Journal of Communication Disorders 35 (2002) 171186

Cognition and aphasia: a discussion and a study

Nancy Helm-Estabrooks*
Harold Goodglass Aphasia Research Center 12A, Boston Veterans Administration Medical Center, Boston University School of Medicine, 150 South Huntington Avenue, Boston, MA 02130, USA Received 23 October 2001; received in revised form 18 December 2001; accepted 18 December 2001

Abstract The relation between other aspects of cognition and language status of individuals with aphasia is not well-established, although there is some evidence that integrity of nonlinguistic skills of attention, memory, executive function and visuospatial skills can not be predicted on the basis of aphasia severity. At the same time, there is a growing realization among rehabilitation specialists, based on clinical experience and preliminary studies, that all domains of cognition are important to aphasia therapy outcomes. This paper describes a new study of the relation between linguistic and nonlinguistic skill in a group of individuals with aphasia. No signicant relationship was found between linguistic and nonlinguistic skills, and between nonlinguistic skills and age, education or time post onset. Instead, individual proles of strengths and weaknesses were found. The implications of these ndings for management of aphasia patients is discussed. Learning outcomes: Readers of this papers will be able to: list ve primary domains of cognition and relate each to an aspect of aphasia therapy; describe at least three studies that examined the relation between cognition and aphasia; describe four nonlinguistic tasks of cognition that can be used with a wide range of aphasia patients. # 2002 Elsevier Science Inc. All rights reserved.
Keywords: Aphasia; Cognition; Cognitive examination; Neuropsychological examination

Present address: 150 Medway Street, Providence, RI 02906, USA. Tel.: 1-617-232-9500x5844/ 401-331-2866; fax: 1-401-861-0631. E-mail address: nancyhe@bu.edu (N. Helm-Estabrooks). 0021-9924/02/$ see front matter # 2002 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 1 - 9 9 2 4 ( 0 2 ) 0 0 0 6 3 - 1

172

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

1. Introduction 1.1. Cognition and aphasia According to Sarno (1998), ``aphasia rehabilitation must be viewed as a process of patient management in the broadest sense'' (p. 615). She went on to list many factors that aphasia therapists must consider for execution of effective interventions. Among these factors are ``neuropsychological decits.'' Yet, despite the inherent importance of the neuropsychological (cognitive) status of aphasia patients to the development of treatment plans and approaches and to expectations of positive outcomes, most aphasia therapists are guided solely by the results of language exams. Some exceptions are Luria (1966), a Russian neuropsychologist actively engaged in aphasia rehabilitation, who looked beyond language in developing his approaches, and Chapey (1994) who conceptualized aphasia therapy as ``cognitive intervention.'' Interestingly, however, no chapter in the widely used text edited by Chapey (2001) is devoted to the neuropsychological examination of aphasia patients although in this fourth edition, several excellent chapters on cognitive-psychological approaches to aphasia therapy have been added. A chapter describing neuropsychological assessment does appear in the Manual of Aphasia Therapy (Helm-Estabrooks & Albert, 1991) but it is geared toward neuropsychologists. The second edition of this text is in preparation and will have a chapter directed toward speech and language pathologists interested in the cognitive examination of their aphasia patients. As a basis for developing an approach to determining the cognitive status of aphasia patients, it is advisable to begin with a denition of cognition. Neisser (1967) dened cognition as ``all the processes by which sensory input is transformed, reduced, elaborated, stored, recovered and used'' (p. 4). More recently, Bayles (2001) stated that ``cognition refers to what we know and the processes that enable us to acquire and manipulate information.'' If cognition is to be formally examined, however, we must go beyond these broad denitions and consider the components or domains of cognition. Cognition may be regarded as having ve primary domains: attention, memory, executive functions, language, and visuospatial skills. Pertinent to the topic of this paper, consider that each of these cognitive domains are recruited and used to varying extents during the aphasia rehabilitation process. Most often, aphasia therapy is directly related to language with proles of spared and impaired language functions serving as the principle guides for making treatment decisions. At the same time, attention is a powerful variable as it is basic and critical to all activities. Failure to attend results in failure to process information despite what may be relatively spared ability to understand spoken or graphic stimuli. Furthermore, there is no question that aphasia therapy is a learning experience and that learning relies upon memory processes. Moreover, it would be a rare treatment protocol that did not call upon some aspect of visuospatial skills,

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

173

e.g., recognition and/or production of pictured, orthographic or gestured stimuli. Finally, the ultimate goal of aphasia therapy is to improve individuals' ability to communicate within everyday settings with all their unpredictable demands and uctuating conditions. This task requires goal-oriented behavior and exible problem-solving, both hallmarks of executive functions. Thus, one can make a strong case for factoring the relative integrity of all domains of cognition into the treatment process for a particular patient with aphasia. To establish individual proles of cognitive functions, a neuropsychological assessment is required. Unfortunately, many commonly used neuropsychological tests have linguistic processing and/or production demands that make them largely invalid for use with aphasic individuals. Thus, studies of cognition in people with aphasia typically have employed cognitive tests with no obvious linguistic demands. For example, one of the tests used by Basso, DeRenzi, Scotti and Spinnler (1973) was the Raven's coloured progressive matrices (Raven, Court, & Raven, 1979), a test of visual analogic thinking in which each item consists of a visual matrix with a missing piece. The task is to select the piece that best completes the particular design from an array of six. Interestingly, Basso et al., found that the correlation between scores earned by 33 subjects with left hemisphere damage on the Raven's and scores earned on a language test of naming and comprehension was ``practically zero.'' In other words, it was impossible to predict analogic thinking ability on the basis of language test performance. In a study of cognition and aphasia, Helm-Estabrooks, Bayles, Ramage and Bryant (1995) found that scores earned on a battery of ``nonverbal'' cognitive tasks did not correlate signicantly with aphasia severity scores of 32 aphasic patients, none of whom had global aphasia. Thus, this study provided additional evidence that clinicians cannot predict the relative integrity of other domains of cognition on the basis of language decits in aphasic stroke patients. A study by Van Mourik, Vershaeve, Boon, Paquier, et al. (1992) focused on cognition in 17 patients with global aphasia. They used a battery of nonlinguistic tasks that they referred to as the global aphasic neuropsychological battery (GANBA), and a test of auditory comprehension. (Although not explained by the authors, this reader assumes that none of the patients had meaningful verbal output given their diagnosis, so this was not tested). The GANBA was comprised of six tasks, tests and subtests from published sources, and one task designed by the investigators. Together, the GANBA tasks targeted the areas of attention/ concentration, memory, intelligence (as tested by the Raven's progressive matrices), visual recognition and nonverbal auditory recognition. All subjects could perform the tasks. Van Mourik et al. reported that scores earned on the GANBA were independent of the level of spoken language comprehension. Although GANBA proles reected patient heterogeneity, test results also allowed the investigators to identify two main groups of globally aphasic patients that they discussed in terms of implications for treatment. Group 1 patients performed well on the GANBA and were thought to have sufciently intact

174

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

cognitive functions to respond to language-oriented aphasia treatment. Group 2 patients had variable patterns of decits (e.g., impaired concentration, visual skills), suggesting that these impairments must be addressed directly before language treatment is initiated. Van Mourik et al. also described a third group of globally aphasic patients who were excluded from their study because they could not perform the basic requirements of the GANBA tasks. Given the apparent severity of their cognitive decits, it was suggested that therapy time be directed toward work with the communicative partners of these individuals. Although Van Mourik et al. only discussed the implications of aspects of cognition for language treatment, Hinckley, Carr, and Patterson (2001) studied the relationship between cognitive abilities, and specic treatment type (``contextbased'' versus ``skill-based'' and treatment duration with 18 chronically aphasic patients. The goal of the treatment was to improve performance on the functional task of ordering items from a catalogue. Nine patients received ``context-based'' therapy that involved the use of individualized compensatory strategies such as using notebooks for graphic cueing and pressing keys for electronic speech output. The other nine received ``skill-based'' therapy in which clinicians used cueing hierarchies to improve retrieval and production of words related to catalogue-ordering. Hinckley et al. found that the lower the scores on the Raven's and Wisconsin card sort test (WCST; Grant & Berg, 1993), the longer it took patients to achieve performance criterion with ``context-based'' therapy. Furthermore, scores earned on these tests by individuals receiving the context-based treatment were signicantly related to the ability to perform the functional task 6 months after treatment ended. In discussing their ndings, these investigators discuss the roles that executive functions, learning and memory skills played in the success of their particular treatment protocols. 1.2. Assessing cognitive status of individuals with aphasia Ideally, aphasia therapists obtain information as to their patients' current cognitive status from neuropsychologists experienced in testing of individuals with aphasia. Unfortunately, neuropsychological test data are rarely available in most clinical practices. Furthermore, few tools exist for speech and language pathologists to briey examine the neuropsychological status of their aphasia patients. One exception is the Raven's progressive coloured matrices (Raven, 1995), which is easily obtained and can be administered to most individuals with aphasia independent of their language decits. This test, however, mainly targets only one type of cognitive behavior, i.e., visual analogic thinking. To test a broader spectrum of skills, researchers typically assemble batteries of nonlinguistic tests and tasks (e.g., Van Mourik et al., 1992; Helm-Estabrooks et al., 1995) that may include subtests of standardized assessment tools, or nonstandardized tasks. Not only are these batteries difcult and costly to assemble, their administration time may be prohibitively long within the current environment of

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

175

health care. Furthermore, some tests such as the Wechsler memory scale (Wechsler, 1997), from which speechlanguage pathologists might wish to extract certain subtests, can be purchased only by Ph.D. level professionals. The cognitive linguistic quick test (CLQT; Helm-Estabrooks, 2001) was recently developed to meet the needs of clinicians who want to obtain basic information about the relative status of attention, memory, executive functions, language, and visual spatial skills of their patients. The CLQT consists of 10 tasks and can be administered in 1530 minutes. Field testing was conducted with both nonclinical (no known neurological dysfunction) and neurologic examinees who were demographically representative of the United States adult population according to sex, age, education level, race/ethnicity, and region. The target ages were 1889 years. One pilot test and three research studies led to the nal version of the CLQT. In Study 1, 30 certied speechlanguage pathologists and licensed psychologists tested 92 individuals, 28 with one of the following neurologic diagnoses: left CVA, right CVA, bilateral CVA's, Alzheimer's disease, and closed head injury. Each of these individuals was matched by age, race/ ethnicity and educational level to at least two clinical examinees (except for those clinical cases with 11 years or less of education who had one match each) for a total of 64 nonclinical examinees. In Study 2, 61 clinicians in 31 states tested 154 nonclinical examinees. In Study 3 another 119 examinees (38 clinical and 81 nonclinical) were tested. On the basis of scores earned by nonclinical examinees during the research studies, two sets of normal cut-off scores (ages 1869 and 70 89) were established for all tasks, cognitive domains, and overall cognitive performance. 1.3. Purpose of present study The purpose of the present study was to further elucidate the status of cognitive abilities in a group of individuals whose aphasia ranged from mild to severe levels. To accomplish this, eight CLQT tasks (four linguistic and four nonlinguistic) were used to (1) test the relationships between linguistic and nonlinguistic task performance, and the relationships between nonlinguistic performance and age, education, and time post onset of aphasia, and (2) examine group and individual proles of spared and impaired cognitive functions vis-a-vis overall performance on the sets of linguistic and nonlinguistic tasks and on the individual tasks. 2. Methods 2.1. Subjects Subjects were 13 right-handed, left hemisphere stroke patients (ve females and eight males) from various settings across the United States (e.g., rehabilitation

176

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

and extended care facilities, own home). Eleven of the 13 had right hemiplegia that forced them to use their non-preferred left hand for graphomotor activities. They ranged in age from 4677 years (mean 62 years, S:D: 10:6) and in education from 8 to 22 years (mean 14 years, S:D: 4). Time post onset at time of testing ranged from 2 to 118 months (mean 34 months, S:D: 38). 2.2. CLQT tasks Four linguistic and four nonlinguistic CLQT tasks were administered in one session. All administration and scoring methods, and test materials were standardized. All tasks had time limits except for personal facts (see below). As a group these tasks were chosen to assess the ve primary domains of cognition, i.e., attention, memory, executive, language, and visuospatial skills. The tasks were as follows. 2.3. Linguistic tasks 2.3.1. Personal facts This task primarily assesses episodic memory, word retrieval and verbal language production. Clinicians ask four questions relating to the examinee's date and place of birth, current age and address. One point is given for each correct element given from memory. 2.3.2. Confrontation naming This task primarily assesses semantic conceptual knowledge, word retrieval and production. Ten common, pictured items are presented one-at-a-time for naming. A 3-point scoring system allows for full, partial or no credit depending on correctness of responses. 2.3.3. Story retelling and paragraph comprehension The purposes of this task are to assess immediate/working memory for verbally presented facts embedded in a story narrative and auditory comprehension of three key elements of the story. A short story is read aloud and patients are asked to repeat the story verbatim. A check-list of the story's 18 information units allows clinicians to quickly record the items accurately recalled. Story retelling is followed by three pairs of ``yes''/``no'' questions to probe auditory comprehension. Raw scores are converted to a 10-point scale based on normative data. 2.3.4. Generative naming This task primarily assesses executive functions, working memory and verbal language skills related to semantic and phonological knowledge. Examinees are given 1 minute to list as many animals as possible and 1 minute to list words (excluding proper nouns) beginning with the letter ``m.'' Total raw scores for both lists are converted to a 9-point system based on normative data.

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

177

2.4. Nonlinguistic cognitive tasks 2.4.1. Symbol cancellation The purpose of this task is to test visual attention, scanning, discrimination, inhibition, and response shifting within four quadrants of space. In this paper and pen task, 36 abstract symbols are arranged in a pseudo-random fashion with the target stimulus appearing three times in each quadrant of space to allow for assessment of visual eld decits and visual neglect. Foils bear resemblance to the target to increase visual attention demands. The task is to cross out target symbols. One point is given for each correct symbol cancelled and one point is subtracted for each incorrect symbol cancelled. 2.4.2. Alternating symbol trails The primary purpose of this task is to assess the executive functions of planning, working memory, and mental exibility without placing demands on the language system, but the task also calls upon visual attention and perception. Two learning trials involving the single concepts of size and shape are used in preparation for the test item which involves drawing a single line to connect a total of 11 circles and triangles in an alternating fashion according to size and shape beginning with the smallest circle. The maximum possible score is 10 points. 2.4.3. Memory for designs The purpose of this task is to test immediate/working visual memory and attention without language demands. Three target abstract designs are presented one-at-a-time for memorization. They must then be identied immediately from arrays of six that include four foils similar to the targets. The highest possible score is six points. 2.4.4. Mazes This task was designed to assess executive functions, specically those involved in planning a course of action, rejecting/inhibiting incorrect choices, and correcting mistakes when made. Other skills assessed are attention and visual perception. Two mazes of two levels of difculty are used. The object of both is to trace a continuous line through the maze ``alleys'' without entering any dead-ends or crossing any line. The highest possible score for each maze is four (correct solution) for a total of eight points. Even with a correct solution, one point is subtracted each time the examinee's line travels at least halfway up an incorrect path but is self-corrected. 2.5. Administration time Five of the eight tasks have established time limits and total time for testing was approximately 22 minutes. All were administered in the following temporal

178

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

order: personal facts, symbol cancellation, confrontation naming, story retelling, symbol trails, generative naming, design memory, mazes. (Note: A nonverbal clock drawing/setting task was administered between confrontation naming and story retelling but the scores on this task were not used for this study). 3. Results Individual data for linguistic and nonlinguistic task performance, education, age, and time post onset at time of testing as well as correlation values obtained are presented in Table 1. 3.1. Overall performance on the linguistic tasks The highest possible score for the four linguistic tasks was 37 points. Scores for the 13 aphasic subjects ranged from 0 to 26 points with a mean of 10.5 points (S:D: 11). Based on data collected from nonclinical subjects used for CLQT standardization, the normal cut-off score for the four tasks was 29 for those 1869 years of age and 27 for those 7089 years. Thus, no aphasic patient earned an overall linguistic score at or above normal cut-off, and the severity of their language decits ranged from severely to mildly impaired.
Table 1 Individual data for linguistic and nonlinguistic task scores, education, age, and time post onset for 13 right-handed aphasic stroke patients Subject no. 1 2 3 4 5 6 7 8 9 10 11 12 13 Mean S.D. Range Gender F M M F M M M M M F M F F Linguistic score 25 6 21 22 1 26 21 12 2 0 0 0 0 10.46 10.91 26.00 Nonlinguistic score 24 11 19 28 30 33 18 27 19 12 30 24 34 23.77 7.51 23.00 Years of education 14 12 12 12 22 12 8 12 14 14 22 12 12 13.69 3.99 14 Age 54 71 53 65 86 64 77 50 53 77 68 55 46 61.62 10.36 31.00 Months post onset 4 15 2 3 18 72 102 118 32 18 22 17 14 33.62 38.42 116.00

Correlations: linguistic score nonlinguistic score, r 0:08; nonlinguistic score years of education, r 0:32; nonlinguistic score age, r 0:42; nonlinguistic score months post onset, r 0:09.

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

179

3.2. Overall performance on nonlinguistic tasks The highest possible score for the four nonlinguistic tasks was 36 points. Scores for the 13 aphasic subjects ranged from 11 to 34 points with a mean of 23.8 points (S:D: 7:5). Based on data collected from nonclinical subjects for CLQT standardization, the normal cut-off score for the four tasks was 32 for those 1869 years of age and 24 for those 7089 years. Two aphasic subjects (6 and 13) earned scores above the normal cut-off (see Table 1). Note that subject 6 earned an overall nonlinguistic score of 26 (only three points below normal cut-off), indicating mild impairment in linguistic processing. While no other aphasic subject earned an overall linguistic score at or above normal cut-off, two subjects (5 and 11) missed by just two points. Note also, that subject 13, who scored above the cut-off on the nonlinguistic tasks, earned a linguistic score of zero. 3.3. Correlations A correlation computed between performance scores on the four linguistic and four nonlinguistic cognitive scores was nonsignicant (r 0:08). In other words, performance on tasks involving attentional, executive, memory, and visuospatial processes could not be predicted on the basis of performance on the linguistic tasks. Likewise, correlations between nonlinguistic scores and years of education (r 0:32), age (r 0:42) and months post onset at time of testing (r 0:09) were nonsignicant. 3.4. Profiles of spared and impaired abilities In addition to analyzing group data for linguistic and nonlinguistic tasks, scores earned by aphasic individuals on each CLQT task were examined vis-a-vis the cutoff scores established for normal performance on that task for age groups 1869 and 7089 years. The scores earned by each individual were also measured against the normal cut-off scores for the four linguistic and four nonlinguistic tasks combined for the two age groups. As mentioned above, of the possible 37 total points for the linguistic tasks, the normal cut-off is 29 for the 1869 years group, and 27 for the 7089 years group for a difference of two points. In contrast, of the possible 36 total points for nonlinguistic tasks, the normal cut-off is 32 for the younger group and 24 for the older group for a difference of eight points. These scores reect the wellstudied observation that nonlinguistic skills are more vulnerable to the effects of old age than are linguistic skills. By examining individual performance proles, therefore, each patient could be measured against the normal cut-scores for his or her age group. This information is displayed in Table 2. 3.5. Overall linguistic and nonlinguistic scores for individuals When overall scores earned by aphasic individuals for four linguistic tasks were compared with the overall linguistic cut-off scores for normal performance,

180 N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

Table 2 Individual aphasia subject scores on CLQT linguistic and nonlinguistic tasks and deviation from normal cut-off scores for age Subject no. Linguistic tasks Personal Confrontation Story Generative facts naming retelling naming 1 2 3 4 5 6 7 8 9 10 11 12 13 8a 1 7 7 0 8a 7 0 0 0 0 0 0 10a 2 9 9 0 9 10a 9 2 0 0 0 0 1010 5 2 3 4 1 6a 3 3 0 0 0 0 0 65 2 1 2 2 0 3 3 1 0 0 0 0 0 54 Nonlinguistic tasks Symbol Symbol Design Mazes cancellation trails memory 12a 0 4 12a 12a 10 10a 12a 11a 7 12a 11a 12a 1110 6 2 7 7 9a 10a 4 4 2 3 9a 1 9a 96 6a 5a 4 5a 5a 5a 4a 3 2 2 5a 5a 5a 54 0 4a 4 5a 4 8a 0 7.5a 4 0 4 7a 8a 74 Overall Difference Overall Difference linguistic in score nonlinguistic in score score score 25 6 21 22 1 26 21 12 2 0 0 0 0 2927 5 21 8 7 29 3 6 17 27 27 29 29 29 24 11 19 28 30 33a 18 27 19 12 30 24 34a 3224 8 13 13 4 2 1 6 5 13 12 2 8 2

() Age cut-off scores 88

() Cut-off scores for normal task performance based on nonclinical subjects for two age groups: 1869 years and 7089 years (see CLQT manual). a Score is at or above normal cut-off.

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

181

it was found that the difference scores ranged from 29 to 3, with a mean difference of 18.2 points. Two individual earned scores at cut-off for relating personal facts and confrontation naming. One person earned a normal cut-off score for story retelling, but none achieved a normal cut-off score for generative naming, which was clearly the most difcult task for the group. When overall scores earned by aphasic individuals for the four nonlinguistic were compared with the overall nonlinguistic cut-off scores for normal performance, it was found that the differences ranged from 13 to 2 points with a mean difference of 6.4 points. Two patients (6 and 13) exceeded normal cut-off scores for overall nonlinguistic task performance, as we mentioned earlier. Note that patient 6 had mild aphasia with an overall linguistic score just three points below normal cut-off while patient 13 had severe aphasia earning an overall linguistic score of zero. Closer examination of individuals with very severe aphasia (patients 5, 9, 10, 11, 12) showed great variability on nonlinguistic task performance. Two of these patients (5 and 2) earned good nonlinguistic scores (two points below normal cut-off for their age), two patients (9 and 10) had notable nonlinguistic impairment, while one patient (12) showed moderate nonlinguistic impairment. 3.6. Task specific scores for individuals Table 2 indicates that the scores earned on a particular task by an individual patient was at or above the normal cut-off for his or her age. Not surprisingly, given that these patients had the language disorder of aphasia, the best performances were seen on nonlinguistic tasks. The easiest tasks for these aphasic patients were tasks of symbol cancellation and design memory with nine or 69% of individuals earning scores at or above normal cut-off for their age. The next easiest task was mazes with six patients (46%) earning normal or above normal scores, followed by symbol trails (four subjects or 31%). Two patients (15%) earned normal cut-off scores for personal facts and confrontation naming, one (8%) for story retelling and none (0%) for generative naming. The pattern of task difculty for the 13 aphasia patients is displayed in Table 3. The four easiest tasks, all of which were nonlinguistic, must be completed within prescribed time frames. The two easiest tasks were symbol cancellation and design memory with nine patients (69%) performing at or above normal cut-off. One of the two, symbol cancellation, calls upon visual attention, scanning and discrimination. It also requires the ability to inhibit stimulus ``pull'' to symbols similar to the target and response shifting within four quadrants of space. The other easiest task, design memory, calls upon immediate/working visual memory and attention for difcult-to-encode visual stimuli. Six out of 13 patients (46%) performed at or above normal cut-off on mazes which (in addition to visuospatial skills), requires executive skills of planning and foresight. Symbol trails was the most difcult of the nonlinguistic tasks with 4/13 (31%) of patients reaching or surpassing normal cut-off scores. This demanding

182

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

Table 3 Pattern of task difficulty for 13 individuals with aphasia CLQT tasks Easiest Symbol cancellation and design memory Mazes Symbol trail Personal facts and confrontation naming Story retelling Generative naming Number/earning scores at or above normal cut-off (%) 9 6 4 2 1 (69%) (46%) (31%) (15%) (8%)

Hardest

0 (0%)

task requires working memory, attention, and visuospatial skills as well as the executive abilities of mental exibility in planning and shifting responses. All of the linguistic tasks were difcult for the aphasic patients although two individuals (1 and 6) were able to answer the four personal fact questions pertaining to name, date of birth, age and current address. One of these individuals (1) also named all 10 pictured objects correctly (10 is the normal cut-off for both age groups). One other person (7) achieved this score for confrontation naming. Story retelling which calls upon memory as well as language comprehension and production was difcult for most individuals with only one patient (6) scoring at cut-off. No patient was able to earn a normal cut-off score for generative naming, which not only calls upon language skills related to semantic and phonological knowledge, but requires some executive skills and working memory so as not to repeat previously listed items. 4. Discussion There is a growing realization among aphasia therapists that aspects of cognition, other than language, may account for the variability in response that some patients with the same aphasia severity level have to the same treatment protocols. Furthermore, it may be that higher level cognitive skills, i.e., those that allow us to plan intentional activities while exibly adjusting our goal-directed strategies in keeping with situational changes (so-called executive functions), may account for differences that some patients of equal aphasia severity show in their functional communication skills. Treatment planning would be made easier if clinicians could assume a direct relation between linguistic and nonlinguistic cognitive skills; having some condence that mildly aphasic patients will have mild problems in the areas of attention, memory, executive functions and visuospatial skills, and that these problems will increase as the aphasia severity increases. Instead, we have mounting evidence that it is impossible to predict the status of nonlinguistic cognitive skills on the basis of language skills. Helm-Estabrooks et al. (1995)

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

183

found this to be true for 32 patients representing a range of aphasia severity excluding global aphasia. Similarly, Van Mourik et al. (1992) studied only patients with global aphasia and found that not only were cognitive test scores unrelated to auditory comprehension but that three subgroups of globally aphasic patients could be identied; those with relatively spared nonlinguistic cognitive skills, those with variable patterns of decits, and those who could not respond at all to testing. One of the measures used in their study was the Raven's progressive matrices, used earlier by Basso et al. (1973) who found that virtually no relation existed between Raven's scores and a test of auditory comprehension and naming in a larger group of patients with a wider range of aphasia severity. To the mounting evidence that it is impossible to predict the status of other aspects of cognition on the basis of language test scores, we can add the results of the present study. Scores earned on a group of four linguistic tasks taken from the newly published CLQT 2001 by 13 individuals were not signicantly correlated with scores earned on a group of four CLQT nonlinguistic tasks chosen to briey assess visual attention and memory, executive functions, and visuospatial skills. Examination of individual proles of task performance underscored the poor relation between linguistic and nonlinguistic task performances. For example, of the two patients earning the overall nonlinguistic scores that were above the normal cut-off for their age, one had mild language decits (three points below normal cutoff) and one had severe decits (29 points below normal cut-off). Among all six patients with severe linguistic decits, three had relatively spared nonlinguistic skills, earning nonlinguistic scores near or above normal cut-off, two had notably impaired nonlinguistic skills, and one demonstrated moderate impairment. It would appear that second to language, executive functions are the cognitive skills most vulnerable to the effects of brain damage associated with aphasia. The results of the present study showed that the greater the task-demands for executive skills the more difcult the task for the group in general. This was true in both the nonlinguistic and linguistic domains with the ability to create trails between symbols according to progressive size and alternating shape the most difcult nonlinguistic task, and generative naming the most difcult linguistic task. Even so, four patients (two with severe aphasia) achieved normal cut-off scores on symbol trails. No one, however, was able to achieve normal cut-off for generative naming. The importance of executive functions to treatment response is underscored by the results of the Hinckley et al. (2001) study in which they found that the lower the scores on the Raven's and Wisconsin card sort test, the longer it took patients to achieve performance criterion for ``context-based'' therapy. Furthermore, scores earned by patients receiving this type of treatment on the Raven's and the Wisconsin were signicantly related to the ability to perform the functional task 6 months after treatment. These investigators concluded their discussion on these results with the following statement. ``Furthermore, if conrmed and replicated, aspects of executive function and cognitive status may become important contributors to clinical decision-making about treatment type and amount.''

184

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

In the present study, visual discrimination and attention skills were most spared with 9/13 achieving scores at or above normal cut-off for symbol cancellation. Interestingly, 2/4 patients with low symbol cancellation scores were two of the three individuals over 70 years of age. Because, the CLQT normal cut-off scores are adjusted for age, this suggests that visual attention may be particularly sensitive to the effects of brain damage in the elderly. The relation between attentional skills and aphasia has been a topic of growing interest. For a review of this literature see Murray (1999). Given that various attentional mechanisms may be impaired in the presence of aphasia and inuence language performance, some clinical researchers have designed treatment protocols to directly or indirectly treat forms of attention. For example, Strum and Willmes (1991) used a computerized treatment protocol to train accurate response to single and multiple nonlinguistic stimuli presented in auditory and visual modalities. Subjects improved in speed and accuracy of the attentional tasks but showed little concurrent improvement on tasks used to measure other cognitive functions (e.g., Raven's matrices). In contrast, the study of two severely aphasic patients, for whom repeated baseline measures were obtained before beginning a nonlinguistic attention treatment program, Helm-Estabrooks, Connor and Albert (2000) found that Raven coloured progressive matrices scores did improve with attention training. At the same time, modest gains occurred in auditory comprehension test scores of these two subjects. Crosson (2000) and his colleagues are currently engaged in the study of the effects of attention treatments on remediation of aphasic naming disorders. Crosson was referring to attention vis-a-vis aphasia rehabilitation when he pointed out, ``there is a considerable need for research'' (p. 392), but the same may be said of other domains of cognition and their importance to treatment of aphasia. Indeed, it would appear that this is an area ripe for careful investigation as we rightfully move away from the conceptualization of language as being separate from cognition and accept that language is one aspect of cognition. If aphasia therapists are to consider their patients' overall neuropsychological proles in developing or adopting appropriate treatment approaches, then they must be prepared to at least screen these abilities. The CLQT tasks used in the present study appear to be useful for a brief, standardized assessment of all domains of cognition. More in-depth testing with domain-specic tests probably will be required if notable attention, memory, executive, and visuospatial decits are identied. Certainly, there is much to be learned about the relation between aspects of cognition and aphasia treatment outcomes. Fortunately, it would appear that some clinical investigators are now pursuing this line of research. Acknowledgments Some of the data used in this study were standardization data of the CLQT. Copyright # 2000 by The Psychological Corporation. Used by permission. All

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

185

rights reserved. Dr. Helm-Estabrooks' work was supported, in part, by a grant from the McDonnell Foundation. Appendix A. Continuing education 1. Which of the following is a test of visual analogic thinking? a. Wisconsin card sort test b. Porteus mazes c. Raven's progressive matrices d. Cognitive linguistic quick test e. None of the above 2. In the present study what was the correlation between linguistic and nonlinguistic cognitive skills? a. Highly significant b. Non significant c. Weakly significant d. Significant in the negative direction e. Not determinable 3. What skill is needed for successfully completing symbol trails? a. Mental flexibility b. Semantic knowledge c. Verbal mediation d. Episodic memory e. All of the above 4. Nonlinguistic cognitive skills in the presence of aphasia are related to: a. Years of education b. Time post onset c. Severity of aphasia d. Linguistic skills e. None of the above 5. What is the cognitive domain most likely to be impaired in aphasia patients? a. Executive functions b. Analogic thinking c. Attention d. Language e. Memory References
Basso, A., DeRenzi, E., Faglioni, P., Scotti, G., & Spinnler, H. (1973). Neuropsychological evidence for the existence of cerebral areas critical to the performance of intelligence tasks. Brain, 96, 715728.

186

N. Helm-Estabrooks / Journal of Communication Disorders 35 (2002) 171186

Bayles, K.A. 2001. Understanding the neuropsychological syndrome of dementia. Seminars in Speech and Language, 22(4), 251259. Chapey, R. (1994). Cognitive intervention: stimulation of cognition, memory, convergent thinking, convergent thinking, and evaluative thinking. In R. Chapey (Ed.), Language intervention strategies in adult aphasia (3rd ed.) (pp. 220245). Baltimore, MD: Williams & Wilkins. Chapey, R. (2001). Language intervention strategies in aphasia and related neurogenic communication disorders (4th ed.). Baltimore, MD: Williams & Wilkins. Crosson, B. (2000). Systems that support language processes: attention. In S. E. Nadeau, L. Gonzalez, J. Rothi, & B. Crosson (Eds.), Aphasia and language (pp. 372398). New York: The Guilford Press. Helm-Estabrooks, N., & Albert, M. L. (1991). Manual of aphasia therapy. Austin, TX: Pro-Ed Inc. Helm-Estabrooks, N., Bayles, K., Ramage, A., & Bryant, S. (1995). Relationship between cognitive performance and aphasia severity, age and education: females versus males. Brain and Language, 51(1), 139141. Helm-Estabrooks, N. (2001). Cognitive Linguistic Quick Test. San Antonio, TX: The Psychological Corporation. Helm-Estabrooks, N., Connor, L. T., & Albert, A. L. (2000). Treating attention to improve auditory comprehension in aphasia. Brain and Language, 74, 469472. Hinckley, J. J., Carr, T. H., & Patterson, J. P. (2001). Relationships between cognitive abilities, treatment type and treatment time in aphasia. Paper presented at the Clinical Aphasiology Conference, Sante Fe, NM, June. Grant, D. A., & Berg, E. A. (1993). Wisconsin card sorting test. Tampa, FL: Psychological Assessment Resources. Luria, A. R. (1966). Higher cortical functions in man. New York: Basic Books. Murray, L. L. (1999). Attention and aphasia: theory, research and clinical implications, research and clinical implications. Aphasiology, 13(3), 91111. Neisser, U. (1967). Cognitive psychology. New York: Appleton Century-Crofts. Raven, J. C., Court, J. H., & Raven, J. (1979). Manual for Raven's progressive matrices and vocabulary scales. London: Lewis & Co. Raven, J. C. (1995). Raven's coloured progressive matrices. San Antonio, TX: The Psychological Corporation. Sarno, M. T. (1998). Recovery and rehabilitation in aphasia. In M. T. Sarno (Ed.), Acquired aphasia (3rd ed.). (pp. 595631). New York: Academic Press. Strum, W., & Willmes, K. (1991). Efficacy of a reaction training on various attentional and cognitive functions in stroke patients. Neuropsychological Rehabilitation, 1, 259280. Van Mourik, M., Verschaene, M., Boon, P., Paquier, P. et al. (1992). Cognition in global aphasia: indicators for therapy. Aphasiology, 6(5), 491499. Wechsler, D. (1997). Wechsler memory scaleIII. San Antonio, TX: The Psychological Corporation.