15 questions about how they have felt over the past week. Scores on the GDS-15 were graded as normal (0–4), mild (5–9), and moderate to severe (10–15).

A series of neuropsychological tests were grouped a priori into the following six conceptual domains according to the principal cognitive function captured by the tests.

Processing speed. The SDMT oral version (Smith, 1982). Participants were given a sheet of paper at the top of which is a printed key that pairs each abstract symbol with a different number (1–9). Eight rows containing small blank squares, each paired with a randomly assigned symbol, are presented below the key. Following a practice trial on the first 10 squares, the examiner, on a copy of the test sheet, recorded in the empty squares the numbers the participant associated orally with the symbols within the 90-s time limit.

Verbal memory. The Chinese version of the Wechsler Memory Scale-Third Edition Logical Memory subtest (WMS-III;Wechsler, 1997). Participants were asked to immediately reproduce two short passages read to them. Thirty minutes after this immediate recall trial, delayed recall was tested.

Visual memory. The Rey-Osterrieth Complex Figure Test immediate and delayed recall trials (ROCFT; Meyers & Meyers, 1995). Participants were first shown a complex geometric figure and asked to draw the same figure. Immediately after the copy trial, they were instructed to draw the design from memory. After 30 min delayed recall followed.

Visuoperceptual function. The ROCFT copy trial (Meyers & Meyers, 1995) and the Hooper Visual Organization Test (HVOT; Western Psychological Services, 1983). The HVOT consists of 30 drawings of common objects and animals that have been cut into two or more pieces. Participants were requested to identify and name each item after reorganizing these pieces mentally into a coherent whole.

Language. The Boston Naming Test-Second Edition (BNT-2; Kaplan, Goodglass, & Weintraub, 2001) short form. Participants were presented with 15 line drawings of everyday objects and asked to name each one. Failure to spontaneously name the object leads to the administration of semantic and phonemic cues as necessary.

The Chinese version of the Luria–Nebraska Neuropsychological Battery (LNNB;Golden, Purisch, & Hammeke, 1985) expressive and receptive speech scales. On the expressive scale, participants were asked to repeat simple and complex phrases, describe pictures, and make speeches on a given topic after seeing a picture or hearing a short story. On the receptive speech scale, participants were required to discriminate phonemes, read words, comprehend the meaning of words and sentences, and follow verbal commands.

Cognitive flexibility. The Wisconsin Card Sorting Test-64 card version (WCST-64;Kongs, Thompson, Iverson, & Heaton, 2000). Participants were given one pack containing 64 response cards, which have designs similar to those on the stimulus cards, varying in color, geometric form, and number. Participants were then asked to match the response cards with the stimulus cards and instructed to infer the matching principle from the feedback provided (either “right” or “wrong”). Two indices of the WCST were computed, including perseverative responses and conceptual level responses.

Data Analyses

Statistical analyses were carried out with SPSS for Windows (version 15.0). Demographic differences between stroke patients and healthy controls were compared using independent sample t-tests for continuous variables, and χ2 tests for categorical variables. Pearson correlations were implemented to assess the interrelations between cognitive measures, depression, and stroke severity. The associations between cognitive and clinical variables were evaluated using the Kruskal–Wallis and Mann–Whitney U-tests as appropriate.

As normative data on Chinese adults are not available for the subtests in the language domain, raw neuropsychological test scores of the stroke group were re-scaled to obtain z-scores using the mean and SD of the controls. The sign of some variables for which high scores were in the impaired direction (such as LNNB speech scales and WCST perseverative responses) was reversed, so that negative z-scores reflected poor performance. Composite scores were then constructed by averaging the z-scores of the respective test measures selected as representative of each domain except PS which contained only one test. The purpose of creating domain-specific composite scores was to increase the power of the analyses by reducing the number of comparisons owing to the relatively large number of dependent variables. While there is a lack of consensus on how cognitive impairment should be defined, a z-score cutoff of −1.65 (corresponding to the fifth percentile) was adopted in the present study as it is significant at the .05 level for a one-tailed test. Cognitive domain and individual test scores equal to or lower than this cutoff were defined as deficient. For each cognitive variable, the frequency of patients who scored in the impaired range was calculated.

To examine whether stroke patients have different neurocognitive test profiles when compared with healthy controls, all subtests and domain scores were treated as dependent variables. Multivariate analysis of covariance (MANCOVA) and univariate analysis of covariance (ANCOVA) were applied. Age, gender, education, and depression were included as covariates in all analyses based on literature. Adjustment of α for multiple comparisons was performed with Bonferroni statistics.

Our conjecture that impaired PS may contribute significantly to disturbances in other more complex cognitive functions was tested under the assumption that statistical control of the variance in PS relative to other cognitive factors would lead to the greatest attenuation of the group differences in the remaining cognitive domains. As a result, six MANCOVAs were performed, each involving only one cognitive covariate (PS and then the rest of the cognitive domains), to determine whether stroke-related cognitive impairments diminished in a different way depending on which domain was used as a covariate.

The effect size indicator partial eta squared (η2) was reported with each ANCOVA/MANCOVA results as a measure of the strength of the association, with η2 = 0.01, 0.06, and 0.14 representing small, medium, and large effect sizes, respectively (Cohen, 1988). The partial η2 (analogous to the R2 statistic in regression) denotes the proportion of variance in cognitive performance associated with group (patients/controls) that is not attributed to the covariates.

Results

Sample Characteristics

The patients and controls did not differ significantly in age (t = −0.06, p = .96) or education (t = −1.59, p = .12), though they did differ in sex ratio (χ2 = 5.41, p = .02), with more women in the control group and more men in the patient group. There was a marginally significant difference between groups in GDS-15 scores (t = 1.95, p = .056).

Intercorrelations between Measures of Cognitive Function and Their Correlations with Depression and Stroke Severity

In the stroke group, the SDMT correlated significantly with LM immediate recall (r = 0.39), LM delayed recall (r = 0.40), ROCFT immediate recall (r = 0.42), ROCFT delayed recall (r = 0.40), ROCFT copy (r = 0.43), HVOT (r = 0.41), LNNB expressive speech (r = −0.52), WCST perseverative responses (r = −0.41), and conceptual level responses (r = 0.53), but not with BNT (r = 0.30) and LNNB receptive speech (r = −0.35) (Table 2).

Table 2.

Intercorrelations between cognitive measures and depression in patients and controls

 Stroke patients
 
Healthy controls
 
PS VM VSM VP Lang CF PS VM VSM VP Lang CF 
PS             
VM 0.42**      0.69***      
VSM 0.42** 0.26     0.39* 0.34     
VP 0.47** 0.24 0.57**    0.54** 0.45* 0.71***    
Lang 0.50** 0.67*** 0.27 0.30   0.69*** 0.58** 0.60*** 0.61***   
CF 0.48** 0.25 0.31 0.54** 0.21  0.34 0.27 0.35 0.56** 0.50**  
GDS −0.10 0.16 −0.12 −0.06 −0.02 0.28 −0.25 −0.21 −0.04 −0.27 −0.31 −0.44* 
mNIHSS −0.52** −0.09 −0.22 −0.29 −0.07 −0.17 — — — — — — 
 Stroke patients
 
Healthy controls
 
PS VM VSM VP Lang CF PS VM VSM VP Lang CF 
PS             
VM 0.42**      0.69***      
VSM 0.42** 0.26     0.39* 0.34     
VP 0.47** 0.24 0.57**    0.54** 0.45* 0.71***    
Lang 0.50** 0.67*** 0.27 0.30   0.69*** 0.58** 0.60*** 0.61***   
CF 0.48** 0.25 0.31 0.54** 0.21  0.34 0.27 0.35 0.56** 0.50**  
GDS −0.10 0.16 −0.12 −0.06 −0.02 0.28 −0.25 −0.21 −0.04 −0.27 −0.31 −0.44* 
mNIHSS −0.52** −0.09 −0.22 −0.29 −0.07 −0.17 — — — — — — 

Notes: PS = processing speed; VM = verbal memory; VSM = visuospatial memory; VP = visuoperceptual function; Lang = language; CF = cognitive flexibility; GDS = Geriatric Depression Scale; mNIHSS = modified National Institute of Health Stroke Scale.

*p < .05; **p < .01; ***p < .001.

At the domain level, significant correlations between SDMT and other measures of cognitive function were observed in the patient group, whereas in healthy controls, the SDMT was significantly correlated to all but cognitive flexibility which approached but did not reach statistical significance (p = .066). The correlations were in the expected direction in patients, with the strongest association found between verbal memory and language. In the control group, almost all correlations between cognitive measures were >.30, with the highest correlation found between visual memory and visuoperceptual function. Depression did not correlate with performance on any cognitive domain in stroke patients, while in the control group depression was only correlated with cognitive flexibility. mNIHSS score correlated moderately with PS in stroke patients, but it did not correlate significantly with other cognitive domains.

Relationships between Cognitive Domains and Clinical Factors

Table 3 describes the impact of clinical variables on the presence of cognitive impairment. The Mann–Whitney test revealed no significant differences between patients with left and right hemispheric stroke in any of the cognitive domains, nor did any significant cognitive differences exist between ischemic stroke and ICH. Visual memory was the only domain that was significant between patients with hypertension and those with other vascular risk factors (i.e., DM and heart disease). Patients with stroke confined to BG had significantly higher scores on cognitive flexibility than patients with non-BG lesions. As for the effect of onset–assessment interval, only cognitive flexibility showed significance in a Kruskal–Wallis test, with patients who had stroke for >3 years performing worst.

Table 3.

Relationship of clinical variables to specific cognitive domain scores

 PS VM VSM VP Lang CF 
Lesion side 
 LH (n = 10) −2.93 (0.79) −1.17 (1.45) −1.19 (1.12) −1.30 (2.30) −2.41 (2.25) −2.10 (2.25) 
 RH (n = 20) −2.67 (0.96) −1.09 (0.94) −1.73 (1.21) −2.40 (2.56) −1.72 (1.71) −2.15 (2.15) 
Mann–Whitney U test 84.00 94.00 72.00 69.50 84.50 95.50 
Stroke type 
 Ischemia (n = 15) −2.80 (0.91) −1.34 (0.93) −1.81 (1.13) −2.62 (2.60) −2.19 (2.08) −2.69 (2.31) 
 ICH (n = 15) −2.72 (0.92) −0.89 (1.25) −1.29 (1.23) −1.45 (2.32) −1.71 (1.73) −1.58 (1.88) 
Mann–Whitney U test 108.00 86.00 85.00 67.00 97.00 78.50 
Lesion site 
 BG (n = 12) −2.60 (0.98) −0.96 (1.17) −1.38 (1.37) −1.76 (2.47) −2.04 (1.73) −1.03 (1.39) 
 Others (n = 10) −2.97 (0.81) −1.09 (1.22) −1.42 (1.27) −2.26 (3.01) −1.75 (2.32) −3.02 (1.98) 
Mann–Whitney U test 45.50 49.00 56.00 52.50 43.50 22.00* 
Vascular risk factors 
 Hypertension (n = 22) −2.65 (0.93) −1.15 (1.17) −1.24 (1.15) −1.22 (1.74) −1.880 (1.81) −1.82 (1.87) 
 Others (n = 5) −2.75 (0.87) −0.72 (0.59) −2.63 (0.66) −3.31 (2.69) −1.74 (1.71) −1.77 (2.61) 
Mann–Whitney U test 50.000 39.500 18.000* 27.000 55.000 50.000 
Onset–assessment time 
 ≤1 year (n = 11) −2.31 (0.98) −0.93 (1.20) −1.10 (1.36) −1.05 (1.83) −1.72 (2.02) −0.85 (1.84) 
 1–3 years (n = 13) −3.03 (0.83) −1.42 (1.19) −1.59 (0.95) −2.17 (2.67) −2.24 (2.02) −2.67 (2.19) 
 >3 years (n = 6) −2.99 (0.65) −0.80 (0.59) −2.28 (1.12) −3.54 (2.73) −1.76 (1.58) −3.33 (1.54) 
Kruskal–Wallis test 4.42 2.87 3.95 4.99 0.78 8.68* 
 PS VM VSM VP Lang CF 
Lesion side 
 LH (n = 10) −2.93 (0.79) −1.17 (1.45) −1.19 (1.12) −1.30 (2.30) −2.41 (2.25) −2.10 (2.25) 
 RH (n = 20) −2.67 (0.96) −1.09 (0.94) −1.73 (1.21) −2.40 (2.56) −1.72 (1.71) −2.15 (2.15) 
Mann–Whitney U test 84.00 94.00 72.00 69.50 84.50 95.50 
Stroke type 
 Ischemia (n = 15) −2.80 (0.91) −1.34 (0.93) −1.81 (1.13) −2.62 (2.60) −2.19 (2.08) −2.69 (2.31) 
 ICH (n = 15) −2.72 (0.92) −0.89 (1.25) −1.29 (1.23) −1.45 (2.32) −1.71 (1.73) −1.58 (1.88) 
Mann–Whitney U test 108.00 86.00 85.00 67.00 97.00 78.50 
Lesion site 
 BG (n = 12) −2.60 (0.98) −0.96 (1.17) −1.38 (1.37) −1.76 (2.47) −2.04 (1.73) −1.03 (1.39) 
 Others (n = 10) −2.97 (0.81) −1.09 (1.22) −1.42 (1.27) −2.26 (3.01) −1.75 (2.32) −3.02 (1.98) 
Mann–Whitney U test 45.50 49.00 56.00 52.50 43.50 22.00* 
Vascular risk factors 
 Hypertension (n = 22) −2.65 (0.93) −1.15 (1.17) −1.24 (1.15) −1.22 (1.74) −1.880 (1.81) −1.82 (1.87) 
 Others (n = 5) −2.75 (0.87) −0.72 (0.59) −2.63 (0.66) −3.31 (2.69) −1.74 (1.71) −1.77 (2.61) 
Mann–Whitney U test 50.000 39.500 18.000* 27.000 55.000 50.000 
Onset–assessment time 
 ≤1 year (n = 11) −2.31 (0.98) −0.93 (1.20) −1.10 (1.36) −1.05 (1.83) −1.72 (2.02) −0.85 (1.84) 
 1–3 years (n = 13) −3.03 (0.83) −1.42 (1.19) −1.59 (0.95) −2.17 (2.67) −2.24 (2.02) −2.67 (2.19) 
 >3 years (n = 6) −2.99 (0.65) −0.80 (0.59) −2.28 (1.12) −3.54 (2.73) −1.76 (1.58) −3.33 (1.54) 
Kruskal–Wallis test 4.42 2.87 3.95 4.99 0.78 8.68* 

Notes: PS = processing speed; VM = verbal memory; VSM = visuospatial memory; VP = visuoperceptual function; Lang = language; CF = cognitive flexibility; LH = left hemisphere; RH = right hemisphere; ICH = intracerebral hemorrhage; BG = basal ganglia.

*p < .05.

Profile and Prevalence of Cognitive Impairment

Regarding the overall group differences, a MANCOVA was executed to examine the effect of the groups on the six domain scores, considered collectively as dependent variables. Results of this test showed a significant omnibus effect (Wilks' Λ = 0.18, F(6,49) = 36.60, p < .0001, η2 = 0.82) of group on overall cognitive functioning. Univariate tests revealed a significance for the PS (F(1,54) = 190.27, p < .0001, η2 = 0.78), verbal memory (F(1,54) = 10.13, p = .002, η2 = 0.16), visual memory (F(1,54) = 33.17, p < .0001, η2 = 0.38), visuoperceptual function (F(1,54) = 16.02, p < .0001, η2 = 0.23), language (F(1,54) = 17.16, p < .0001, η2 = 0.24), and cognitive flexibility (F(1,54) = 28.21, p < .0001, η2 = 0.34). Pairwise comparisons, adjusted by using the Bonferroni correction, demonstrated that stroke patients performed significantly worse than controls on all cognitive domains.

At the subtest level, a one-way MANCOVA was performed to compare the performance of the stroke group to the control group when the two verbal memory subtests were entered as dependent variables. This MANCOVA was significant beyond the .0001 level (Table 4). Follow-up ANCOVAs revealed the stroke group performed poorer than the control group on immediate and delayed trials of the LM stories. Similarly, significant differences between the groups were obtained for all subtests in their respective domains of visual memory, visuoperceptual function, language, and cognitive flexibility.

Table 4.

Neuropsychological performance by study group and cognitive domain

 Patients Control group Partial η2 p-value 
Processing speed F(1, 54) = 190.27 
SDMT 27.2 (8.2) 52.3 (9.1) 0.78 <.0001 
Verbal memory Wilks' Λ = 0.67, F(2, 53) = 13.10, p < .0001; 0.33 
 WMS-III logical memory immediate recall 28.8 (10.6) 38.4 (9.1) 0.12 .01 
 WMS-III logical memory delayed recall 17.9 (7.5) 25.5 (6.5) 0.30 <.0001 
Visuospatial memory Wilks’ Λ = 0.56, F(2, 53) = 21.23, p < .0001; 0.45 
 ROCFT immediate recall 14.3 (9.3) 25.1 (7.4) 0.32 <.0001 
 ROCFT delayed recall 12.4 (8.2) 23.9 (7.0) 0.42 <.0001 
Visuoperceptual function Wilks’ Λ = 0.66, F(2, 53) = 13.90, p < .0001; 0.34 
 HVOT 17.4 (4.3) 22.6 (4.1) 0.32 <.0001 
 ROCFT copy 30.6 (7.0) 35.0 (1.6) 0.14 .004 
Language Wilks’ Λ = 0.71, F(3, 52) = 7.01, p < .0001; 0.29 
 BNT 10.8 (2.2) 13.1 (1.8) 0.25 <.0001 
 LNNB expressive speech scale 3.8 (3.7) 0.8 (1.6) 0.17 .002 
 LNNB receptive speech scale 8.0 (6.1) 2.8 (1.9) 0.15 .003 
Cognitive flexibility Wilks’ Λ = 0.66, F(2, 53) = 13.86, p < .0001; 0.34 
 WCST perseverative responses 23.2 (16.6) 8.5 (5.2) 0.32 <.0001 
 WCST conceptual level responses 25.0 (16.8) 42.4 (12.4) 0.26 <.0001 
 Patients Control group Partial η2 p-value 
Processing speed F(1, 54) = 190.27 
SDMT 27.2 (8.2) 52.3 (9.1) 0.78 <.0001 
Verbal memory Wilks' Λ = 0.67, F(2, 53) = 13.10, p < .0001; 0.33 
 WMS-III logical memory immediate recall 28.8 (10.6) 38.4 (9.1) 0.12 .01 
 WMS-III logical memory delayed recall 17.9 (7.5) 25.5 (6.5) 0.30 <.0001 
Visuospatial memory Wilks’ Λ = 0.56, F(2, 53) = 21.23, p < .0001; 0.45 
 ROCFT immediate recall 14.3 (9.3) 25.1 (7.4) 0.32 <.0001 
 ROCFT delayed recall 12.4 (8.2) 23.9 (7.0) 0.42 <.0001 
Visuoperceptual function Wilks’ Λ = 0.66, F(2, 53) = 13.90, p < .0001; 0.34 
 HVOT 17.4 (4.3) 22.6 (4.1) 0.32 <.0001 
 ROCFT copy 30.6 (7.0) 35.0 (1.6) 0.14 .004 
Language Wilks’ Λ = 0.71, F(3, 52) = 7.01, p < .0001; 0.29 
 BNT 10.8 (2.2) 13.1 (1.8) 0.25 <.0001 
 LNNB expressive speech scale 3.8 (3.7) 0.8 (1.6) 0.17 .002 
 LNNB receptive speech scale 8.0 (6.1) 2.8 (1.9) 0.15 .003 
Cognitive flexibility Wilks’ Λ = 0.66, F(2, 53) = 13.86, p < .0001; 0.34 
 WCST perseverative responses 23.2 (16.6) 8.5 (5.2) 0.32 <.0001 
 WCST conceptual level responses 25.0 (16.8) 42.4 (12.4) 0.26 <.0001 

Notes: SDMT = Symbol Digit Modalities Test; WMS-III = Wechsler Memory Scale-Third Edition; ROCFT = Rey-Osterrieth Complex Figure Test; HVOT = Hooper Visual Organization Test; BNT = Boston Naming Test; LNNB = Luria–Nebraska Neuropsychological Battery; WCST = Wisconsin Card Sorting Test.

It was apparent from the correlation analysis above that there existed shared variance between SDMT and other cognitive measures. Subsequently, we re-ran the ANCOVA to adjust for the effects of these cognitive domains. The results showed that patients remained significantly impaired on the SDMT (F(1,49) = 69.54, p < .0001, η2 = 0.59), supporting the test's sensitivity to stroke pathology as previously reported (Lezak et al., 2004) and attesting to the robustness of PS impairment in this population.

As seen in Table 5, the proportion of patients who were classified as cognitively impaired on each test ranged from 26.7% (LM immediate recall) to 90% (SDMT). Domain wise, the highest prevalence was in PS, and the lowest prevalence was in verbal memory. Two (6.7%) patients demonstrated no deficits across cognitive domains, 4 (13.3%) were deficient in one domain, 3 (10%) were deficient in two domains, and 21 (70%) were deficient in three or more domains.

Table 5.

Mean z-scores and the ratio of stroke patients performing in the impaired range

Neuropsychological tests/domains Mean (SDRange Impairment ratioa (%) 
SDMT −2.76 (0.9) −4.42 to −0.91 90 
Verbal memory −1.12 (1.1) −3.23 to 1.47 30 
 WMS-III logical memory immediate recall −1.05 (1.2) −3.44 to 1.38 26.7 
 WMS-III logical memory delayed recall −1.18 (1.2) −3.17 to 1.78 36.7 
Visual memory −1.55 (1.2) −3.31 to 0.77 53.3 
 ROCFT immediate recall −1.46 (1.3) −3.33 to 0.8 50 
 ROCFT delayed recall −1.64 (1.2) −3.29 to 0.73 56.7 
Visuoperceptual function −2.03 (2.5) −8.32 to 0.55 43.3 
 HVOT −1.27 (1.1) −3.33 to 0.72 40 
 ROCFT copy −2.80 (4.4) −15.52 to 0.94 40 
Language −1.95 (1.9) −6.65 to 0.80 46.7 
 BNT −1.29 (1.2) −4.48 to 1.05 60 
 LNNB expressive speech scale −2.73 (3.3) −11.79 to 1.51 56.7 
 LNNB receptive speech scale −1.84 (2.3) −7.5 to 0.47 40 
Cognitive flexibility −2.13 (2.1) −6.68 to 0.96 53.3 
 WCST-64 perseverative responses −2.86 (3.2) −10.18 to 0.67 50 
 WCST-64 conceptual level responses −1.40 (1.4) −3.42 to 1.25 46.7 
Neuropsychological tests/domains Mean (SDRange Impairment ratioa (%) 
SDMT −2.76 (0.9) −4.42 to −0.91 90 
Verbal memory −1.12 (1.1) −3.23 to 1.47 30 
 WMS-III logical memory immediate recall −1.05 (1.2) −3.44 to 1.38 26.7 
 WMS-III logical memory delayed recall −1.18 (1.2) −3.17 to 1.78 36.7 
Visual memory −1.55 (1.2) −3.31 to 0.77 53.3 
 ROCFT immediate recall −1.46 (1.3) −3.33 to 0.8 50 
 ROCFT delayed recall −1.64 (1.2) −3.29 to 0.73 56.7 
Visuoperceptual function −2.03 (2.5) −8.32 to 0.55 43.3 
 HVOT −1.27 (1.1) −3.33 to 0.72 40 
 ROCFT copy −2.80 (4.4) −15.52 to 0.94 40 
Language −1.95 (1.9) −6.65 to 0.80 46.7 
 BNT −1.29 (1.2) −4.48 to 1.05 60 
 LNNB expressive speech scale −2.73 (3.3) −11.79 to 1.51 56.7 
 LNNB receptive speech scale −1.84 (2.3) −7.5 to 0.47 40 
Cognitive flexibility −2.13 (2.1) −6.68 to 0.96 53.3 
 WCST-64 perseverative responses −2.86 (3.2) −10.18 to 0.67 50 
 WCST-64 conceptual level responses −1.40 (1.4) −3.42 to 1.25 46.7 

Notes: SDMT = Symbol Digit Modalities Test; WMS-III = Wechsler Memory Scale-Third Edition; ROCFT = Rey-Osterrieth Complex Figure Test; HVOT = Hooper Visual Organization Test; BNT = Boston Naming Test; LNNB = Luria–Nebraska Neuropsychological Battery; WCST = Wisconsin Card Sorting Test.

aA cognitive deficit was defined as a z-score of ≤−1.65 (below the fifth percentile).

Between-Group Differences after Controlling for PS and Other Cognitive Factors

After inclusion of the SDMT as an additional covariate, the effect of group on cognition was no longer significant (Wilks'Λ = 0.83, F(5,49) = 2.08, p = .08, η2 = 0.18), with a major attenuation in group-related variance (64%). ANCOVA tests presented in Table 6 found no significant between-group differences in the remaining five domains.

Table 6.

Follow-up ANCOVA results for differences in cognitive performance between patients and controls after controlling for neuropsychological domains

Cognitive domain PS as a covariate
 
VM as a covariate
 
VSM as a covariate
 
VP as a covariate
 
Lang as a covariate
 
CF as a covariate
 
F partial η2 F partial η2 F partial η2 F partial η2 F partial η2 F partial η2 
PS — — 131.94*** 0.713 105.44*** 0.665 132.22*** 0.714 136.08*** 0.720 110.45*** 0.676 
VM 0.14 0.003 — — 10.56** 0.166 9.25** 0.149 0.98 0.018 9.06** 0.146 
VSM 3.54 0.063 25.95*** 0.329 — — 14.88*** 0.219 17.81*** 0.251 18.63*** 0.260 
VP 0.09 0.002 9.10** 0.147 1.52 0.028 — — 6.81* 0.114 1.82 0.033 
Lang 2.09 0.038 4.78* 0.083 4.81* 0.083 7.79** 0.128 — — 9.28*** 0.149 
CF 2.01 0.037 19.70*** 0.271 14.56*** 0.215 11.37** 0.177 18.94*** 0.263 — — 
Cognitive domain PS as a covariate
 
VM as a covariate
 
VSM as a covariate
 
VP as a covariate
 
Lang as a covariate
 
CF as a covariate
 
F partial η2 F partial η2 F partial η2 F partial η2 F partial η2 F partial η2 
PS — — 131.94*** 0.713 105.44*** 0.665 132.22*** 0.714 136.08*** 0.720 110.45*** 0.676 
VM 0.14 0.003 — — 10.56** 0.166 9.25** 0.149 0.98 0.018 9.06** 0.146 
VSM 3.54 0.063 25.95*** 0.329 — — 14.88*** 0.219 17.81*** 0.251 18.63*** 0.260 
VP 0.09 0.002 9.10** 0.147 1.52 0.028 — — 6.81* 0.114 1.82 0.033 
Lang 2.09 0.038 4.78* 0.083 4.81* 0.083 7.79** 0.128 — — 9.28*** 0.149 
CF 2.01 0.037 19.70*** 0.271 14.56*** 0.215 11.37** 0.177 18.94*** 0.263 — — 

Notes: ANCOVA = analysis of covariance; PS = processing speed; VM = verbal memory; VSM = visuospatial memory; VP = visuoperceptual function; CF = cognitive flexibility.

*p < .05; **p < .01; ***p < .001.

In contrast, controlling for verbal memory did not change the overall between-group differences in the results (Wilks' Λ = 0.22, F(5,49) = 35.44, p < .0001, η2 = 0.78), nor did the control of visual memory (Wilks' Λ = 0.29, F(5,49) = 23.48, p < .0001, η2 = 0.71), visuoperceptual function (Wilks' Λ = 0.24, F(5,49) = 31.63, p < .0001, η2 = 0.76), language (Wilks' Λ = 0.24, F(5,49) = 30.96, p < .0001, η2 = 0.76), and cognitive flexibility (Wilks' Λ = 0.28, F(5,49) = 25.49, p < .0001, η2 = 0.72). Follow-up univariate analyses revealed that between-subjects effect was reduced to non-significance for visuoperceptual function when covarying for visual memory (Table 6). The relatively high correlations between these two domains may account for the attenuation of group differences. Likewise, after partialling out the effects of language and cognitive flexibility, between-group differences dropped to non-significance for verbal memory and visuoperceptual function, respectively.

Discussion

Our stroke cohort displayed varying degrees of impairment across a range of cognitive domains including PS, visual memory, visuoperceptual function, language, and cognitive flexibility. Consistent with some studies (Hochstenbach et al., 1998; Reed et al., 2007), verbal memory was less affected compared with the other cognitive domains. An alternative explanation may lie in the overrepresentation of patients with right hemisphere stroke in our research as verbal memory seems to be more sensitive to left hemisphere lesions (Gillespie, Bowen, & Foster, 2006). Apart from this, inclusion of SAH patients with damage to the basal forebrain and frontal lobes secondary to ruptured anterior communicating artery aneurysms may alter this cognitive profile since these brain regions are known to be importantly involved in memory (Damasio, Graff-Radford, Eslinger, Damasio, & Kassell, 1985; Ravnik et al., 2006). In all, 70% of patients still experienced deficiency in three or more domains 3 months or more after onset. More than 50% of patients performed in the impaired range on SDMT, ROCFT immediate and delayed recall, BNT, expressive speech scale, and perseverative responses. Among these, PS was found to be the cognitive function that is invariably affected in terms of both prevalence and severity. In particular, PS remains conspicuously deficient after adjusting for demographic features and other cognitive factors. Zinn, Bosworth, Hoenig, and Swartzwelder (2007) who defined impairment by a cutoff score of 1.5 SD below the mean, which is less rigorous than the statistical criterion chosen in this study, reported a lower percentage (75%) of patients with acute stroke scoring below this cutoff point on oral version of the SDMT. This leads to speculation that the PS deficit in stroke may be far more pervasive than previously assumed. Longitudinal studies that evaluate temporal changes in the recovery of PS deficit after first-ever stroke are critical.

The primacy of PS in governing performance on the more complex aspects of cognitive function in stroke patients is confirmed, echoing the results of the research on aging and other cognitively impaired populations (Butters et al., 2004; Diamond et al., 2008; McGrath et al., 2011; McKinlay et al., 2009; Rodríguez-Sánchez et al., 2007; Willmott et al., 2009). That is, partialling out the effect of PS substantially attenuated group-related variance in each of the cognitive domains examined, whereas statistical control of other cognitive factors failed to yield the same effects of attenuation. Despite its prominent contribution to cognitive functioning, the nature of PS in stroke remains unclear and largely unexplored. There has been some research into the impact of hemispheric lateralization on PS after stroke. For example, using reaction time paradigms, Gerritsen, Berg, Deelman, Visser-Keizer, and Meyboom-de Jong (2003) found that right hemisphere-damaged stroke patients performed worse than those with left hemisphere lesions on visuomotor decision times that placed high demands on spatial attentional capacity. The left hemisphere-damaged patients were significantly slower compared with healthy controls only in the most complex tasks, the categorization tasks (semantic and pictorial). In our study, no such group differences were detected using perceptual speed measure. This is congruent with previous research supporting the view that the SDMT is sensitive to bilateral processing (Smith, 1982), which is further substantiated by results from brain imaging studies in that better performance on coding was generally associated with structural integrity of the WM in the whole brain (Ferrer et al., 2013; Penke et al., 2010; Vernooij et al., 2009). These neuroimaging findings are not unexpected, considering that the cognitive processes required for performing the SDMT are multiple, including motor speed, selective attention, working memory, relational memory, response selection and execution, and visual scanning and matching, which undoubtedly depend on recruitment and integration of large-scale functional neural circuitry that are also involved in most, if not all, higher cognitive functions (Rypma et al., 2006; Salthouse, 1996; Turken et al., 2008). Turken and colleagues (2008) took aim at exploring the relation between PS and structural properties of WM pathways in stroke patients and found that during performance on the digit-symbol substitution test, posterior parietal regions are necessary for establishing and maintaining the digit-symbol pairing, which in turn are mediated through mechanisms of attention control and regulation of actions in frontal–striatal circuit. The temporal lobe WM regions are critical for relaying the analysis of visual features to frontal lobes responsible for the control of visual scanning and response selection.

Although we did not restrict our analyses to patients with subcortical stroke, the fact that a high prevalence of PS deficit was found in a sample in which at least 66.7% of patients had lesions located in subcortical areas may reinforce the key role of subcortical WM in PS. This corroborates with a recent imaging study by Duering and colleagues (2011) showing that subcortical WM hyperintensities were associated with slowed processing in patients with cerebral autosomal arteriopathy with subcortical infarcts and leukoencephalopathy. Taken together, PS deficit is a salient feature of various strokes (Ballard et al., 2003; Brand et al., 2014), which may be partially accounted for by the WM defects in cortical–subcortical circuits that have been reported to be relatively common in this heterogeneous disease (Mäntylä et al., 1999; Zhu et al., 2012).

Another possible underlying neural mechanism for PS disturbances pertains to the abnormalities in multiple neurotransmitter systems. Several lines of evidence demonstrate that PS is modulated by neurotransmitters such as glutamate, dopamine, and gamma-aminobutyric acid in normal and neuropsychiatric disease states (Dickinson & Gold, 2007). However, research on stroke in this respect is scarce and awaits further investigation.

Corresponding with the above neurobiological underpinnings of PS, Salthouse (1996) has postulated that performance in many complex cognitive tasks relies on the execution speed of a limited set of elementary cognitive processes with associated neural substrates. Individuals with age or disease-related changes in these neural substrates often have slower processing rates, consequently affecting a wide variety of abilities. According to Salthouse, there are two factors that may contribute to the relation between speed and cognition: limited time and simultaneity. In the context of limited time mechanism, it is suggested that as we age, we may not be able to complete all pertinent mental operations in the time allotted, thereby restricting time available for later operations. Simultaneity mechanism contends that a slower speed of processing may lead to decay or displacement of earlier presented information such that relevant information may no longer be available for integrative higher-level processing. The fact that the oral SDMT was moderately correlated with all but naming and receptive speech in our stroke sample lent support to these propositions that the impact of slow PS is systemic, which manifests itself on most tasks including those that are not overtly speeded (Dickinson & Gold, 2007). Collectively, slower speed of processing, a universal phenomenon common to the healthy elderly and patients with different brain damage etiologies, can be attributed to both cognitive and neurobiological mechanisms (Brønnick, 2010).

It is worth noting that the SDMT had significant correlation with stroke severity, which is in line with prior studies showing that PS is sensitive to symptom severity associated with schizophrenia (Leeson et al., 2010), traumatic brain injury (Thatcher et al., 2001), and multiple sclerosis (Knowles, Weiser, Davidson, David, & Reichenberg, 2010). Because of its superiority in correlating with underlying brain pathology (e.g., cerebral atrophy) of multiple sclerosis (Rao et al., 2014), the SDMT was selected as the sole measure of cognition to be included in all research for this disease founded by the National Institute of Neurological Disorders and Stroke (NINDS Common Data Elements). The NINDS also recommended use of the SDMT for measuring PS in all research for stroke and TBI. Together, these results provide convincing evidence on the high susceptibility of PS to the deleterious effects of brain insult. Further studies are necessary in order to investigate the sensitivity of the SDMT to other neuropathological changes such as microinfarcts and microbleeds which have been identified as important correlates of slowed PS and dementia (Smith, Schneider, Wardlaw, & Greenberg, 2012).

So far, there is no gold standard for the assessment of PS. The problem is compounded by the multifactorial nature of most of the existing PS measures, each of which may also be tapping deficiencies in other cognitive areas such as attention and working memory. This may result in the use of different tests within the same cognitive domain or vice versa which often leads to a poorly integrated literature on PS. Cepeda, Blackwell, and Munakata (2013) posit that the choice of PS measure (complex perceptual speed versus simpler psychomotor speed) could strongly influence the types of conclusions that can be drawn. The implications for future research are a further delineation of the role the different PS measures fulfill in the brain pathology of stroke, their impact on patients' functional outcomes, and their relationships to higher cognitive functions.

The strengths of our study are the rather comprehensive assessment of neuropsychological functions and the well-matched demographic characteristics (age and education) of participants included in the patient and control groups. Several limitations need to be acknowledged. First, our patients were on average younger than subjects of other studies and did not include those with aphasia to minimize bias due to oral administration of the SDMT. Thus, our sample is not fully representative. This underscores the need for the development of PS tests with minimal motor and verbal response demands that are applicable to stroke patients who experience problems in fine motor skills or verbal communication. However, given the favorable results, the reported findings should be verified and extended in a larger sample of older stroke patients and those with SAH, or a homogeneous sample of subcortical strokes. Second, the failure to take into account premorbid intelligence may introduce biases such as its confounding effect on PS as well as under- or overestimation of a patient's level of cognitive decline (Griffin, Mindt, Rankin, Ritchie, & Scott, 2002). Third, although useful as a screening tool for dementia, the MMSE is sensitive to the brain atrophy (Apostolova et al., 2006; Fjell, Amlien, Westlye, & Walhovd, 2009), but not to the presence of silent brain infarcts and WM lesions (Swirsky-Sacchetti et al., 1992; Vermeer et al., 2003). Accordingly, the use of the MMSE may not be able to exclude the confounding contribution of these insidious neurodegenerative changes to cognitive function in older healthy individuals. Fourth, the scope of the present study does not permit a comprehensive analysis of the cognitive demands of the SDMT. It highlights the need to decompose the basic processes contributing to performance on the SDMT in stroke. Fifth, the small and unequal subsample sizes in our own work preclude an in-depth examination of the relationship between types of stroke, BG lesions, hypertension, and PS, although some of these relationships have been supported by other studies with more balanced samples in this regard (Batista et al., 2012; Gorelick et al., 2011). A final caveat is the use of a single measure of PS in this study, in spite of the fact that PS is a multidimensional construct in itself (Knowles et al., 2010). Combining several measures of PS to form a composite is strongly recommended for future studies to improve the reliability and validity of the assessment. In addition, considering that no other speeded tests were included in the test battery, it is of interest to compare the relative contributions of the oral SDMT and the other verbal measure of PS (e.g., verbal fluency) to the cognitive functions in stroke. Despite these, the significant interplay between PS and other cognitive functions is unlikely to have occurred by chance because, as stated above, controlling for demographic and other cognitive factors failed to reduce group differences.

In conclusion, our results suggest that PS deficit is a cardinal feature of stroke that underpins post-stroke cognitive impairments. Although no causal inference from the current data can be made, these findings complement qualitative interview studies on PS and cognition in stroke. More importantly, early targeted interventions aimed at reducing PS deficits are crucial for reducing post-stroke cognitive decline.

Funding

This study was supported by the National Science Council of Taiwan under grant NSC 97-2314-B-037-046-MY3.

References

Allan
L. M.
Rowan
E. N.
Firbank
M. J.
Thomas
A. J.
Parry
S. W.
Polvikoski
T. M.
, et al.  . 
Long term incidence of dementia, predictors of mortality and pathological diagnosis in older stroke survivors
Brain
 , 
2011
, vol. 
134
 (pg. 
3716
-
3727
)
Apostolova
L.
Lu
P.
Rogers
S.
Dutton
R.
Hayashi
K.
Toga
A.
, et al.  . 
3D mapping of mini-mental state examination performance in clinical and preclinical Alzheimer disease
Alzheimer Disease and Associated Disorders
 , 
2006
, vol. 
20
 (pg. 
224
-
231
)
Ballard
C.
Stephens
S.
Kenny
R.
Kalaria
R.
Tovee
M.
O'Brien
J.
Profile of neuropsychological deficits in older stroke survivors without dementia
Dementia and Geriatric Cognitive Disorders
 , 
2003
, vol. 
16
 (pg. 
52
-
56
)
Barker-Collo
S.
Feigin
V. L.
Krishmnamurthi
R.
Differences in neuropsychological profiles of long-term intracerebral hemorrhage and subarachnoid hemorrhage survivors
International Journal of Stroke
 , 
2013
, vol. 
8
 pg. 
E14
 
Barker-Collo
S.
Feigin
V. L.
Parag
V.
Lawes
C. M.
Senior
H.
Auckland stroke outcomes study. Part 2: Cognition and functional outcomes 5 years poststroke
Neurology
 , 
2010
, vol. 
75
 (pg. 
1608
-
1616
)
Barker-Collo
S.
Starkey
N.
Lawes
C. M.
Feigin
V.
Senior
H.
Parag
V.
Neuropsychological profiles of 5-year ischemic stroke survivors by Oxfordshire stroke classification and hemisphere of lesion
Stroke
 , 
2012
, vol. 
43
 (pg. 
50
-
55
)
Batista
S.
Zivadinov
R.
Hoogs
M.
Bergsland
N.
Heininen-Brown
M.
Dwyer
M. G.
, et al.  . 
Basal ganglia, thalamus and neocortical atrophy predicting slowed cognitive processing in multiple sclerosis
Journal of Neurology
 , 
2012
, vol. 
259
 (pg. 
139
-
146
)
Brand
C.
Alber
B.
Fladung
A. K.
Knauer
K.
König
R.
Oechsner
A.
, et al.  . 
Cognitive performance following spontaneous subarachnoid haemorrhage versus other forms of intracranial haemorrhage
British Journal of Neurosurgery
 , 
2014
, vol. 
28
 (pg. 
68
-
80
)
Brønnick
K.
Emre
M.
Cognitive profile in Parkinson's disease dementia
Cognitive impairment and dementia in Parkinsons disease
 , 
2010
New York
Oxford University Press
(pg. 
27
-
44
)
Bugg
J. M.
Zook
N. A.
DeLosh
E. L.
Davalos
D. B.
Davis
H. P.
Age differences in fluid intelligence: contributions of general slowing and frontal decline
Brain and Cognition
 , 
2006
, vol. 
62
 (pg. 
9
-
16
)
Butters
M. A.
Whyte
E. M.
Nebes
R. D.
Begley
A. E.
Dew
M. A.
Mulsant
B. H.
, et al.  . 
The nature and determinants of neuropsychological functioning in late-life depression
Archives of General Psychiatry
 , 
2004
, vol. 
61
 (pg. 
587
-
595
)
Cepeda
N. J.
Blackwell
K. A.
Munakata
Y.
Speed isn't everything: Complex processing speed measures mask individual differences and developmental changes in executive control
Developmental Science
 , 
2013
, vol. 
16
 (pg. 
269
-
286
)
Cohen
J.
Statistical power analysis for the behavioral sciences
 , 
1988
2nd ed
Hillsdale, NJ
Lawrence Erlbaum Associates
Cumming
T. B.
Marshall
R. S.
Lazar
R. M.
Stroke, cognitive deficits, and rehabilitation: Still an incomplete picture
International Journal of Stroke
 , 
2013
, vol. 
8
 (pg. 
38
-
45
)
Damasio
A. R.
Graff-Radford
N. R.
Eslinger
P. J.
Damasio
H.
Kassell
N.
Amnesia following basal forebrain lesions
Archives of Neurology
 , 
1985
, vol. 
42
 (pg. 
263
-
271
)
Diamond
B. J.
Johnson
S. K.
Kaufman
M.
Graves
L.
Relationships between information processing, depression, fatigue and cognition in multiple sclerosis
Archives of Clinical Neuropsychology
 , 
2008
, vol. 
23
 (pg. 
189
-
199
)
Dickinson
D.
Gold
J. M.
Deluca
J.
Kalmar
J. H.
Processing speed and the Digit Symbol Substitution test in schizophrenia
Information processing speed in clinical populations
 , 
2007
New York
Taylor and Francis
(pg. 
125
-
152
)
Douiri
A.
Rudd
A. G.
Wolfe
C. D.
Prevalence of poststroke cognitive impairment: South London Stroke Register 1995-2010
Stroke
 , 
2013
, vol. 
44
 (pg. 
138
-
145
)
Duering
M.
Zieren
N.
Hervé
D.
Jouvent
E.
Reyes
S.
Peters
N.
, et al.  . 
Strategic role of frontal white matter tracts in vascular cognitive impairment: A voxel-based lesion-symptom mapping study in CADASIL
Brain
 , 
2011
, vol. 
134
 (pg. 
2366
-
2375
)
Elgamal
S. A.
Roy
E. A.
Sharratt
M. T.
Age and verbal fluency: The mediating effect of speed of processing
Canadian Geriatrics Journal
 , 
2011
, vol. 
14
 (pg. 
66
-
72
)
Ferrer
E.
Whitaker
K. J.
Steele
J. S.
Green
C. T.
Wendelken
C.
Bunge
S. A.
White matter maturation supports the development of reasoning ability through its influence on processing speed
Developmental Science
 , 
2013
, vol. 
16
 (pg. 
941
-
951
)
Fjell
A. M.
Amlien
I. K.
Westlye
L. T.
Walhovd
K. B.
Mini-mental state examination is sensitive to brain atrophy in Alzheimer's disease
Dementia and Geriatric Cognitive Disorders
 , 
2009
, vol. 
28
 (pg. 
252
-
258
)
Folstein
M. F.
Folstein
S. E.
McHugh
P. R.
“Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician
Journal of Psychiatric Research
 , 
1975
, vol. 
12
 (pg. 
189
-
198
)
Gerritsen
M. J.
Berg
I. J.
Deelman
B. G.
Visser-Keizer
A. C.
Meyboom-de Jong
B.
Speed of information processing after unilateral stroke
Journal of Clinical and Experimental Neuropsychology
 , 
2003
, vol. 
25
 (pg. 
1
-
13
)
Gillespie
D. C.
Bowen
A.
Foster
J. K.
Memory impairment following right hemisphere stroke: A comparative meta-analytic and narrative review
The Clinical Neuropsychologist
 , 
2006
, vol. 
20
 (pg. 
59
-
75
)
Golden
C. J.
Purisch
A. D.
Hammeke
T. A.
Manual for the Luria-Nebraska neuropsychological battery: Forms I and II
 , 
1985
Los Angeles
Western Psychological Services
Gorelick
P. B.
Scuteri
A.
Black
S. E.
Decarli
C.
Greenberg
S. M.
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
, vol. 
42
 (pg. 
2672
-
2713
)
Gottesman
R. F.
Hillis
A. E.
Predictors and assessment of cognitive dysfunction resulting from ischaemic stroke
Lancet Neurology
 , 
2010
, vol. 
9
 (pg. 
895
-
905
)
Griffin
S. L.
Mindt
M. R.
Rankin
E. J.
Ritchie
A. J.
Scott
J. G.
Estimating premorbid intelligence: Comparison of traditional and contemporary methods across the intelligence continuum
Archive of Clinical Neuropsychology
 , 
2002
, vol. 
17
 (pg. 
497
-
507
)
Hannay
H. J.
Howieson
D. B.
Loring
D. W.
Fischer
J. S.
Lezak
M. D.
Lezak
M. D.
Howieson
D. B.
Loring
D. W.
Hannay
H. J.
Fischer
J. S.
Neuropathology for neuropsychologists
Neuropsychological assessment (4th ed.)
 , 
2004
New York
Oxford University Press
Hochstenbach
J.
Mulder
T.
van Limbeek
J.
Donders
R.
Schoonderwaldt
H.
Cognitive decline following stroke: A comprehensive study of cognitive decline following stroke
Journal of Clinical and Experimental Neuropsychology
 , 
1998
, vol. 
20
 (pg. 
503
-
517
)
Hurford
R.
Charidimou
A.
Fox
Z.
Cipolotti
L.
Werring
D. J.
Domain-specific trends in cognitive impairment after acute ischemic stroke
Journal of Neurology
 , 
2013
, vol. 
260
 (pg. 
237
-
241
)
Jaillard
A.
Grand
S.
Le Bas
J. F.
Hommel
M.
Predicting cognitive dysfunctioning in nondemented patients early stroke
Cerebrovascular Diseases
 , 
2010
, vol. 
29
 (pg. 
415
-
423
)
Jokinen
H.
Kalska
H.
Mäntylä
R.
Pohjasvaara
T.
Ylikoski
R.
Hietanen
M.
, et al.  . 
Cognitive profile of subcortical ischaemic vascular disease
Journal of Neurology, Neurosurgery, and Psychiatry
 , 
2006
, vol. 
77
 (pg. 
28
-
33
)
Joy
S.
Kaplan
E.
Fein
D.
Speed and memory in the WAIS-III Digit Symbol – Coding subtest across the adult lifespan
Archives of Clinical Neuropsychology
 , 
2004
, vol. 
19
 (pg. 
759
-
767
)
Kail
R.
Salthouse
T. A.
Processing speed as a mental capacity
Acta Psychologica
 , 
1994
, vol. 
86
 (pg. 
199
-
225
)
Kaplan
E.
Goodglass
H.
Weintraub
S.
Boston naming test
 , 
2001
2nd ed
Austin, TX
Pro-Ed
Kauhanen
M.
Korpelainen
J. T.
Hiltunen
P.
Brusin
E.
Mononen
H.
Määttä
R.
, et al.  . 
Poststroke depression correlates with cognitive impairment and neurological deficits
Stroke
 , 
1999
, vol. 
30
 (pg. 
1875
-
1880
)
Kilander
L.
Nyman
H.
Boberg
M.
Hansson
L.
Lithell
H.
Hypertension is related to cognitive impairment: A 20-year follow-up of 999 men
Hypertension
 , 
1998
, vol. 
31
 (pg. 
780
-
786
)
Knowles
E. E. M.
Weiser
M.
Davidson
M.
David
A. S.
Reichenberg
A.
The multidimensional structure of processing speed in healthy individuals and schizophrenia patients
Schizophrenia Research
 , 
2010
, vol. 
117
 pg. 
324
 
Kongs
S. K.
Thompson
L. L.
Iverson
G. L.
Heaton
R. K.
Wisconsin card sorting test-64 card version. Professional manual
 , 
2000
Odessa, FL
Psychological Assessment Resources
Lee
T.
Crawford
J. D.
Henry
J. D.
Trollor
J. N.
Kochan
N. A.
Wright
M. J.
, et al.  . 
Mediating effects of processing speed and executive functions in age-related differences in episodic memory performance: A cross-validation study
Neuropsychology
 , 
2012
, vol. 
26
 (pg. 
776
-
784
)
Leeson
V. C.
Barnes
T. R.
Harrison
M.
Matheson
E.
Harrison
I.
Mutsatsa
S. H.
, et al.  . 
The relationship between IQ, memory, executive function, and processing speed in recent-onset psychosis: 1-year stability and clinical outcome
Schizophrenia Bulletin
 , 
2010
, vol. 
36
 (pg. 
400
-
409
)
Leśniak
M.
Bak
T.
Czepiel
W.
Seniów
J.
Członkowska
A.
Frequency and prognostic value of cognitive disorders in stroke patients
Dementia and Geriatric Cognitive Disorders
 , 
2008
, vol. 
26
 (pg. 
356
-
363
)
Lezak
M. D.
Howieson
D. B.
Lowing
D. W.
Neuropsychological assessment
 , 
2004
New York
Oxford University Press
Lovell
M. R.
Smith
S. S.
Nussbaum
P. D.
Neuropsychological evaluation of subcortical dementia
Handbook of neuropsychology and aging
 , 
1997
New York
Plenum Press
Lyden
P. D.
Lu
M.
Levine
S.
Brott
T. G.
Broderick
J.
NINDS rtPA Stroke Study Group
A modified National Institutes of Health Stroke Scale for use in stroke clinical trials: Preliminary reliability and validity
Stroke
 , 
2001
, vol. 
32
 (pg. 
1310
-
1317
)
Madureira
S.
Guerreiro
M.
Ferro
J. M.
Dementia and cognitive impairment three months after stroke
European Journal of Neurology
 , 
2001
, vol. 
8
 (pg. 
621
-
627
)
Mäntylä
R.
Aronen
H. J.
Salonen
O.
Pohjasvaara
T.
Korpelainen
M.
Peltonen
T.
, et al.  . 
Magnetic resonance imaging white matter hyperintensities and mechanism of ischemic stroke
Stroke
 , 
1999
, vol. 
30
 (pg. 
2053
-
2058
)
McGrath
L. M.
Pennington
B. F.
Olson
R. K.
Santerre-Lemmon
L. E.
Barnard
H. D.
Willcutt
E. G.
, et al.  . 
A multiple deficit model of reading disability and attention-deficit/hyperactivity disorder: Searching for shared cognitive deficits
Journal of Child Psychology and Psychiatry
 , 
2011
, vol. 
52
 (pg. 
547
-
557
)
McKinlay
A.
Dalrymple-Alford
J. C.
Grace
R. C.
Roger
D.
The effect of attentional set-shifting, working memory, and processing speed on pragmatic language functioning in Parkinson's disease
European Journal of Cognitive Psychology
 , 
2009
, vol. 
21
 (pg. 
330
-
346
)
Meyers
J. E.
Meyers
K. R.
Rey complex figure test and recognition trial
 , 
1995
Odessa, FL
Psychological Assessment Resources
Mok
V. C.
Wong
A.
Lam
W. W.
Fan
Y. H.
Tang
W. K.
Kwok
T.
, et al.  . 
Cognitive impairment and functional outcome after stroke associated with small vessel disease
Journal of Neurology, Neurosurgery, and Psychiatry
 , 
2004
, vol. 
75
 (pg. 
560
-
566
)
Moroney
J. T.
Bagiella
E.
Tatemichi
T. K.
Paik
M. C.
Stern
Y.
Desmond
D. W.
Dementia after stroke increases the risk of long-term stroke recurrence
Neurology
 , 
1997
, vol. 
48
 (pg. 
1317
-
1825
)
Narasimhalu
K.
Ang
S.
De Silva
D. A.
Wong
M. C.
Chang
H. M.
Chia
K. S.
, et al.  . 
The prognostic effects of poststroke cognitive impairment no dementia and domain-specific cognitive impairments in nondisabled ischemic stroke patients
Stroke
 , 
2011
, vol. 
42
 (pg. 
883
-
888
)
NINDS Common Data Elements
n.d.
 
Multiple sclerosis. From http://www.commondataelements.ninds.nih.gov/ms.aspx#tab=Data_Standards (Accessed on 15 August 2014)
Nys
G. M.
van Zandvoort
M. J.
de Kort
P. L.
Jansen
B. P.
de Haan
E. H.
Kappelle
L. J.
Cognitive disorders in acute stroke: Prevalence and clinical determinants
Cerebrovascular Diseases
 , 
2007
, vol. 
23
 (pg. 
408
-
416
)
Penke
L.
Maniega
S. M.
Murray
C.
Gow
A. J.
Valdés Hernández
M. C.
Clayden
J. D.
, et al.  . 
A general factor of brain white matter integrity predicts information processing speed in healthy older people
Journal of Neuroscience
 , 
2010
, vol. 
30
 (pg. 
7569
-
7574
)
Rao
S. M.
Martin
A. L.
Huelin
R.
Wissinger
E.
Khankhel
Z.
Kim
E.
Correlations between MRI and information processing speed in MS: A meta-analysis
Multiple Sclerosis International
 , 
2014
, vol. 
2014
 pg. 
975803
 
Rasquin
S.
Lodder
J.
Verhey
F.
The association between psychiatric and cognitive symptoms after stroke: A prospective study
Cerebrovascular Diseases
 , 
2005
, vol. 
19
 (pg. 
309
-
316
)
Ravnik
J.
Starovasnik
B.
Sesok
S.
Pirtosek
Z.
Svigelj
V.
Bunc
G.
, et al.  . 
Long-term cognitive deficits in patients with good outcomes after aneurysmal subarachnoid hemorrhage from anterior communicating artery
Croatian Medical Journal
 , 
2006
, vol. 
47
 (pg. 
253
-
263
)
Reed
B. R.
Mungas
D. M.
Kramer
J. H.
Ellis
W.
Vinters
H. V.
Zarow
C.
, et al.  . 
Profiles of neuropsychological impairment in autopsy-defined Alzheimer's disease and cerebrovascular disease
Brain
 , 
2007
, vol. 
130
 
Pt 3
(pg. 
731
-
739
)
Rodríguez-Sánchez
J. M.
Crespo-Facorro
B.
González-Blanch
C.
Perez-Iglesias
R.
Vázquez-Barquero
J. L.
PAFIP Group Study
Cognitive dysfunction in first-episode psychosis: The processing speed hypothesis
British Journal of Psychiatry Supplement
 , 
2007
, vol. 
51
 (pg. 
s107
-
s110
)
Rypma
B.
Berger
J. S.
Prabhakaran
V.
Bly
B. M.
Kimberg
D. Y.
Biswal
B. B.
, et al.  . 
Neural correlates of cognitive efficiency
Neuroimage
 , 
2006
, vol. 
33
 (pg. 
969
-
979
)
Salthouse
T. A.
The processing speed theory of adult age differences in cognition
Psychological Review
 , 
1996
, vol. 
103
 (pg. 
403
-
428
)
Sheikh
J. I.
Yesavage
J. A.
Brink
T. L.
Geriatric Depression Scale (GDS) recent evidence and development of a shorter version
Clinical gerontology: A guide to assessment and intervention
 , 
1986
New York
The Haworth Press
(pg. 
165
-
173
)
Smith
A.
Symbol digit modalities test
 , 
1982
Los Angeles
Western Psychological Services
Smith
E. E.
Schneider
J. A.
Wardlaw
J. M.
Greenberg
S. M.
Cerebral microinfarcts: The invisible lesions
Lancet Neurology
 , 
2012
, vol. 
11
 (pg. 
272
-
282
)
Swirsky-Sacchetti
T.
Field
H. L.
Mitchell
D. R.
Seward
J.
Lublin
F. D.
Knobler
R. L.
, et al.  . 
The sensitivity of the mini-mental state exam in the white matter dementia of multiple sclerosis
Journal of Clinical Psychology
 , 
1992
, vol. 
48
 (pg. 
779
-
786
)
Tatemichi
T. K.
Desmond
D. W.
Paik
M.
Figueroa
M.
Gropen
T. I.
Stern
Y.
, et al.  . 
Determinants of dementia related to stroke
Annals of Neurology
 , 
1993
, vol. 
33
 (pg. 
568
-
575
)
Tatemichi
T. K.
Desmond
D. W.
Stern
Y.
Paik
M.
Sano
M.
Bagiella
E.
Cognitive impairment after stroke: Frequency, patterns, and relationship to functional abilities
Journal of Neurology, Neurosurgery, and Psychiatry
 , 
1994
, vol. 
57
 (pg. 
202
-
207
)
Thatcher
R. W.
North
D. M.
Curtin
R. T.
Walker
R. A.
Biver
C. J.
Gomez
J. F.
, et al.  . 
An EEG severity index of traumatic brain injury
Journal of Neuropsychiatry and Clinical Neurosciences
 , 
2001
, vol. 
13
 (pg. 
77
-
87
)
Turken
A.
Whitfield-Gabrieli
S.
Bammer
R.
Baldo
J. V.
Dronkers
N. F.
Gabrieli
J. D.
Cognitive processing speed and the structure of white matter pathways: Convergent evidence from normal variation and lesion studies
Neuroimage
 , 
2008
, vol. 
42
 (pg. 
1032
-
1044
)
van der Ham
I. J.
van Zandvoort
M. J.
Frijns
C. J.
Kappelle
L. J.
Postma
A.
Hemispheric differences in spatial relation processing in a scene perception task: A neuropsychological study
Neuropsychologia
 , 
2011
, vol. 
49
 (pg. 
999
-
1005
)
Vermeer
S. E.
Prins
N. D.
den Heijer
T.
Hofman
A.
Koudstaal
P. J.
Breteler
M. M.
Silent brain infarcts and white matter lesions increase stroke risk in the general population: The Rotterdam Scan Study
Stroke
 , 
2003
, vol. 
34
 (pg. 
1126
-
1129
)
Vernon
P. A.
Kantor
L.
Reaction time correlations with intelligence test scores obtained under either timed or untimed conditions
Intelligence
 , 
1986
, vol. 
10
 (pg. 
315
-
330
)
Vernooij
M. W.
Ikram
M. A.
Vrooman
H. A.
Wielopolski
P. A.
Krestin
G. P.
Hofman
A.
, et al.  . 
White matter microstructural integrity and cognitive function in a general elderly population
Archives of General Psychiatry
 , 
2009
, vol. 
66
 (pg. 
545
-
553
)
Wechsler
D.
Wechsler memory scale-third edition
 , 
1997
San Antonio, FL
The Psychological Corporation
Western Psychological Services
Hooper visual organization test (VOT) manual
 , 
1983
Los Angeles
Author
WHO MONICA Project Principal Investigators
The World Health Organization MONICA Project (Monitoring Trends and Determinants in Cardiovascular Disease): A major international collaboration
Journal of Clinical Epidemiology
 , 
1988
, vol. 
41
 (pg. 
105
-
114
)
Willmott
C.
Ponsford
J.
Hocking
C.
Schönberger
M.
Factors contributing to attentional impairments after traumatic brain injury
Neuropsychology
 , 
2009
, vol. 
23
 (pg. 
424
-
432
)
Winkens
I.
Van Heugten
C. M.
Fasotti
L.
Duits
A. A.
Wade
D. T.
Manifestations of mental slowness in the daily life of patients with stroke: A qualitative study
Clinical Rehabilitation
 , 
2006
, vol. 
20
 (pg. 
827
-
834
)
Zhu
Y. C.
Chabriat
H.
Godin
O.
Dufouil
C.
Rosand
J.
Greenberg
S. M.
, et al.  . 
Distribution of white matter hyperintensity in cerebral hemorrhage and healthy aging
Journal of Neurology
 , 
2012
, vol. 
259
 (pg. 
530
-
536
)
Zinn
S.
Bosworth
H. B.
Hoenig
H. M.
Swartzwelder
H. S.
Executive function deficits in acute stroke
Archives of Physical Medicine and Rehabilitation
 , 
2007
, vol. 
88
 (pg. 
173
-
180
)