The Relationship Between Visuoconstructive Abilities and Language Performance in Patients With Aphasia After Stroke

Abstract

This study aimed to investigate the visuoconstructive abilities and the relationship between visuoconstructive function and language performance in aphasic patients. Right-handed 24 aphasic patients (males 14, females 10) with at least 3 months post-stroke and 32 age-matched healthy controls participated in this study. Visuoconstructive function was assessed by 3 levels of task difficulty: simple (drawing objects), intermediate (clock drawing), and complex (copy subtest of Rey complex figure test and block construction). Aphasic patients were divided into 3 sub-groups (mild, moderate to severe, and very severe group) according to severity of aphasia and compared with the control group, respectively. We analyzed the relation all levels of visuoconstructive tasks to aphasia quotient (AQ) and sub-domain scores of K-WAB. Moderate to severe aphasia group demonstrated no significant differences in scores of simple drawing objects compared to controls, but clock drawing, Rey complex figure copy and block design showed significantly decreased scores. Very severe group showed significantly lower scores in all levels of visuoconstructive tasks than the control. Correlation between all levels of visuoconstructive tasks except drawing objects and AQ were found to be statistically significant. Among the tasks, the clock drawing test revealed the highest correlation with language performance. Visuoconstructive abilities varied according to the severity of aphasia and the level of visuoconstructive tasks. Therefore, a thorough individual assessment of visuoconstructive function is needed to plan and predict the treatment and prognosis of aphasia and the clock drawing test may be a useful screening tool to evaluate this function.

Graphical Abstract

Highlights

• • Visuoconstructive and language deficits are significantly correlated.

• • Visuoconstructive abilities varied with the severity of aphasia and task level.

INTRODUCTION

Aphasia is an acquired language disorder after brain injury of dominant hemisphere including stroke, traumatic brain injury and others [1]. Cognitive dysfunction is commonly found in people with aphasia (PWA) and affects language and communication performance, as well as the treatment and prognosis of aphasia. The relations between other aspects of cognition and language of PWA has been actively conducted but is not well-established, especially in terms of aphasia therapy outcomes and prediction on basis of aphasia severity. Cognitive function includes attention, memory, executive function, visuospatial skills, and language function, and are called non-linguistic cognitive functions except language function [2,3]. Most tests that evaluate nonlinguistic cognitive function are performed through language which involves obvious linguistic processing. Therefore, in PWA with impaired language function, it is not clear whether the low cognitive function scores are due to reduced cognitive function or reduced language function. For this reason, many studies of cognitive functions in PWA have used non-verbal cognitive tests with no explicit verbal demands and for decades have focused primarily on non-linguistic abilities [4,5]. Visuoconstructive tasks, such as drawing and block design, are typical non-verbal tests and they can be evaluated relatively independently of language function [4].

Visual construction refers to the ability to assemble the elements of 2- or 3-dimensional objects, respecting their orientations and spatial relationships; an impairment of this ability is called constructional apraxia [6]. The incidence of visual constructive dysfunction is reportedly 15%–40% in unilateral brain injury [7]. Although deficits are more frequent and marked in patients with a right brain damage due to problems of visuospatial processing, it can also be associated with left hemispheric lesions due to executive problems [6].

Previous studies have suggested that cognitive deficits, including visuoconstructive and executive functions [2,8], are common in PWA after stroke and several studies have investigated constructional abilities in PWA [8,9,10,11]. Although most studies have agreed that a visuoconstructive performance in the aphasia group is low compared to healthy controls, the relationship between visuoconstructive function and language performance revealed mixed findings [2,4,9,11,12,13,14,15,16]. Some studies reported no significant relationship between linguistic and non-linguistic tests including visuoconsturctive task, while others reported positive and statistically significant correlations [2,4,5,11,12,13]. Inconsistent results have also been reported depending on the severity of aphasia [3,4,13,15,16]. Certain types of aphasia evidently impaired non-verbal cognitive abilities including visuoconstructive function, while other types of aphasia were within normal limits [4,15,16]. Taken together, the results of previous studies suggest that non-verbal abilities including visuoconstructive function may be affected in PWA, but there is a significant variability [4,5]. In addition, direct comparisons between studies are limited because most studies used different types of visuoconstructive tasks and few studies have examined the drawing skills using different levels of difficulty.

This study was conducted to investigate the visuoconstructive abilities in PWA using various visuoconstructive tasks of different levels of difficulty and to elucidate the relationship between visuoconstructive function and language performance in PWA.

MATERIALS AND METHODS

Participants

Twenty-four patients who had aphasia (14 males and 10 females) after left hemispheric stroke participated for the study. The inclusion criteria were; 1) aphasia confirmed by clinical features and a comprehensive aphasia test battery, 2) first ever stroke with visible left hemispheric lesions based on brain computed tomography or magnetic resonance imaging, 3) right handedness, and 4) no prior history of dementia, psychiatric illness, visuoperceptual disorder or neuromuscular disease involving the upper extremity.

Age and education matched healthy subjects with right hand dominance were enrolled for the control group. Subjects who had decreased Mini-Mental Status Examination score of lower than 1 standard deviation from age and education matched reference value were excluded from the control group [17]. Informed consent for the procedures was obtained from all participants, and the Health Service Human Research Ethics Committee and the Committee on Experimental Procedures Involving Human Subjects of the Korea University Medical Center approved this study (2013AN0040).

Assessment of language and visuoconstructive function

Handedness was evaluated using the Edinburg Handedness Inventory [18] in the aphasia and control groups. Language performance in the PWA was assessed by a Korean version of the Western Aphasia Battery (K-WAB) [19]. According to the aphasia quotient (AQ), aphasia severity is interpreted as follows: 76-above = mild, 51–75 = moderate, 26–50 = severe, 0–25 = very severe [19]. Our study consisted of 24 PWA: 10 mild, 2 moderate, 8 severe, and 4 very severe. Due to the small number of patients with moderate severity, we divided them into 3 sub-groups: 10 mild, 10 moderate to severe, and 4 very severe.

Visuoconstructive abilities were assessed by multiple drawing tasks and block designs in the aphasia and control groups. Visuoconstructive tasks were classified into 3 levels: simple (drawing objects), intermediate (clock drawing), and complex (block design and Rey-Osterrieth Complex Figure Test; RCFT) [6]. The drawings included a free drawing or copy of the simple, intermediate, and complex figures. Block construction was classified as a complex visuoconstructive task. Items of simple drawing were selected from the cognitive function subtest of the WAB. Among the 7 drawings of the WAB, 6 drawings (circle, square, face, tree, house, and cube) except clock were selected. The drawing procedure and scoring system were the same as the WAB and the maximum score was 22 points. If a patient did not understand the instructions, he or she drew after seeing a sample figure. The clock drawing test was determined to be a moderately difficult visuoconstructive task [6]. The subjects were made to draw a clock with a time of 10:15. Among the various scoring systems of the clock drawing test, we used the scoring procedure of the 7-Minute Neurocognitive Screening Battery [20] and the score ranged from 0 to 7 points. A copy subtest of the RCFT was selected as the most difficult level of a drawing task. The maximum score of the RCFT copy test was 36 points. A block design test of the WAB, which is a part of Wechsler Adult Intelligence Scale, was selected as a 3-dimensional visuoconstructive task, and the sum of scores was 9 points. Additionally, one of the non-verbal intelligence tests, Raven Colored Progressive Matrices (RCPM), was evaluated in both groups. RCPM consists of 3 set of 12 questions (total 36 items). The sum of raw score ranged from 0 to 36 points and 1 point can be added when the examinee completes the task within 5 minutes. Many PWA had right hemiplegia involving the upper limb; thus, the drawing was performed using a preferred hand in the aphasia group. We instructed the healthy control subjects to draw using their left hand to compensate for the disadvantage of the aphasia group who drew using their weakened right hand or non-dominant left hand.

The performance of each visuoconstructive tasks and RCPM was compared between 3 aphasia sub-groups (mild, moderate to severe and very severe group) and control groups by independent t-test, respectively. The correlation between visuoconstructional abilities and language performance assessed by the WAB was analyzed using Pearson’s correlation test. SPSS 24.0 software (IBM, Armonk, NY, USA) was used for statistical analysis. The significance level was set at p < 0.05.

RESULTS

Among the PWA, 15 patients were caused by cerebral infarction and 9 patients had intracerebral hemorrhage. The mean age of aphasia group was 59.3 ± 9.1 years and duration of education was 11.4 ± 3.4 years. Total thirty-two healthy subjects (10 males and 22 females) were enrolled for the control group and the mean age was 60.5 ± 10.1 years and mean years of education was 10.0 ± 3.2. There was no statistical difference between the 2 groups.

Based on the aphasia classification system of WAB, anomic aphasia was the most common (8 cases, 33.3%), followed by Broca’s aphasia (6 cases, 24.0%), Wernicke’s aphasia (13.0%), and global aphasia (13.0%). The mean AQ score was 56.9 ± 27.4 points; the details of the WAB results are presented in Table 1.

Table 1. Clinical characteristics of aphasia group.

No.	Sex	Age	Education (yr)	Lesion of stroke	Post-onset (mon)	Type of aphasia	AQ
1	M	40	16	Lt. BG ICH	14.1	Transcortical motor	61.2
2	M	71	16	Lt. MCA infarction	54.5	Anomic	79.4
3	M	50	14	Lt. BG ICH	3.7	Wernicke	45.4
4	F	56	6	Lt. BG ICH	7.0	Broca	45.6
5	M	59	12	Lt. BG ICH	12.5	Anomic	85.8
6	F	53	9	Lt. cerebral	7.7	Anomic	88.0
7	F	73	6	Lt. MCA PCA infarction	11.6	Anomic	89.0
8	M	46	16	Lt. BG ICH	47.9	Global	43.2
9	M	50	14	Lt. MCA infarction	6.8	Broca	62.0
10	F	52	8	Lt. MCA infarction	2.6	Anomic	85.0
11	M	66	12	Lt. MCA infarction	11.6	Broca	31.0
12	M	63	16	Lt. BG ICH	4.4	Anomic	92.8
13	F	56	9	Lt. BG ICH	1.7	Global	14.6
14	F	64	9	Lt. thalamic infarction	1.0	Anomic	84.2
15	M	70	12	Lt. MCA infarction	12.1	Wernicke	49.4
16	M	55	12	Lt. MCA infarction	6.0	Anomia	87.1
17	M	66	12	Lt. MCA infarction	1.7	Broca	29.0
18	F	77	9	Lt. MCA infarction	4.9	Broca	20.8
19	F	54	16	Lt. MCA infarction	22.1	Conduction	81.8
20	M	51	14	Lt. BG ICH	2.2	Conduction	77.8
21	M	59	6	Lt. temporoparietal ICH	35.2	Wernicke	11.4
22	M	63	12	Lt. MCA infarction	38.5	Transcortical motor	45.4
23	F	63	9	Lt. MCA infarction	8.6	Global	38.4
24	F	65	9	Lt. ICA infarction	21.1	Broca	17.6

AQ, aphasia quotient; BG, basal ganglia; ICA, internal carotid artery; ICH, intracerebral hemorrhage; MCA, middle cerebral artery; Lt., left; PCA, posterior cerebral artery.

The mean score of the Edinburgh Handedness Inventory was 9.6 points for both groups representing right hand dominance. The aphasia group earned low mean scores for the entire visuoconstructive tasks than the control group (Table 2).

Table 2. Comparison of demographic factor between the aphasia and control groups.

Tests		Aphasia (n = 24)	Control (n = 32)	p value
Age (yr)		59.3 ± 9.1	60.5 ± 10.1	0.644
Education (yr)		11.4 ± 3.4	10.0 ± 3.2	0.125
Edinburgh Handedness Inventory (10)		9.6 ± 1.0	9.6 ± 1.1	0.898
Language test (WAB)
	Spontaneous speech (20)	11.0 ± 5.9	-
	Auditory comprehension (10)	6.8 ± 2.4	-
	Repetition (10)	5.7 ± 3.5	-
	Naming (10)	5.5 ± 3.1	-
	Aphasia quotient (100)	56.9 ± 27.4	-

Data are mean ± standard deviation. Numbers in parenthesis are the maximum scores. Statistical test: independent t-test.

WAB, Western Aphasia Battery.

^*p < 0.05 was considered statistically significant.

The results of comparison between 3 aphasia sub-groups and the control group were as follows. The mean scores of visuoconstructive tests and RCPM according to aphasia sub-groups and control are presented in Table 3. There were no significant differences between the mild aphasia group and the control group for all visuoconstructive tasks and RCPM. Among the mild PWA, 8 cases in simple drawing, 8 cases in clock drawing, 4 cases in RCFT copies, and 8 cases in block test earned similar or higher scores than the means of the control group. In the mild to severe aphasia group the mean scores of the clock drawing, RCFT copy, and block design were significantly lower than those of the control group (p < 0.05), but there were no statistically significant differences between the 2 groups in the simple objects drawing and RCPM (p > 0.05) (Table 3). The very severe aphasia group showed significantly lower mean scores on all visuoconstructive tasks and RCPM compared to the control group.

Table 3. Visuoconstructive tests and RCPM scores according to aphasia sub-groups and control.

Tests	Mild group (n = 10)	Moderate to severe group (n = 10)	Very severe group (n = 4)	Control (n = 32)	t-statistic	p value
Drawing objects (22)	15.4 ± 2.1	13.6 ± 5.8	9.5 ± 3.1	15.4 ± 2.8	−0.953	0.362
Clock drawing (7)	5.7 ± 1.3	3.0 ± 2.4	1.0 ± 0.8	6.0 ± 1.1	−3.828	0.003*
RCFT copy (36)	20.6 ± 6.1	14.9 ± 6.8	8.4 ± 5.4	23.1 ± 6.0	−2.495	0.032*
Block design (9)	7.7 ± 1.5	5.3 ± 3.6	3.5 ± 2.6	8.2 ± 1.3	−3.673	0.001*
RCPM (37)	23.9 ± 5.2	21.9 ± 8.3	20.0 ± 7.8	25.8 ± 4.9	−1.396	0.190

Data are mean ± standard deviation. Numbers in parenthesis are the maximum scores. Statistical test: independent t-test. t-statistic, p value: comparison of visuoconstructive tests and RCPM between moderate to severe aphasia and control groups.

RCFT, Rey-Osterrieth Complex Figure Test; RCPM, Raven Colored Progressive Matrices.

^*p < 0.05.

Scores of visuoconstructive tasks, including clock drawing, RCFT copy, and block design, was significantly correlated with overall aphasia severity (AQ). Although a trend of a weak positive correlation was noted between the drawing objects and AQ scores, statistical significance level was not reached (r = 0.390, p = 0.060). Among the 4 domains of the language tests, the entire visuoconstructive task demonstrated a significant correlation with fluency and naming, but auditory comprehension and repetition revealed inconsistent results based on the visuoconstructive tests. The strongest correlation was noted in the clock drawing test and all the parameters of the WAB. The RCPM results did not correlate with any parameter of the language tests (Table 4, Fig. 1).

Table 4. Correlation between test scores and language performance in the aphasia group.

Tests	AQ	Fluency	Auditory comprehension	Repetition	Naming
Drawing objects	0.390 (0.060)	0.435* (0.034)	0.243 (0.253)	0.252 (0.236)	0.410* (0.047)
Clock drawing	0.695† (0.000)	0.722† (0.000)	0.488* (0.016)	0.506* (0.012)	0.722† (0.000)
RCFT copy	0.534† (0.007)	0.538† (0.007)	0.380 (0.067)	0.390 (0.059)	0.502† (0.003)
Block design	0.493* (0.014)	0.452* (0.027)	0.466* (0.022)	0.390 (0.059)	0.580* (0.012)
RCPM	0.239 (0.261)	0.288 (0.172)	0.025 (0.908)	0.268 (0.206)	0.174 (0.416)

Numbers in parenthesis are p value. Statistical test: Pearson’s correlation.

AQ, aphasia quotient; RCFT, Rey-Osterrieth Complex Figure Test; RCPM, Raven Colored Progressive Matrices.

^*p < 0.05; ^†p < 0.01.

Fig. 1. Scatterplot matrix of correlation.

AQ, aphasia quotient; RCFT, Rey-Osterrieth Complex Figure Test; RCPM, Raven Colored Progressive Matrices.

DISCUSSION

One of the common misconceptions is that visuospatial ability may be preserved in the left hemispheric lesion due to the extreme lateralization of this ability in the right hemisphere [21,22]. Visuoconstructive ability correlates with visuospatial perception and spatial manipulation, and this correlation is more evident in patients with right brain damage [10,11]. However, compared to the right hemispheric lesion, individuals with a left hemispheric lesion showed simplified drawing with an absence of detail, reduced in size, and mixed perspectives [23,24], and mainly the perceptuomotor or executive defect [11]. A coexistence of visuoconstructive deficits and aphasia is common and a lesion involving both the parietal and frontal structures was reported as a predictor of the poor visuoconstructional skills [25]. Our findings of the moderate to severe group showed similar results to previous studies [8,9,10] that entire visuoconstructive performances are significantly low in aphasia group than normal controls. In this study, 17%–34.8% of the PWA earned similar or higher scores in the different types of visuoconstructive tasks than the normal controls. These findings suggest a greater possibility of coexistence of the visuoconstructive deficits in PWA, but large individual variability do exist. Regarding the relationship between constructional ability and severity of aphasia, previous studies have shown mixed results [2,4,9,11,12,13,14,15,16]. In our study, the correlation between the score of each visuoconstructive task was moderately correlated with the scores of global aphasic severity (AQ) and subcomponents of language. However, the correlation was not evident in the simple drawing tasks and correlations that are more significant were noted in the intermediate to complex visuoconstructive tasks. Moreover, the scores of the nonverbal intelligence test by RCPM did not demonstrate a significant correlation with the severity of aphasia. Therefore, we cannot assume that visuoconstructive abilities are predicted solely by the impairment of language.

Aphasia can recover completely in some patients, but most aphasics remain with persistent language impairments and communication problems. During the past several decades, various approaches, including restorative and functional communication trainings, were applied to improve the language and communication skills [26,27]. Although restorative therapy may be the best approach for aphasics, functional communication training using alternative communication techniques is essential for severe aphasics. These include gesture, drawing, writing, picture board, and computer-based communication [28,29,30,31,32,33]. Drawing is one of the oldest human communication methods and it is relatively simple and easy to express one’s thoughts and ideas. Several studies have suggested that drawing is useful not only as a compensatory strategy for improving functional communication, but also as a facilitative tool to improve verbal output in PWA [24,34]. They reported that drawings, when combined with residual forms of language, contribute to improving verbal communication in PWA. Drawing promotes deeper levels of semantic processing by increasing attention to the structural and perceptual aspects of the object involved in right hemisphere activation. In most PWA, the right hemisphere is intact, and drawing can potentially be considered a non-verbal intervention to access semantic knowledge in the right brain [24,34]. Therefore, a thorough evaluation of visuoconsturctive function including drawing is important in terms of aphasia therapy and training according to the severity of aphasia.

In our results, simple drawing objects showed a comparable performance in the mild and moderate to severe PWA and healthy controls, and the correlation was not significant with the severity of language impairment. However, intermediate to complex visuoconstructive tasks showed higher correlations with language components. Among the multiple visuoconstructive tasks, the clock drawing test showed the highest correlation with language performance. Clock drawing has an intermediate level of difficulty and it is commonly used to evaluate cognitive function in patients with brain injury. Proper attention, working memory, visuoconstructive function, and executive functions, such as planning, are needed to complete the clock drawing successfully. In the authors’ view, the clock drawing may be a useful screening method to evaluate the visuoconstructive function in PWA.

In this study, we included various visuoconstructive tasks according to the level of complexity and recruited age matched controls to investigate this function thoroughly. We also tried to eliminate the disadvantage of the aphasia group while drawing with their non-dominant hand by making the control group draw with their non-dominant hand. A limitation of this study is the small sample size; thus, we did not compare the visuoconstructive abilities in the various subtypes of aphasia.

In summary, visuoconstructive abilities were significantly related to the performance of language in PWA. However, our results demonstrated that there is a large variance according to the severity of aphasia and the level of visuoconstructive tasks. Simple drawings were less affected in the mild and moderate to severe aphasic individuals, but the defect was more evident in the very severe group. Intermediate to complex visuoconstructive tasks were significantly affected in the moderate to severe and very severe aphasic individuals. Therefore, a thorough individual assessment of visuoconstructive function may provide useful information regarding the selection of drawing levels for the treatment and prognosis of aphasia.