Heuropsycholoqm, Vol. 28, No. Prmted in Great Bntain.
3, pp. 231-241.
1990 C
0028-393290 13.00+0.w 1990 Pcrgamon Press
plc
IMPAIRED VERBAL REASONING AND CONSTRUCTIONAL APRAXIA IN SUBJECTS WITH RIGHT HEMISPHERE DAMAGE LARRY I. BENOWITZ,*$$T[ KENNETH L. MOYA§
and DAVID N. LEVINE~~~
Departmentsof *Psychiatry and tNeurology and :Program in Neuroscience, §Mailman Research Center, Massachusetts General
McLean Hospital
Harvard Medical School; Hospital, Belmont, Massachusetts, U.S.A.; and liNeurology Service, and Spaulding Rehabilitation Hospital, Massachusetts, U.S.A.
(Received 20 March
1989; accepted
3 October 1989)
Abstract-In addition to causing visuospatial deficits, damage to the right cerebral hemisphere also impairs other cognitive abilities, including those requiring higher-order aspects of language. The present study used a standardized test battery to examine the relationship between visuospatial abilities and comprehension of narrative material in subjects having unilateral right hemisphere damage (RHD). In a series of 41 consecutively admitted RHD subjects, impairments in abstracting information from narrative passages were as prevalent and as severe in magnitude as constructional apraxia. Moreover, the extent of the visuospatial and linguistic impairments were highly correlated. Although age, educational levels, and degree of premorbid brain atrophy were all found to influence performance, analysis of a select subgroup of the population established that the covariation of visuospatial and verbal impairments is related to right hemisphere damage per se. Clinically, these findings may be of significance for understanding the pervasive cognitive impairments that are often evidenced by RHD patients.
INTRODUCTION THE INVOLVEMENT of the right cerebral hemisphere in understanding spatial organization has long been inferred from the striking impairments in drawing, copying, or recognizing spatial objects that result from damage to this side of the brain [7, 203. More subtle, but perhaps equally prevalent, are the defects right hemisphere damaged (RHD) subjects show in certain language-based functions. In contrast to subjects with left hemisphere damage, RHD patients seldom have difficulties with phonology, syntax, or semantics, and will carry on a conversation which at first glance seems normal. Nevertheless, impairments in less obvious faculties, including understanding the metaphoric or humorous use of language [3, 10,241, connotations of words [4], motivations of characters and logical interrelationships among events in a narrative passage [6,12,18,19,22], and in verbal problem-solving [S, 1l] have all been described. These observations raise the question of whether the visuospatial and verbal defects seen after RHD reflect disruption of two independent faculties that happen to coexist in the right cerebral hemisphere, or whether the two might instead be related to a more basic mechanism that is common to both cognitive domains. A previous study had suggested a significant correlation between the severity of RHD subjects’ visuospatial and verbal impairments [ 193. However, due to the limited size of the population, the unavailability of CT scans for some subjects, and the presence of cortical CCorrespondence to be addressed Belmont, MA 02178, U.S.A.
to: Dr Larry I. Benowitz,
231
Mailman
Research
Center,
315, McLean
Hospital,
232
LARRY I. BENOWITZ, KENNETH L. MOYA and DAVID N. LEVINE
atrophy in the relatively aged population used, the question remained unresolved as to whether the obseved covariation might have been caused artifactually by both visuospatial and linguistic capacities being exacerbated by such variables as overall brain damage, low educational levels, or advanced age. In the present study, we have sought to resolve this issue by expanding our population to include more than twice as many subjects, many of whom are younger, have higher levels of education, and have CT data. Our results indicate that the visuospatial and verbal reasoning defects after right-sided damage are strongly associated even after partialing out the confounding variables.
METHODS Subjects
Subjects included 44 stroke patients hospitalized in the Neurological Unit of the Spaulding Rehabilitation Hospital and 10 controls from the Volunteer Department of the Massachusetts General Hospital. The 44 patients included: 14 subjects (Ss) who had CT scans to verify the extent and location of the brain damage, and who met the criteria of being less than 70 years of age and having at least 10 years ofeducation (Group A. Table 1): 4 Ss of similar age and education without CT scans, but with the extent of damage inferred from the clinical examination (Group B); 11 Ss with CT verification of right cortical damage but not meeting the criteria of the select group. i.e. either their age was greater than 70 years (up to age 83) or their level of education was below high school. or both (Group C); 12 Ss with age > 70, or education 59 years, or both, and without CT scans (Group D); and 3 subjects with no rightsided damage evident by CT, but who showed left-sided motor impairments without cognitive defects (Group E). The latter group was inferred to have only lacunar lesions restricted to the internal capsule, and were behaviorally similar to the controls. All but one subject were tested at least 3 weeks after stroke, with the range being 18-540 days: the median time of testing was 43 days. The 10 age-matched, non-neurologically impaired subjects constituted Group F. Informed consent was obtained in all cases.
Table 1. Summary
Behavioral
of neurological
and control
Description
Group
N
A B
14 4
age < 70, education age < 70, education
C D
11 12
age > 70 or education age > 70 or education
E F
3 10
neurological controls non-neurological controls
> 9, CT > 9. no CT < 9. CT < 9, no CT
groups
included
in study
Age
Education
54.lk2.7 59.3 * 5.7
14.6+0.8 12.5+ 1.0
65.1 +6.0 73.3k3.6
9.1_+1.0 10.2* 1.5
53.7k6.2 66.3 & 2.2
14.0+ 1.15 14.7kO.5
tests
Subjects were given a standardized test battery that included (a) recall of material based upon 5 brief narrative passages (Verbal Recall test), (b) copying figures and drawing from memory (Visuospatial test), and (c)forward and reverse digit span. These tests have been described in some detail 1191 and will be summarized here only briefly. Verbal recall battery. Five brief narrative passages (51-71 words in length) were read aloud slowly to each subject. Immediately after hearing each passage, the S was asked to recall as much of the story as possible. The passage was then read a second time and a series of questions were asked probing (a) memory for specific details (e.g. what time of day did the events occur? What was the name of a character in the story?). or (b) an understanding of more abstract qualities, including relationships among events or characters, the ability to draw appropriate inferences based upon the information given (e.g. what motivated a character to do something?), or to evaluate the overall theme of the story (selecting an appropriate title for the passage from among 5 alternatives). Errors, including confabulations, reduplication of items from previous stories, and elements of personal reference, were scored as well. Verbatim transcripts were prepared from tape recordings and were scored afterwards. Scores for the total number of details recalled in the 5 passages varied over a 5-fold range, while scores for the total number of abstractions varied over a IO-fold range. Visuosparial test. Each S was asked to copy and draw from memory a series offigures, including a slightly modified Rey--0sterreith diagram and cubes oriented rightwards and leftwards: to till in either the missing vertices of a cube
COGNITIVE
DEFICITS
AFTER
RIGHT
HEMlSPHERE
233
DAMAGE
(when shown segments of the sides only) or the missing sides (when shown the vertices only); and to draw from memory the face of a clock (showing the hands pointing to 3:45), a pear. a teacup, and a flower. Using the standardized protocol described previously [19], figures were evaluated for the number of specific details included. the organization of the figures (e.g. appreciating the relationships among elements of the ReyyOsterreith figure, closure, perspective), and for left spatial neglect (number of details drawn correctly on the left side of the figures normalized by the total number of details rendered correctly). Errors. including perseverations or insertion of extraneous components, were scored as well.
r
1
f-
-f
L
L
I
COPY
A
FILL
A CLOCK
DRAW
A
FLOWER
IN
-I
f L
DRAW
r
L
J
B
Fig. I, Examples of RHD subjects’ performance on visuospatial tasks. Left: Subjects were asked to copy the modified Rey-Osterrieth diagram shown at the top. Subject A showed severe left-sided neglect. but drew the right-most side accurately; B drew two copies that contained some correct details but were oriented incorrectly and missing much of the outline; c’s drawing was essentially correct, but also missed some of the outline. Left-center: subjects were shown the vertices ofa cube and asked to fill in the sides. A again showed left-sided neglect: the other two subjects drew in many interal details but failed to appreciate the overall organization of the figure. Right: when asked to draw a clock with the hands at 3:45 and to draw a flower. A showed considerable disorganization and left-sided neglect; B’s drawing contained many correct elements, but the clock was poorly organized; C drew both well.
Digir .sp~~rt.Forward
and reverse digit span were evaluated
in a standard
manner
[23].
CT scans suitable for mapping the extent of right brain damage were available from 26 Ss, as summarized in Table 2. To assess the location of the lesion. the hemisphere was divided into I2 non-overlapping areas. Within each area. the extent of the damage was rated either present or absent. The size of the lesion was described as the total number of areas affected. and ranged from 0 to Il. Areas most frequently damaged included the pre- and postcentral gyri. the parteto-temporal convexity, and the fronto-temporo-insular region. General cortical atrophy. over and above the focal damage. was rated according to techniques described by LISHMAN [16]. A qualitative index of sulcal widening was scored on a &5 scale (0 = none, 2 = moderate, 3 or more =extensive atrophy). The measures of atrophy were made on the left hemisphere because of possible distortions of the right hemisphere from the stroke.
234
LARRY I. BENOWITZ. KENNETH L. MOYA and DAVID N. LEVINE Table 2. Samples
Original Jbhni
story (numbered
segments
of immediate
are basis for scoring
individual
drove o:t of town/ on a stArmy day.,/ Rain cl&ds,
hand side.: The :iver,/ which w& on his right.,’ began side-vi&
mirror!
John no&ed
a car: skidding
he saw only algbuard rail. separating
recall for Story 4
obscured
rising
tb3wards him.
details) the c&y’s skyline,,
and soon overflo;ed As he swer:;d
which was on his Igft
onto the road.’
In his
to avoid it; his hea:? leapt
as
hj& from the river.
Sample recalls: A. Jbhn was,’ trying
to dri:e
out of town.
B. J&hn,’ drove out ti a storm.: C. Jbhnldrove
’ There was a sivere storm.,’
Ah, some began, rain.:
Ri;ers’
o:t of town;’ on a ra?ny day.; He was presumably
was on’the right,;
and the river overflgwed
came fzller up. (Intrusion) going west on Storrow Drive..
The ryver.
onto the road.
Data analysis
Test results werescored independently by two investigators who were blind to Ss’identity. extent ofbrain damage, and peformance on other test items. Inter-rater reliability scores were all greater than 0.90. and most were higher than 0.95. The principal analyses included comparing average scores of the RHD and control groups on the different visuospatial and verbal performance categories (l-tailed f-tests), correlating visuospatial and story recall scores (Pearson product&moment correlations), and correlating behavioral test scores with the extent of brain damage where CT scans were available (Pearson product-moment correlations). Partial correlations were done to eliminate the effects of age. education. total lesion size. and brain atrophy on the correlations between visuospatial and verbal abilities.
RESULTS Sample story recalls for selected right hemisphere damaged subjects are shown in Table 2. Test scores for all subjects having CT scans are shown in Table 3: Ss 1-14 comprise the select group who were below 70 years of age and had 10 or more years of education (Group A); 15-25 represent Group C, who were either 70 years of age or older, or who had 9 years or less of education (or both); S 26 had no evidence of cortical damage. Average scores for all subject groups are shown in Table 4 (mean? SEM). The 3 subjects with left-sided motor impairments but no evidence of right cortical damage (Group E) were nearly identical in all respects to the normal subjects (Group F) and these two groups were therefore combined for statistical purposes to give a somewhat enlarged group of normal and non-cortically damaged neurological controls (E + F. Table 4). Groups C and D, composed of subjects over the age of 70 or having I 9 years ofeducation (with and without CT scans, respectively) were similar to each other in behavioral test scores, and were combined to form a right hemisphere damaged group (RHD II) which had a number of complications that made direct comparison with the controls difficult (e.g. lower education and a possibly higher degree of cortical atrophy). However, Groups A and B (age ~70, education 2 10 years, with and without CT scans, respectively). when combined to form the RHD I Group, were well matched to the controls with respect to education and age. As shown in Fig. 2, the RHD I group showed striking impairments in all visuospatial and verbal recall scores (P
3, pp. 231-241.
1990 C
0028-393290 13.00+0.w 1990 Pcrgamon Press
plc
IMPAIRED VERBAL REASONING AND CONSTRUCTIONAL APRAXIA IN SUBJECTS WITH RIGHT HEMISPHERE DAMAGE LARRY I. BENOWITZ,*$$T[ KENNETH L. MOYA§
and DAVID N. LEVINE~~~
Departmentsof *Psychiatry and tNeurology and :Program in Neuroscience, §Mailman Research Center, Massachusetts General
McLean Hospital
Harvard Medical School; Hospital, Belmont, Massachusetts, U.S.A.; and liNeurology Service, and Spaulding Rehabilitation Hospital, Massachusetts, U.S.A.
(Received 20 March
1989; accepted
3 October 1989)
Abstract-In addition to causing visuospatial deficits, damage to the right cerebral hemisphere also impairs other cognitive abilities, including those requiring higher-order aspects of language. The present study used a standardized test battery to examine the relationship between visuospatial abilities and comprehension of narrative material in subjects having unilateral right hemisphere damage (RHD). In a series of 41 consecutively admitted RHD subjects, impairments in abstracting information from narrative passages were as prevalent and as severe in magnitude as constructional apraxia. Moreover, the extent of the visuospatial and linguistic impairments were highly correlated. Although age, educational levels, and degree of premorbid brain atrophy were all found to influence performance, analysis of a select subgroup of the population established that the covariation of visuospatial and verbal impairments is related to right hemisphere damage per se. Clinically, these findings may be of significance for understanding the pervasive cognitive impairments that are often evidenced by RHD patients.
INTRODUCTION THE INVOLVEMENT of the right cerebral hemisphere in understanding spatial organization has long been inferred from the striking impairments in drawing, copying, or recognizing spatial objects that result from damage to this side of the brain [7, 203. More subtle, but perhaps equally prevalent, are the defects right hemisphere damaged (RHD) subjects show in certain language-based functions. In contrast to subjects with left hemisphere damage, RHD patients seldom have difficulties with phonology, syntax, or semantics, and will carry on a conversation which at first glance seems normal. Nevertheless, impairments in less obvious faculties, including understanding the metaphoric or humorous use of language [3, 10,241, connotations of words [4], motivations of characters and logical interrelationships among events in a narrative passage [6,12,18,19,22], and in verbal problem-solving [S, 1l] have all been described. These observations raise the question of whether the visuospatial and verbal defects seen after RHD reflect disruption of two independent faculties that happen to coexist in the right cerebral hemisphere, or whether the two might instead be related to a more basic mechanism that is common to both cognitive domains. A previous study had suggested a significant correlation between the severity of RHD subjects’ visuospatial and verbal impairments [ 193. However, due to the limited size of the population, the unavailability of CT scans for some subjects, and the presence of cortical CCorrespondence to be addressed Belmont, MA 02178, U.S.A.
to: Dr Larry I. Benowitz,
231
Mailman
Research
Center,
315, McLean
Hospital,
232
LARRY I. BENOWITZ, KENNETH L. MOYA and DAVID N. LEVINE
atrophy in the relatively aged population used, the question remained unresolved as to whether the obseved covariation might have been caused artifactually by both visuospatial and linguistic capacities being exacerbated by such variables as overall brain damage, low educational levels, or advanced age. In the present study, we have sought to resolve this issue by expanding our population to include more than twice as many subjects, many of whom are younger, have higher levels of education, and have CT data. Our results indicate that the visuospatial and verbal reasoning defects after right-sided damage are strongly associated even after partialing out the confounding variables.
METHODS Subjects
Subjects included 44 stroke patients hospitalized in the Neurological Unit of the Spaulding Rehabilitation Hospital and 10 controls from the Volunteer Department of the Massachusetts General Hospital. The 44 patients included: 14 subjects (Ss) who had CT scans to verify the extent and location of the brain damage, and who met the criteria of being less than 70 years of age and having at least 10 years ofeducation (Group A. Table 1): 4 Ss of similar age and education without CT scans, but with the extent of damage inferred from the clinical examination (Group B); 11 Ss with CT verification of right cortical damage but not meeting the criteria of the select group. i.e. either their age was greater than 70 years (up to age 83) or their level of education was below high school. or both (Group C); 12 Ss with age > 70, or education 59 years, or both, and without CT scans (Group D); and 3 subjects with no rightsided damage evident by CT, but who showed left-sided motor impairments without cognitive defects (Group E). The latter group was inferred to have only lacunar lesions restricted to the internal capsule, and were behaviorally similar to the controls. All but one subject were tested at least 3 weeks after stroke, with the range being 18-540 days: the median time of testing was 43 days. The 10 age-matched, non-neurologically impaired subjects constituted Group F. Informed consent was obtained in all cases.
Table 1. Summary
Behavioral
of neurological
and control
Description
Group
N
A B
14 4
age < 70, education age < 70, education
C D
11 12
age > 70 or education age > 70 or education
E F
3 10
neurological controls non-neurological controls
> 9, CT > 9. no CT < 9. CT < 9, no CT
groups
included
in study
Age
Education
54.lk2.7 59.3 * 5.7
14.6+0.8 12.5+ 1.0
65.1 +6.0 73.3k3.6
9.1_+1.0 10.2* 1.5
53.7k6.2 66.3 & 2.2
14.0+ 1.15 14.7kO.5
tests
Subjects were given a standardized test battery that included (a) recall of material based upon 5 brief narrative passages (Verbal Recall test), (b) copying figures and drawing from memory (Visuospatial test), and (c)forward and reverse digit span. These tests have been described in some detail 1191 and will be summarized here only briefly. Verbal recall battery. Five brief narrative passages (51-71 words in length) were read aloud slowly to each subject. Immediately after hearing each passage, the S was asked to recall as much of the story as possible. The passage was then read a second time and a series of questions were asked probing (a) memory for specific details (e.g. what time of day did the events occur? What was the name of a character in the story?). or (b) an understanding of more abstract qualities, including relationships among events or characters, the ability to draw appropriate inferences based upon the information given (e.g. what motivated a character to do something?), or to evaluate the overall theme of the story (selecting an appropriate title for the passage from among 5 alternatives). Errors, including confabulations, reduplication of items from previous stories, and elements of personal reference, were scored as well. Verbatim transcripts were prepared from tape recordings and were scored afterwards. Scores for the total number of details recalled in the 5 passages varied over a 5-fold range, while scores for the total number of abstractions varied over a IO-fold range. Visuosparial test. Each S was asked to copy and draw from memory a series offigures, including a slightly modified Rey--0sterreith diagram and cubes oriented rightwards and leftwards: to till in either the missing vertices of a cube
COGNITIVE
DEFICITS
AFTER
RIGHT
HEMlSPHERE
233
DAMAGE
(when shown segments of the sides only) or the missing sides (when shown the vertices only); and to draw from memory the face of a clock (showing the hands pointing to 3:45), a pear. a teacup, and a flower. Using the standardized protocol described previously [19], figures were evaluated for the number of specific details included. the organization of the figures (e.g. appreciating the relationships among elements of the ReyyOsterreith figure, closure, perspective), and for left spatial neglect (number of details drawn correctly on the left side of the figures normalized by the total number of details rendered correctly). Errors. including perseverations or insertion of extraneous components, were scored as well.
r
1
f-
-f
L
L
I
COPY
A
FILL
A CLOCK
DRAW
A
FLOWER
IN
-I
f L
DRAW
r
L
J
B
Fig. I, Examples of RHD subjects’ performance on visuospatial tasks. Left: Subjects were asked to copy the modified Rey-Osterrieth diagram shown at the top. Subject A showed severe left-sided neglect. but drew the right-most side accurately; B drew two copies that contained some correct details but were oriented incorrectly and missing much of the outline; c’s drawing was essentially correct, but also missed some of the outline. Left-center: subjects were shown the vertices ofa cube and asked to fill in the sides. A again showed left-sided neglect: the other two subjects drew in many interal details but failed to appreciate the overall organization of the figure. Right: when asked to draw a clock with the hands at 3:45 and to draw a flower. A showed considerable disorganization and left-sided neglect; B’s drawing contained many correct elements, but the clock was poorly organized; C drew both well.
Digir .sp~~rt.Forward
and reverse digit span were evaluated
in a standard
manner
[23].
CT scans suitable for mapping the extent of right brain damage were available from 26 Ss, as summarized in Table 2. To assess the location of the lesion. the hemisphere was divided into I2 non-overlapping areas. Within each area. the extent of the damage was rated either present or absent. The size of the lesion was described as the total number of areas affected. and ranged from 0 to Il. Areas most frequently damaged included the pre- and postcentral gyri. the parteto-temporal convexity, and the fronto-temporo-insular region. General cortical atrophy. over and above the focal damage. was rated according to techniques described by LISHMAN [16]. A qualitative index of sulcal widening was scored on a &5 scale (0 = none, 2 = moderate, 3 or more =extensive atrophy). The measures of atrophy were made on the left hemisphere because of possible distortions of the right hemisphere from the stroke.
234
LARRY I. BENOWITZ. KENNETH L. MOYA and DAVID N. LEVINE Table 2. Samples
Original Jbhni
story (numbered
segments
of immediate
are basis for scoring
individual
drove o:t of town/ on a stArmy day.,/ Rain cl&ds,
hand side.: The :iver,/ which w& on his right.,’ began side-vi&
mirror!
John no&ed
a car: skidding
he saw only algbuard rail. separating
recall for Story 4
obscured
rising
tb3wards him.
details) the c&y’s skyline,,
and soon overflo;ed As he swer:;d
which was on his Igft
onto the road.’
In his
to avoid it; his hea:? leapt
as
hj& from the river.
Sample recalls: A. Jbhn was,’ trying
to dri:e
out of town.
B. J&hn,’ drove out ti a storm.: C. Jbhnldrove
’ There was a sivere storm.,’
Ah, some began, rain.:
Ri;ers’
o:t of town;’ on a ra?ny day.; He was presumably
was on’the right,;
and the river overflgwed
came fzller up. (Intrusion) going west on Storrow Drive..
The ryver.
onto the road.
Data analysis
Test results werescored independently by two investigators who were blind to Ss’identity. extent ofbrain damage, and peformance on other test items. Inter-rater reliability scores were all greater than 0.90. and most were higher than 0.95. The principal analyses included comparing average scores of the RHD and control groups on the different visuospatial and verbal performance categories (l-tailed f-tests), correlating visuospatial and story recall scores (Pearson product&moment correlations), and correlating behavioral test scores with the extent of brain damage where CT scans were available (Pearson product-moment correlations). Partial correlations were done to eliminate the effects of age. education. total lesion size. and brain atrophy on the correlations between visuospatial and verbal abilities.
RESULTS Sample story recalls for selected right hemisphere damaged subjects are shown in Table 2. Test scores for all subjects having CT scans are shown in Table 3: Ss 1-14 comprise the select group who were below 70 years of age and had 10 or more years of education (Group A); 15-25 represent Group C, who were either 70 years of age or older, or who had 9 years or less of education (or both); S 26 had no evidence of cortical damage. Average scores for all subject groups are shown in Table 4 (mean? SEM). The 3 subjects with left-sided motor impairments but no evidence of right cortical damage (Group E) were nearly identical in all respects to the normal subjects (Group F) and these two groups were therefore combined for statistical purposes to give a somewhat enlarged group of normal and non-cortically damaged neurological controls (E + F. Table 4). Groups C and D, composed of subjects over the age of 70 or having I 9 years ofeducation (with and without CT scans, respectively) were similar to each other in behavioral test scores, and were combined to form a right hemisphere damaged group (RHD II) which had a number of complications that made direct comparison with the controls difficult (e.g. lower education and a possibly higher degree of cortical atrophy). However, Groups A and B (age ~70, education 2 10 years, with and without CT scans, respectively). when combined to form the RHD I Group, were well matched to the controls with respect to education and age. As shown in Fig. 2, the RHD I group showed striking impairments in all visuospatial and verbal recall scores (P
