| | Executive Function Deficits in Acute StrokeAbstract Zinn S, Bosworth HB, Hoenig HM, Swartzwelder HS. Executive function deficits in acute stroke. ObjectivesTo establish the frequency of executive dysfunction during acute hospitalization for stroke and to examine the relationship of that dysfunction to stroke severity and premorbid characteristics. DesignInception cohort study. SettingInpatient wards at a Veterans Affairs hospital. ParticipantsConsecutive sample of inpatients with radiologically or neurologically confirmed stroke. Final sample included 47 patients screened for aphasia and capable of neuropsychologic testing. Two nonstroke inpatient control samples (n=10 each) with either transient ischemic attack (TIA) or multiple stroke risk factors were administered the same research procedure and tests. InterventionsNot applicable. Main Outcome MeasuresComposite cognitive impairment ratio (CIR), calculated from 8 scores indicative of executive function on 6 neuropsychologic tests by dividing number of tests completed into the number of scores falling below cutoff point, defined as 1.5 standard deviations below normative population mean. ResultsStroke patients had a mean CIR of .61, compared with .48 for TIAs and .44 for stroke-risk-only. Analysis of variance revealed that CIRs of stroke-risk-only patients but not TIAs were lower than those of the stroke patients (P=.02). Impairment frequencies were at least 50% for stroke patients on most test scores. The Symbol Digit Modalities Test (75% impairment) and a design fluency measure distinguished stroke from nonstroke patients. CIR was not related to stroke severity in the stroke patient sample, but was related to estimated premorbid intelligence. ConclusionsExecutive function deficits are common in stroke patients. The data suggest that limitations in information processing due to these deficits may require environmental and procedural accommodations to increase rehabilitation benefit. PATIENTS WITH RECENT STROKE who are beginning rehabilitation often have cognitive impairment, either predating or acquired with the stroke. Deficits in the particular cognitive processes known as executive functions, which manage goal-oriented behavior, are common poststroke1, 2 and reduce stroke treatment effectiveness.3, 4 Current knowledge of executive function deficits is typically based on studies conducted 3 months or more poststroke. Prevalence of executive function deficits may be even higher early after stroke, when the initial rehabilitative training occurs, but this is unknown. No neuropsychologic studies of executive functions in acute stroke have been conducted, to our knowledge. Although executive functions may broadly be said to manage goal-oriented behavior, several component processes have been identified. Important components of the executive functions include starting and stopping behavior at appropriate times, persisting at a task or switching tactics as needed, and selecting behaviors in novel situations based on context and higher-level or long-term goals. These components are also denoted (respectively) as initiation/perseveration, cognitive persistence and flexibility, self-monitoring, and abstract thinking (including planning). Working memory, the capacity for storing and manipulating data during problem solving, is dependent on frontal cortex (and other brain regions) and is often included as an executive function. Combined deficits of these components create impairments that can compromise rehabilitation treatment in varying ways. For example, rehabilitation patients with deficits in initiation and persistence may have a reduced capacity to initiate, sequence, and sustain a series of exercises, due to their executive function impairment, and thus have reduced functional recovery after stroke.3 Deficits of initiation and perseveration may also result in an impairment, producing compulsive repetition of a behavioral sequence.5 Impairments of planning and/or problem-solving can lead to unsafe physical maneuvers and increase the risk of falls.6, 7 When there is a generalized deficit of attention and/or cognitive speed in addition to executive function deficits, the ability to process novel or complex information is curtailed, leading to limits in information processing capacity.8 Impaired planning, reduced prospective memory (remembering to remember something), and reduced information-processing capacity may make it difficult for rehabilitation patients to remember and follow the complex treatment regimens, often provided at discharge, that are designed to promote functional gains and reduce their risk of stroke recurrence. Post-rehabilitation functional improvement has been related to executive function scores9 and providing cognitive remediation has improved performance of activities of daily living (ADLs) in stroke patients.10 Poststroke cognitive impairment of any type has been repeatedly related to stroke severity,11, 12 but recent studies suggest that executive function decline may begin prior to completed strokes.13, 14, 15, 16, 17 It appears that small vessel ischemic disease in white and subcortical gray matter leads to decrements in executive functioning18, 19, 20 that primarily affect processing speed21 and cognitive flexibility.15 Functional decline, especially in instrumental ADLs, has been related to executive function impairment even in community-dwelling samples.22, 23, 24, 25 Thus any executive function deficit occurring as a sequelae of stroke may be amplified by prestroke executive function decline. We believe that stroke outcomes can be improved if executive function deficits are identified and compensatory techniques are incorporated into treatment early. The study reported here was designed to establish the frequency and correlates of executive function deficits occurring during acute hospitalization for stroke. For this study, acute hospitalization was defined as the initial inpatient period (typically 3−7 days at our facility) during which symptoms were stabilized, diagnostic studies were completed, and rehabilitation was initiated. Identifying executive function deficits in acute stroke patients through neuropsychologic assessment is also needed because these deficits are more subtle than aphasia or neglect, and health care providers working with stroke patients have often been taught little about this area of cognitive functioning. Furthermore, identification of deficits is difficult because executive functions are multifaceted, and are most greatly activated in novel or unstructured situations. Most studies of any type of cognitive impairment in acute stroke involve brief screening tests such as the Mini-Mental Status Examination (MMSE),26, 27 yet the MMSE is relatively insensitive to executive function deficits.28 Studies using more in-depth testing, thereby including some executive function measures, are typically conducted at 3 months2 or more poststroke.29 The only study we found that conducted neuropsychologic tests earlier than 3 months performed assessment at 1 month, but included only a single letter sorting task that might be considered an executive function measure.30 In general, executive function assessment in research studies of stroke has typically been conducted using 1 or 2 tests. If these are reported individually, often only the mean and standard deviation is presented. Rates of impairment are rarely reported, so it is difficult to determine whether a highly impaired minority or a broad reduction in scores are responsible for lowered means. Furthermore, the multiple aspects of executive functioning are not well captured by 1 or 2 test scores, especially in a stroke population who have broad variability among their deficits. Therefore, we administered a battery of executive function tests to patients recently admitted for workup of acute stroke. We wanted to determine the rates of executive function impairment and to examine their relationship with the severity of the index stroke, prestroke functional decline, and patient characteristics such as age and intelligence. We examined both acute symptoms (stroke severity) and premorbid functioning as correlates of executive function deficits, and examined patients without stroke who may evidence this decline. Age, education, and intelligence are common correlates of many aspects of cognitive functioning, so they were examined as well. We predicted a strong relationship between rates of impairment and stroke severity. To better appreciate the context of our expected impairment rates and gain some understanding of possible prestroke decline and other factors, we administered the same neuropsychologic battery to small samples of patients with either transient ischemic attack (TIA) or stroke risk factors alone, who had not had a stroke. We hypothesized, however, that the stroke patients overall would have higher rates of executive function impairment than vascular disease patients without stroke. Methods  Design and Sample We conducted a prospective cohort study of consecutive patients presenting to a Veterans Affairs medical center in the southeastern United States over a 2.5-year period with symptoms of stroke. All patients were evaluated by an admitting neurologist. We exluded patients with global aphasia, dementia or psychosis, or who were unconscious or otherwise too impaired to sign their own consent (when no proxy was available), or whose strokes had occurred more than 10 days prior to enrollment or were related to other brain disease. During screening for this study, patients were interviewed and screened for aphasia or confusion severe enough to hinder test administration, using the sequential commands and auditory verbal comprehension subtests from the Western Aphasia Battery.31, 32 Stroke was confirmed by diffusion-weighted magnetic resonance imaging (MRI), or in several cases where MRI was precluded or inconclusive, from clinical examination, chart review and computed tomography scan by a neurologist. Recruitment of the TIA sample was similar. Stroke-risk-only patients were recruited from the inpatient general medicine wards over a 2-month period; the screening and chart review process was the same. Both the stroke and risk factor protocols were approved by the medical center’s institutional review board and all patients gave informed consent. Patient Enrollment Of 325 patients admitted with symptoms suggestive of stroke, 139 were evaluated by the admitting neurologist as having a diagnosis other than stroke. Of the 186 probable stroke patients, 83 were enrolled and 103 were not enrolled for the reasons noted in figure 1. Of the 83 enrolled, 23 patients were subsequently ruled out for stroke, which left 60 with confirmed stroke, and of these, 47 completed at least part of the neuropsychologic battery. Although our original target enrollment was 100 stroke patients, funding and administrative considerations prevented the extension of our enrollment period when recruitment rates fell below our predictions. Fifteen of the patients ruled out for acute stroke had a TIA; those who had no prior stroke became our TIA sample. The other 8 patients, who had several stroke risk factors and represented less severe cerebrovascular disease, were retained if tested and free of stroke as a convenience sample (n=4) to allow comparison of their executive functioning with that of patients with acute stroke. We supplemented their numbers with additional inpatients who had a minimum of 3 stroke risk factors to create control samples of 10 each of TIA and stroke-risk-only patients. One TIA patient, however, had several severely impaired test scores that were outliers, suggestive of a “silent” stroke, and was subsequently excluded from the analyses. Thus the final sample used in the analyses included 47 stroke patients, 9 TIA patients and 10 patients with at least 3 stroke risk factors. Data Collection At the time of enrollment, demographic, health history, mood status, and pre-admission functional ability information were obtained by patient self-report during the screening interview. Next, a battery of neuropsychologic tests was administered by the primary investigator or a research assistant trained in standard administration of the instruments. Neuropsychologic tests were administered in a fixed order using rote instructions. Because these assessments were collected during the acute stay, completion of testing was sometimes precluded by clinical treatment procedures, patient fatigue, or hospital discharge. Sequelae of stroke, such as paresis, aphasia or visual dysfunction, also precluded administration of certain instruments to particular patients. We used chart review to tabulate stroke risk factors, which included hypertension, diabetes, hyperlipidemia, cardiovascular disease, peripheral vascular diseases, migraine, smoking, cocaine or alcohol dependence, and sleep apnea. Instruments We measured stroke severity by using the National Institutes of Health Stroke Scale (NIHSS).33 These scores were determined by a trained rater from review of the admitting neurologist’s chart note. Scores ranged from 0 to 31, with higher scores indicating greater severity. Functional ability was measured using the questionnaires for physical and instrumental activities of daily living developed for the Older Americans Resources and Services (OARS) study.34, 35 Physical and instrumental ADL scales each contained 7 questions scored as 0 (fully dependent), 1 (requiring some assistance), or 2 (independent) points for the individual’s level of independence in various daily living skills (14 total points possible). We asked patients to use the timeframe of 2 weeks prior to their admission in responding. An estimation of their premorbid intelligence quotient (IQ) was derived using the Barona method from demographic information obtained during the interview.36 The validity and reliability of these instruments have been established as noted in the cited references. We selected the neuropsychologic battery to sample multiple aspects of executive functioning while minimizing the effects of paresis, apraxia, or aphasia. This was difficult because the complex behaviors that recruit executive functioning are likely to be affected by one of these sequelae. Using a variety of tasks to measure a given component can reduce this confounding effect. Because our assessments were done in an acute care environment, we avoided instruments, such as the Wisconsin Card Sorting Test, that take a long time to administer. Our battery included tests that assessed working memory, initiation/perseveration, cognitive persistence and flexibility, self-monitoring, and abstract thinking. Selection of particular tests was guided in part by clinical experience with vascular disease patients. These tests all have well-established validity and reliability. It included digit span and picture arrangement subtests from the Wechsler Adult Intelligence Scale, Third Edition,37 Symbol Digit Modalities Test (SDMT, oral version),38, 39 and the design fluency and trail making subtests from the Delis-Kaplan Executive Function System (DKEFS).40 The Hopkins Verbal Learning Test−Revised (HVLT)41 was administered to assess the effect of executive functioning on verbal memory, relevant to the learning of treatment regimens.42 In addition, differences in memory processes (ie, encoding versus retrieval) have been found to distinguish vascular cognitive impairment from that of Alzheimer’s disease.43 This difference, which may rely in part on strategy usage, appears to involve the frontostriatal circuits that form a major axis of executive functioning.44, 45, 46 Digit span requires the individual to repeat random digit strings of increasing length, and then to repeat similar strings in reverse order. Digit span forward is thought to assess basic attention and span memory for auditory stimuli, whereas digit span backward requires complex attention and working memory.47 The difference between forward and backward scores was used to capture the additional demands of working memory while accounting for the effect of age-related reductions in forward span. Picture arrangement consists of cartoon pictures depicting a brief story that are presented in scrambled order and must be reordered “to tell a story that makes sense” within a time limit. This test assesses working memory, cognitive persistence, social abstraction, and reasoning. As a timed test, it also requires sustained attention and processing speed. The SDMT is a coding task in which digits are paired with abstract symbols and the correct digit must be determined for a long array of symbols. Because the response is a digit and not a symbol, the SDMT may be conducted by having the subject call out the numbers corresponding to the symbols; this was done for all patients in our samples. Functions targeted by the SDMT include processing speed, working memory, cognitive flexibility, cognitive persistence, and incidental learning. We selected the design fluency task and a trail making test from the DKEFS battery. The DKEFS design fluency task involves drawing 4 straight lines between dot templates to create abstract designs, with the instruction to create as many different designs as possible in 1 minute. This test assesses processing speed, working memory, initiation, perseveration, cognitive flexibility, and nonverbal abstraction. The first 2 trials require the test-taker to draw lines between the black dots only or the white dots only, but the third switching trial involves alternating between black and white dots. The total production score summarizes the number of unique designs generated across trials, whereas the accuracy score reflects the proportion of scoreable designs to those that are invalid or perseverative. DKEFS trail making has both alphabetical and numerical trail trials, and scanning and speed trials, as well as a switching trial. Only the switching trial was selected for this study, on the grounds that it would most strongly reflect executive functioning,47 assessing cognitive flexibility and perseveration, as well as processing speed. The HVLT is a word list learning task with learning trials, a delayed free recall trial, and a recognition trial. The total immediate recall score (number recalled on all learning trials) reflects attentional capacity and the ability to benefit from repetition (learning). The delayed recall scores measure free recall delayed (“long-term”) memory. Statistics We examined mean differences between the groups for demographic characteristics using analysis of variance (ANOVA).a The proportion of tests impaired were examined using Fisher exact statistics. To use equivalent cutpoints for determining impaired performance, scores on the neuropsychologic tests were standardized by age-stratified norms into z scores (HVLT, SDMT) or scaled scores (picture arrangement, DKEFS subtests; all norms used were the manual norms). Standardized z scores were classified as impaired or unimpaired based on a cutpoint of 1.5 standard deviations (SDs) below the mean (a commonly used criterion; see Lopez et al48). Note that for a normally distributed population, 9% of all individuals, or about 5 of 50 people, would fall below this cutpoint. For scaled scores, this cutpoint translated into a scaled score of 5 or below. Because the norms for translating digit span scores into scaled scores confound digits forward and backward, which likely represent different cognitive operations, the raw scores of the digits-backward task only was used. To determine the cutpoint for digit span, we used the Lezak criterion47 for evaluating the impairment of backward span (a working memory index) as a guideline. The Lezak criterion states that the backward span scores are typically 2 points less than the forward span scores. We reasoned that a difference greater than 4 between the 2 scores would indicate impairment. We used these binary scores (representing impairment if below the 1.5 SD cutoff and no impairment if above) to summarize performance across tests as a composite impairment ratio (CIR). Because a large proportion of patients did not complete the full battery, using a composite impairment ratio for those tests completed enabled us to compare participants irrespective of the number of tests completed. We calculated the CIR as the percentage of tests failed divided by the number of tests completed. ANOVA was used to examine whether patients with stroke had worse CIRs than the other 2 groups, as we hypothesized. Because a large proportion of participants were unable to complete the full battery, we assessed whether there was a difference in executive function test performance between those who completed the battery and those who did not complete it, using a t test on the CIR. We also examined whether stroke severity affected battery completion by examining the relationship of severity and number of tests completed. The frequencies of scores above and below cutoff on each executive function test score were compiled for each group. We evaluated the significance of any group differences in these scores by using the Fisher exact statistic because some cell sizes were small. We also conducted Spearman correlations in the stroke sample to determine whether the CIR was associated with stroke severity (as predicted) or preadmission functionality. The NIHSS score and the 2 OARS functional ability scales (physical and instrumental ADLs) were used in these analyses. Age, years of education, and estimated premorbid IQ were relatively normally distributed, so the relationship of these variables to the CIR for the combined samples were evaluated using Pearson correlations. Results Sample Characteristics A majority of the sample was male; only 5% were women. Other demographic characteristics for stroke and nonstroke groups are shown in table 1. The groups did not differ significantly on any of the listed characteristics. These data indicate that all patients in our samples were relatively high functioning prior to their stroke, with mild loss of independence in only 1 area on average (physical ADLs, 13.0/14 possible points). For the stroke sample, stroke severity was also relatively low, with mean severity at 3.4 out of a possible 31 (SD=3.0). The mean duration between stroke and assessment was 4.6 days. The presence of stroke risk factors was consistent across samples with 3 to 4 risk factors identified on average. Test Results The amount of missing data varied for each test used. The frequency of missing scores, out of 66, listed in order of administration within the battery, was as follows: digit span, 0 missing; HVLT, 1 missing; design fluency, 11 missing; picture arrangement, 15 missing; SDMT, 21 missing; and trail making, 25 missing. Note that tests administered toward the end of the battery were less likely to be completed. There was no relationship between stroke severity and number of tests completed, nor was there any difference in the CIR between those who completed the battery and those who did not, for whatever reason. The mean proportions of tests impaired (CIR) for the samples are shown at the top of table 2. Patients with acute stroke on average were impaired in 60% of the tests completed. The ANOVA for group differences was significant (F2,63=4.03, P=.02), indicating that the proportion of impairment was higher for stroke patients than for nonstroke control patients. This was examined further with post hoc comparisons using the Tukey-Kramer adjustment, which revealed that the stroke-risk-only group alone had significantly less impairment than the stroke patient group. Mean frequencies of impairment by individual test score are also given in table 2. Note that a 10% impairment rate would be expected in the normal population; our clinical samples would be expected to have somewhat higher rates than that. About half of the stroke patients performed in the impaired range on nearly all of the test scores examined. Two exceptions were picture arrangement, which had a fairly low proportion of impairment, and SDMT, which was impaired at rather high frequencies in all groups. The SDMT, design fluency production, and trail making switching scores were significantly different between groups. These tests all assessed cognitive flexibility, working memory and processing speed. Notably, the stroke-risk-only group was more often impaired on the verbal memory indices (both immediate and delayed) than were either of the other 2 groups. For other tests, there appeared to be nearly equivalent frequencies of impairment across groups. The finding of worse verbal memory in the risk-only group was unexpected and raised concerns about that sample. We went back to the data and compared the raw scores on the memory test between the 3 groups using ANOVA, and found no significant difference. There were 2 scores in the stroke-risk-only group (20% of the sample) that were classified as impaired by virtue of being just below the cutoff point; these elevated the proportion of impaired scores. For the evaluation of acute versus premorbid sources of executive function deficit, there was no relationship of CIR to stroke severity (ρ=.29, P=.07) for patients with stroke. Self-reported prestroke functioning was not related to the CIR in patients with stroke, but neared significance in the combined samples for instrumental ADLs (ρ=−.22, P=.07). The CIR was not related to age or education in the combined sample, but was related to estimated premorbid IQ (ρ=−.24, P=.05). Discussion Rates of impairment on neuropsychologic tests assessing executive functions were close to 50% on most scores examined in acute stroke patients, but were also pronounced in patients with TIA or other stroke risk factors. Tests assessing working memory, cognitive flexibility, and processing speed were more impaired in stroke patients than in patients with risk factors only. Mean rates of executive function impairment, however, in our small TIA sample did not differ significantly from those of patients with stroke. Neither stroke severity nor prestroke functioning was related to our composite index of executive functioning, but estimated premorbid IQ was modestly related to CIR. Given the high rates of executive function deficits in the nonstroke groups, it is reasonable to consider that there may be a gradual loss of executive function processing in cerebrovascular disease prior to stroke that becomes worse with greater lesion load.15 Rao et al16 reported that executive function impairment contributed to global cognitive impairment in TIA patients. Hypertension alone is associated with executive function deficits.49 The fact that our TIA patients did not differ significantly from our stroke patients on mean rates of impairment suggests that executive function deficits exist on a continuum in cerebrovascular disease.15 As noted earlier, evidence is mounting that microvascular disease causes executive dysfunction in patients with vascular risk factors.18, 19, 20 Minute petechial strokes or white matter loss, known as microvascular disease, are associated with cognitive slowing.21, 50 It is plausible that slower processing speed is reflected as reduced executive functioning. Given that the purpose of executive functioning is to coordinate and guide goal-oriented behavior, then the cerebral regions supporting executive functioning must be interconnected with the other brain regions providing the “raw materials” for choices and behavior. Indeed, the frontal lobes are heavily interconnected with posterior regions, regions supporting memory, and subcortical regions.51 Reductions in mental processing speed due to degradation of white matter tracts connecting these regions affect executive functioning and thus performance on tests of executive function.52 Microvascular disease also affects subcortical nuclei, and subcortical vascular syndromes are strongly related to executive function deficits.19, 53 In each of these vascular syndromes, slowed processing speed may be a central feature that leads to executive dysfunction, yet frontal processing is affected54 and the behavioral manifestations are similar to those of working memory and cognitive flexibility deficits.15 If microvascular disease is responsible for a portion of the executive function measure variance, then this could explain why stroke severity was not significantly related to the CIR. It is reasonable to expect that the effects of an acute cerebrovascular event would be observed with a larger sample, more severe strokes, or a different neuropsychologic measure than our composite index. The finding of a relationship between executive function impairment and estimated premorbid intelligence has interesting implications. Studies have shown that IQ is not generally predictive of executive function performance.55 It is also hard to envision how slowed processing speed is related to premorbid intelligence. Our IQ estimation formula, however, relies on demographic data such as race, years of education, and occupation. Although this is speculative, the most likely explanation is that our IQ measure is acting as a measure of what neuropsychologists call cognitive reserve. Studies from Alzheimer’s patients have shown that the effects of deterioration in the central nervous system on functioning are mitigated by advanced education or cognitively complex occupations.56 It is thought that this sort of lifetime cognitive enrichment creates additional neuronal connections that enable normal functioning to be maintained longer despite neurologic decline.57 This suggests that in our study, the behavioral effect of executive function deficits and slowed processing speed is more pronounced in patients with less education and a history of less intellectually challenging occupations. This hypothesis would need to be supported by further investigation in larger and demographically different samples. If it persists, perhaps “prehabilitation” can be used to bolster cognitive reserve and reduce functional decline in people with cerebrovascular disease who have not had strokes. Although we did not find a strong relationship between executive dysfunction and prestroke instrumental activities functioning, other studies have suggested that a slide toward disability may be occurring prior to the stroke. Patients with subcortical strokes, associated with microvascular disease, were more likely to have had both executive function and mobility declines prior to stroke than patients with other types of strokes.58 Community-based studies of daily activities performance in independent older individuals have found an association of executive function levels and instrumental activities performance.23, 25 We suggest that this is an important area for future research because it may lead to preventive interventions. The pronounced impairment of processing speed and executive functioning has broad implications for rehabilitative treatment. Learning how to put a shirt on with a paretic arm or learning how to transfer from a wheelchair are both novel activities and require rapid simultaneous processing of motor and sensory feedback. Our results suggest that when faced with novel situations or increased processing requirements, cerebrovascular disease patients may be affected by processing limitations. Several recent studies bear this out. Reduced information processing speed has been related to FIM scores in stroke patients.59 The integration of perception, intention, and motor programs needed for movement correction under quickly changing conditions is impaired in stroke patients with executive function deficits.60 Poor fine motor control ipsilateral to the lesion may be related to generalized impairment of processing capacity.61 Upper-body dressing difficulties in cognitively impaired patients has been related to failure to learn compensatory strategies,62 and strategy generation requires executive functions. Even postural control requires more attention poststroke and would ostensibly be compromised by reduced cognitive processing capacity.7 Motor retraining techniques need to be developed that compensate for these processing deficits. A longitudinal study that examines executive functioning at both baseline and 3 to 6 months out would not only establish any change in prevalence over time but could determine any specific relationships with changes in both motor and instrumental functioning. Thus, patients with “invisible” executive function deficits are at risk for failure to benefit fully from rehabilitation,63 and not just during the acute phase. Because of their relationship to initiation and self-monitoring, executive functions would plausibly affect both adherence to and performance in rehabilitation after discharge. Unexpectedly high frequencies of executive dysfunction have been found in recovering stroke patients at 3 months who did not have clinically apparent cognitive impairment.64 Executive dysfunction measured 3 to 4 months poststroke has been associated with worse ADLs, both physical and instrumental.2 Attentional deficits are related to poor social participation after stroke.4 Developing and maintaining a new treatment regimen, such as adding new medications or monitoring one’s diet or exercise, also involves cognitive processing of the very type that we found to be impaired. The effect of executive functions on adherence to any sort of treatment plans has not been studied. Effective treatment may require the development and implementation of environmental supports65 to improve physical control and foster new health-related behaviors for patients whose individual cognitive resources are lessened. Studies examining the effect of executive functioning on treatment adherence and its effect on rehabilitation outcome are also needed. The high rates of executive function impairment are particularly noteworthy because the stroke patients in our sample were not excessively disabled. The widespread nature of the impairment suggests that executive function loss can occur irrespective of lesion location and severity. Because the executive functions are more subtle in their expression than localized deficits such as aphasia or apraxia, the extent of a stroke patient’s executive deficits may go unrecognized.64 Most of the stroke patients in our sample were alert and conversant and, aside from limb weakness, would not appear greatly impaired to most clinicians. Screening patients is probably the most efficient means of identifying patients likely to have deficits, given the lack of knowledge about these problems in acute cerebrovascular disease. One benefit of this study is that it suggests that the SDMT is sensitive to these deficits. This test is easily and quickly administered and its written version, which can be used in patients with expressive aphasia, can be easily converted to an oral version for patients with upper extremity weakness.38, 39 Study Limitations Some of the weaknesses of this study are the result of the difficulties inherent in conducting a neuropsychologic study of acute stroke. Often patients are still disoriented or severely aphasic, and are therefore not good candidates for thorough testing. Although this results in exclusion of the most severe patients, such patients are often not candidates for rehabilitation, and if they are, their cognitive difficulties are more apparent. Thus our conclusions remain valid and applicable to rehabilitation populations. Furthermore, it is difficult to assemble a battery of neuropsychologic tests of executive function for stroke patients that avoids the confounders of aphasia, apraxia, and paresis. The executive functions modulate behavior, and most behaviors have a verbal, motor, or activity component that may be affected by stroke. This was likely mitigated somewhat by the attenuated severity of the strokes in our patient sample. Furthermore, our battery was deliberately multidimensional in order to measure the various components of executive functioning through multiple tasks, making it possible to obtain at least partial measurement of executive functioning on every patient irrespective of focal symptoms. This strategy is not ideal, however. Future studies enrolling large samples that would enable these confounds to be statistically controlled are desperately needed. Access to patients is also challenging. Acute stays for stroke often last only long enough to stabilize the patient’s medical condition and conduct diagnostic tests. Researchers must compete with other providers and trips to radiology or other diagnostic laboratories for the patient’s time. The highest attrition on test administration occurred at the end of the battery and was due in part to discharge before testing could be completed. Higher patient dropout before completion may also result from the stress of hospitalization. Stroke patients in particular may have fatigued more quickly than those in the other 2 groups. Although these factors resulted in numerous missing data, several systematic sources of bias were ruled out, including those based on neurological impairments that are of central interest. The number of tests completed did not appear to be due to stroke severity or to the severity of executive function deficits. Our use of impairment frequencies and a composite impairment index was subject to distortion due to the small sample sizes. Using a cutpoint entails the risk of miscategorization and with small samples this may be magnified, as may have occurred in our risk-only group on memory tests (but see next paragraph). The larger number of missing values on tests administered later in the battery compounds the difficulty. Had more stroke patients completed the SDMT or trail making test, it is possible that that group’s CIRs might have changed enough to be significantly different from those of the TIA group. Despite these limitations, our study results point clearly to an extensive cognitive liability in patients with cerebrovascular disease. The small sample sizes may have also affected the memory scores of the stroke risk-only group, which would likely regress toward the mean with a larger sample. Nevertheless, it is possible that, although we sampled inpatients in order to be testing under similar conditions, recruiting inpatients with stroke risk factors resulted in a high proportion of patients with coronary disease and a history of coronary bypass surgery, associated with persistent verbal memory impairment.66, 67 This should be screened for in future studies of cognitive impairment in patients with stroke risk factors. This study is valuable because the inception cohort design and assessment within days of the stroke provided new prevalence information on executive functions during the period of initial rehabilitation assessment. One strength of this study is its assessment of executive function deficits early after stroke with a test battery designed to capture different aspects of executive functioning while minimizing the effects of sequelae. The presence of comparison groups that were demographically similar but with different severities of cerebrovascular disease, although small, highlighted the widespread and probably progressive nature of executive function deficits in this population. The sex composition of the sample (predominantly men) suggests caution in generalizing our findings, although many of our conclusions are supported by other studies with more balanced samples in this regard.16, 21, 64 Conclusions Reduced processing speed, working memory, and cognitive flexibility are common in both stroke and cerebrovascular disease patients. Neuropsychologic screening for such deficits, with a sensitive measure such as the SDMT, can alert providers to the presence of these subtle but crucial information processing impairments. We are developing compensatory rehabilitation techniques that allow stroke patients with executive dysfunction to obtain maximal benefits early from inpatient therapies, and encourage similar projects for outpatient and home-based treatment. The reduction of functional decline, an important target in our aging society, may be facilitated by identifying and remediating executive and physical function deterioration in individuals with cerebrovascular disease before stroke occurs. Further research is needed to achieve this goal. Supplier Acknowledgments The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs. We thank the helpful comments of Morris Weinberger, PhD, and Deborah Koltai Attix, PhD, in the preparation of this manuscript. Meri-Li Douglas was helpful in preparing the revisions. References 1. 1Ballard C, Stephens S, Kenny R, Kalaria R, Tovee M, O’Brien J. Profile of neuropsychological deficits in older stroke survivors without dementia. Dementia Geriatr Cogn Disord. 2003;16:52–56. 2. 2Pohjasvaara T, Leskela M, Vataja R, et al. Post-stroke depression, executive dysfunction and functional outcome. Eur J Neurol. 2002;9:269–275. MEDLINE |
CrossRef
3. 3Mok VC, Wong A, Lam WW, et al. Cognitive impairment and functional outcome after stroke associated with small vessel disease. J Neurol Neurosurg Psychiatry. 2004;75:560–566. MEDLINE |
CrossRef
4. 4McDowd JM, Filion DL, Pohl PS, Richards LG, Stiers W. Attentional abilities and functional outcomes following stroke. J Gerontol B Psychol Sci Soc Sci. 2003;58:P45–P53. MEDLINE 5. 5Evans JJ, Emslie H, Wilson BA. External cueing systems in the rehabilitation of executive impairments of action. J Int Neuropsychol Soc. 1998;4:399–408. MEDLINE 6. 6Hyndman D, Ashburn A. People with stroke living in the community: attention deficits, balance, ADL ability and falls. Disabil Rehabil. 2003;25:817–822. MEDLINE |
CrossRef
7. 7Brown LA, Sleik RJ, Winder TR. Attentional demands for static postural control after stroke. Arch Phys Med Rehabil. 2002;83:1732–1735. Abstract | Full Text |
Full-Text PDF (77 KB)
|
CrossRef
8. 8Torgesen JK. Issues in the assessment of executive function: an information-processing perspective. In: Lyon GR editors. Frames of reference for the assessment of learning disabilities: new views on measurement issues. Baltimore: PH Brookes; 1994;p. 143–162. 9. 9Mysiw WJ, Beegan JG, Gatens PF. Prospective cognitive assessment of stroke patients before inpatient rehabilitation (The relationship of the Neurobehavioral Cognitive Status Examination to functional improvement). Am J Phys Med Rehabil. 1989;68:168–171. MEDLINE |
CrossRef
10. 10Carter LT, Oliveira DO, Duponte J, Lynch SV. The relationship of cognitive skills performance to activities of daily living in stroke patients. Am J Occup Ther. 1988;42:449–455. MEDLINE 11. 11Desmond DW, Moroney JT, Paik MC, et al. Frequency and clinical determinants of dementia after ischemic stroke. Neurology. 2000;54:1124–1131. MEDLINE 12. 12Barba R, Martinez-Espinosa S, Rodriguez-Garcia E, Pondal M, Vivancos J, Del Ser T. Poststroke dementia: clinical features and risk factors. Stroke. 2000;31:1494–1501. MEDLINE 13. 13Pohjasvaara T, Mantyla R, Aronen HJ, et al. Clinical and radiological determinants of prestroke cognitive decline in a stroke cohort. J Neurol Neurosurg Psychiatry. 1999;67:742–748. MEDLINE |
CrossRef
14. 14Burton E, Ballard C, Stephens S, et al. Hyperintensities and fronto-subcortical atrophy on MRI are substrates of mild cognitive deficits after stroke. Dementia Geriatr Cogn Disord. 2003;16:113–118. 15. 15Sachdev PS, Brodaty H, Valenzuela MJ, et al. The neuropsychological profile of vascular cognitive impairment in stroke and TIA patients. Neurology. 2004;62:912–919. 16. 16Rao R, Jackson S, Howard R. Neuropsychological impairment in stroke, carotid stenosis, and peripheral vascular disease (A comparison with healthy community residents). Stroke. 1999;30:2167–2173. MEDLINE 17. 17Elias MF, Sullivan LM, D’Agostino RB, et al. Framingham stroke risk profile and lowered cognitive performance. Stroke. 2004;35:404–409.
CrossRef
18. 18Desmond DW. Cognition and white matter lesions. Cerebrovasc Dis. 2002;13:53–57.
CrossRef
19. 19Roman GC, Erkinjuntti T, Wallin A, Pantoni L, Chui HC. Subcortical ischaemic vascular dementia. Lancet Neurol. 2002;1:426–436. Abstract | Full Text |
Full-Text PDF (388 KB)
|
CrossRef
20. 20Werring DJ, Frazer DW, Coward LJ, et al. Cognitive dysfunction in patients with cerebral microbleeds on T2*-weighted gradient-echo MRI. Brain. 2004;127:2265–2275.
CrossRef
21. 21Prins ND, van Dijk EJ, den Heijer T, et al. Cerebral small-vessel disease and decline in information processing speed, executive function and memory. Brain. 2005;128:2034–2041.
CrossRef
22. 22Cahn-Weiner DA, Malloy PF, Boyle PA, Marran M, Salloway S. Prediction of functional status from neuropsychological tests in community-dwelling elderly individuals. Clin Neuropsychol. 2000;14:187–195. 23. 23Carlson MC, Fried LP, Xue QL, Bandeen-Roche K, Zeger SL, Brandt J. Association between executive attention and physical functional performance in community-dwelling older women. J Gerontol B Psychol Sci Soc Sci. 1999;54:S262–S270. MEDLINE 24. 24Grigsby J, Kaye K, Baxter J, Shetterly SM, Hamman RF. Executive cognitive abilities and functional status among community-dwelling older persons in the San Luis Valley Health and Aging Study. J Am Geriatr Soc. 1998;46:590–596. MEDLINE 25. 25Royall DR, Chiodo LK, Polk MJ. Correlates of disability among elderly retirees with “subclinical” cognitive impairment. J Gerontol A Biol Sci Med Sci. 2000;55:M541–M546. MEDLINE 26. 26Pedersen PM, Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS. General cognitive function in acute stroke: the Copenhagen Stroke Study. J Neurol Rehabil. 1996;10:153–158. 27. 27Ozdemir F, Birtane M, Tabatabaei R, Ekuklu G, Kokino S. Cognitive evaluation and functional outcome after stroke. Am J Phys Med Rehabil. 2001;80:410–415. MEDLINE |
CrossRef
28. 28Malloy PF, Cummings JL, Coffey CE, et al. Cognitive screening instruments in neuropsychiatry: a report of the Committee on Research of the American Neuropsychiatric Association. J Neuropsychiatry Clin Neurosci. 1997;9:189–197. MEDLINE 29. 29Henon H, Pasquier F, Durieu I, et al. Preexisting dementia in stroke patients (Baseline frequency, associated factors, and outcome). Stroke. 1997;28:2429–2436. MEDLINE 30. 30Ebrahim S, Nouri F, Barer D. Cognitive impairment after stroke. Age Ageing. 1985;14:345–348. MEDLINE 31. 31Risser AH, Spreen O. The Western Aphasia Battery. J Clin Exp Neuropsychol. 1985;7:463–470.
CrossRef
32. 32Shewan CM, Kertesz A. Reliability and validity characteristics of the Western Aphasia Battery (WAB). J Speech Hear Disord. 1980;45:308–324. MEDLINE 33. 33Bushnell CD, Johnston DC, Goldstein LB. Retrospective assessment of initial stroke severity: comparison of the NIH Stroke Scale and the Canadian Neurological Scale. Stroke. 2001;32:656–660. 34. 34Fillenbaum G. Multidimensional functional assessment of older adults: the Duke Older Americans Resources and Services procedures. Hillsdale: Erlbaum; 1988;. 35. 35Doble SE, Fisher AG. The dimensionality and validity of the Older Americans Resources and Services (OARS) Activities of Daily Living (ADL) Scale. J Outcome Meas. 1998;2:4–24. 36. 36Paolo AM, Ryan JJ, Troster AI, Hilmer CD. Utility of the Barona demographic equations to estimate premorbid intelligence: information from the WAIS-R standardization sample. J Clin Psychol. 1996;52:335–343. MEDLINE |
CrossRef
37. 37Psychological Corp. WAIS-III - WMS-III technical manual. San Antonio: Harcourt Brace; 1997;. 38. 38Lewandowski LJ. The Symbol Digit Modalities Test: a screening instrument for brain-damaged children. Percept Mot Skills. 1984;59:615–618. MEDLINE 39. 39Selnes OA, Jacobson L, Machado AM, et al. Normative data for a brief neuropsychological screening battery. Percept Mot Skills. 1991;73:539–550. MEDLINE |
CrossRef
40. 40Delis DC, Kaplan E, Kramer J. Delis Kaplan Executive Function System Examiner’s manual. San Antonio: Psychological Corp; 2001;. 41. 41Benedict RH, Schretlen D, Groninger L, Brandt J. Hopkins Verbal Learning Test−Revised: normative data and analysis of inter-form and test-retest reliability. Clin Neuropsychol. 1998;12:43–55. 42. 42Stewart KJ, Gale SD, Diamond PT. Early assessment of post-stroke patients entering acute inpatient rehabilitation: utility of the WASI and HVLT-R. Am J Phys Med Rehabil. 2002;81:223–228. MEDLINE |
CrossRef
43. 43Barr A, Benedict R, Tune L, Brandt J. Neuropsychological differentiation of Alzheimer’s disease from vascular dementia. Int J Geriatr Psychiatry. 1992;7:621–627.
CrossRef
44. 44Hasegawa I, Hayashi T, Miyashita Y. Memory retrieval under the control of the prefrontal cortex. Ann Med. 1999;31:380–387. MEDLINE |
CrossRef
45. 45Lawrence AD, Sahakian BJ. The neuropsychology of fronto-striatal dementias. In: Woods RT editors. Handbook of the clinical psychology of ageing. Chichester: John Wiley & Sons; 1996;p. 243–265. 46. 46Saint-Cyr JA. Frontal-striatal circuit functions: context, sequence, and consequence. J Int Neuropsychol Soc. 2003;9:103–127. MEDLINE 47. 47Lezak MD. Neuropsychological Assessment. 3rd ed.. New York: Oxford Univ Pr; 1995;. 48. 48Lopez OL, Becker JT, Jagust W, et al. Neuropsychological characteristics of mild cognitive impairment subgroups. J Neurol Neurosurg Psychiatry. 2006;77:159–165. MEDLINE |
CrossRef
49. 49Saxby BK, Harrington F, McKeith IG, Wesnes K, Ford GA. Effects of hypertension on attention, memory, and executive function in older adults. Health Psychol. 2003;22:587–591. MEDLINE |
CrossRef
50. 50Ferro JM, Madureira S. Age-related white matter changes and cognitive impairment. J Neurol Sci. 2002;203-204:221–225. 51. 51Filley CM. The behavioral neurology of cerebral white matter. Neurology. 1998;50:1535–1540. MEDLINE 52. 52Gunning-Dixon FM, Raz N. Neuroanatomical correlates of selected executive functions in middle-aged and older adults: a prospective MRI study. Neuropsychologia. 2003;41:1929–1941. MEDLINE |
CrossRef
53. 53Vataja R, Pohjasvaara T, Mantyla R, et al. MRI correlates of executive dysfunction in patients with ischaemic stroke. Eur J Neurol. 2003;10:625–631. MEDLINE |
CrossRef
54. 54Reed BR, Eberling JL, Mungas D, Weiner M, Kramer JH, Jagust WJ. Effects of white matter lesions and lacunes on cortical function. Arch Neurol. 2004;61:1545–1550. MEDLINE |
CrossRef
55. 55Ardila A, Pineda D, Rosselli M. Correlation between intelligence test scores and executive function measures. Arch Clin Neuropsychol. 2000;15:31–36. MEDLINE |
CrossRef
56. 56Alexander GE, Furey ML, Grady CL, et al. Association of premorbid intellectual function with cerebral metabolism in Alzheimer’s disease: implications for the cognitive reserve hypothesis. Am J Psychiatry. 1997;154:165–172. 57. 57Stern RA, Silva SG, Chaisson N, Evans DL. Influence of cognitive reserve on neuropsychological functioning in asymptomatic human immunodeficiency virus-1 infection. Arch Neurol. 1996;53:148–153. MEDLINE 58. 58Pohjasvaara T, Mantyla R, Ylikoski R, Kaste M, Erkinjuntti T. Clinical features of MRI-defined subcortical vascular disease. Alzheimer Dis Assoc Disord. 2003;17:236–242. MEDLINE |
CrossRef
59. 59Loranger M, Lussier J, Pepin M, Hopps SL, Senegal B. Information-processing speed and assessment of early response latency among stroke patients. Psychol Rep. 2000;87:893–900. MEDLINE |
CrossRef
60. 60Dancause N, Ptito A, Levin MF. Error correction strategies for motor behavior after unilateral brain damage: short-term motor learning processes. Neuropsychologia. 2002;40:1313–1323. MEDLINE |
CrossRef
61. 61Yelnik A, Bonan I, Debray M, Lo E, Gelbert F, Bussel B. Changes in the execution of a complex manual task after ipsilateral ischemic cerebral hemispheric stroke. Arch Phys Med Rehabil. 1996;77:806–810. Abstract |
Full-Text PDF (656 KB)
|
CrossRef
62. 62Walker CM, Sunderland A, Sharma J, Walker MF. The impact of cognitive impairment on upper body dressing difficulties after stroke: a video analysis of patterns of recovery. J Neurol Neurosurg Psychiatry. 2004;75:43–48. MEDLINE 63. 63Paolucci S, Antonucci G, Gialloreti LE, et al. Predicting stroke inpatient rehabilitation outcome: the prominent role of neuropsychological disorders. Eur Neurol. 1996;36:385–390. MEDLINE |
CrossRef
64. 64Stephens S, Kenny RA, Rowan E, et al. Neuropsychological characteristics of mild vascular cognitive impairment and dementia after stroke. Int J Geriatr Psychiatry. 2004;19:1053–1057. MEDLINE |
CrossRef
65. 65Ylvisaker M, Szekeres SF, Haarbauer-Krupa J. Cognitive rehabilitation: organization, memory, and language. In: Ylvisaker M editors. Traumatic brain injury rehabilitation: children and adolescents. 2nd ed.. Woburn: Butterworth-Heinemann; 1998;p. 181–220. 66. 66Bokeriia LA, Golukhova EZ, Polunina AG, Davydov DM, Begachev AV. Neural correlates of cognitive dysfunction after cardiac surgery. Brain Res Brain Res Rev. 2005;50:266–274. MEDLINE 67. 67Knipp SC, Matatko N, Wilhelm H, et al. Evaluation of brain injury after coronary artery bypass grafting (A prospective study using neuropsychological assessment and diffusion-weighted magnetic resonance imaging). Eur J Cardiothorac Surg. 2004;25:791–800. Abstract | Full Text |
Full-Text PDF (217 KB)
|
CrossRef
a Research and Development, Veterans Affairs Medical Center, Durham, NC b Health Services and Development, Veterans Affairs Medical Center, Durham, NC c Department of Physical Medicine and Rehabilitation, Veterans Affairs Medical Center, Durham, NC d Department of Psychiatry, Duke University Medical Center, Durham, Durham, NC e Department of Medicine, Duke University Medical Center, Durham, Durham, NC.  Reprint requests to Sandra Zinn, PhD, Research & Development (151), VA Medical Center, 508 Fulton St, Durham, NC 27705
Supported by Veterans Affairs Rehabilitation Research & Development (career development award). No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated. PII: S0003-9993(06)01523-1 doi:10.1016/j.apmr.2006.11.015 © 2007 American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation. Published by Elsevier Inc. All rights reserved. | |
|
FULL TEXT01523-1/journalimage?img=PIIS0003999306015231.gr1.lrg.gif&fig=fig1&kwhquery=&issn=0003-9993&ishighres=false&allhighres=false&free=yes','journalimage');)
