Brain and Cognition 72 (2010) 483–490 Contents lists available at ScienceDirect Brain and Cognition journal homepage: www.elsevier.com/locate/b&c Associations and dissociations of transitive and intransitive gestures in left and right hemisphere stroke patients Vessela Stamenova a,*, Eric A. Roy a,b,c, Sandra E. Black a,b,c,d a Graduate Department of Rehabilitation Science, University of Toronto, Canada Psychology/Kinesiology, University of Waterloo, Canada c Heart and Stroke Foundation Centre for Stroke Recovery (HSFCSR), Department of Medicine, Sunnybrook Health Sciences Centre, University of Toronto, Canada d Division of Neurology, Department of Medicine, University of Toronto, Canada b a r t i c l e i n f o Article history: Accepted 20 January 2010 Available online 18 February 2010 Keywords: Limb apraxia Stroke Hemisphere asymmetries Imitation Pantomime Transitive gestures Intransitive gestures Recovery Skilled movement a b s t r a c t The study investigated performance on pantomime and imitation of transitive and intransitive gestures in 80 stroke patients, 42 with left (LHD) and 38 with right (RHD) hemisphere damage. Patients were also categorized in two groups based on the time that has elapsed between their stroke and the apraxia assessment: acute–subacute (n = 42) and chronic (n = 38). In addition, patterns of performance in apraxia were examined. We expected that acute–subacute patients would be more impaired than chronic patients and that LHD patients would be more impaired than RHD patients, relative to controls. The hemisphere prediction was confirmed, replicating previous findings. The frequency of apraxia was also higher in all LHD time post-stroke groups. The most common impairment after LHD was impairment in both pantomime and imitation in both transitive and intransitive gestures. Selective deficits in imitation were more frequent after RHD for transitive gestures but for intransitive gestures they were more frequent after LHD. Patients were more impaired on imitation than pantomime, relative to controls. In addition, after looking at both gesture types concurrently, we have described cases of patients who suffered deficits in pantomime of intransitive gestures with preserved performance on transitive gestures. Such cases show that the right hemisphere may be in some cases critical for the successful pantomime of intransitive gestures and the neural networks subserving them may be distinct. Chronic patients were also less impaired than acute–subacute patients, even though the difference did not reach significance. A longitudinal study is needed to examine the recovery patterns in both LHD and RHD patients. Ó 2010 Elsevier Inc. All rights reserved. 1. Introduction Limb apraxia is a movement disorder that expresses itself by an inability to perform purposeful movements such as using everyday tools (called transitive gestures) and/or making meaningful gestures not involving tools (called intransitive gestures). Unlike other motor control disorders, limb apraxia is not caused by a physical disability; rather, it results from higher-order cognitive disruptions of the nervous system. It is often defined by exclusion: apraxia is not caused by muscle weakness, paralysis, dystonia, tremor, chorea, myoclonus or defects of sensory feedback. It is also not caused by cognitive deficits such as aphasia, agnosia or inattention (Heilman & Rothi, 1993). Limb apraxia is typically assessed by asking a patient to pantomime (perform a gesture from memory to a verbal command) or to imitate a visually-presented gesture. Therefore, limb apraxia can also be defined as the inability to pantomime or * Corresponding author. Address: 217 Markland Cres, Nepean, ON, Canada K2G 5Z9. E-mail address: viussi@gmail.com (V. Stamenova). 0278-2626/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.bandc.2010.01.004 imitate gestures (Roy, 1996). According to Roy’s model of apraxia (Roy, 1996) three systems are involved in the control of movement: a sensory/perceptual system processing information from the environment, a conceptual system stores our knowledge of gestures and tools, while the production system is responsible for response selection and control of movement. Both pantomime and imitation are dependent on the patient’s preserved ability to organize and plan movements (i.e. they require an intact production system), in order to successfully perform the gesture. Pantomime, however, is also dependent on the patient’s preserved knowledge of tools and actions (i.e. it requires an intact conceptual system). When a patient is given a verbal instruction, he/she must retrieve from memory what a particular tool looks like and what its function is and link this with the representation of the action associated with this tool. Imitation, on the other hand, is dependent on the preserved ability to process the visual information in the gesture performed by the examiner and to translate this information into a movement. Here a patient does not need to know what the gesture means in order to be able to perform the movement, although meaningful gestures are often imitated more accurately 484 V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 due to the support afforded through semantics (Rumiati & Tessari, 2002). Limb apraxia is often observed in patients who have suffered a stroke. A review of studies examining the prevalence of limb apraxia after stroke reported a prevalence of apraxia after left hemisphere damage (LHD) ranging from 28% to 57% (median = 45%), whereas the prevalence after right hemisphere damage (RHD) ranged from 0% to 34% (median = 8%) (Donkervoort, Dekker, van den Ende, Stehmann-Saris, & Deelman, 2000). A series of studies conducted by Roy and colleagues compared the performance of transitive and intransitive gestures in pantomime and imitation in LHD and RHD stroke participants. First, (Roy et al., 2000) showed that deficits in pantomime alone or imitation alone were equally common after damage to either hemisphere, but that deficits in pantomime and imitation together were more common after LHD. Later, a similar study of intransitive gestures reported that imitation alone, pantomime alone, as well as, deficits in both pantomime and imitation were equally likely after damage to either hemisphere (Heath, Roy, Westwood, & Black, 2001). The findings from both studies were at odds with many previous reports showing a greater role of the left hemisphere in the control of movement (Haaland, Harrington, & Knight, 2000; Hanna-Pladdy et al., 2001). Because the studies of Roy and colleagues examined separately transitive and intransitive gestures, we wanted to directly compare the performance of the two gesture types within the same sample. In stroke, accuracy in the performance of transitive gestures is often lower than that of intransitive gestures (Gonzalez-Rothi, Mack, Verfaellie, Brown, & Heilman, 1988; Haaland & Flaherty, 1984; Haaland et al., 2000; Schnider, Hanlon, Alexander, & Benson, 1997). Transitive gestures are also performed with lower accuracy in normal control participants and thus it has been suggested that they are performed at lower accuracy, because they are more difficult than intransitive gestures (Carmo & Rumiati, 2009; Mozaz, Rothi, Anderson, Crucian, & Heilman, 2002). In addition, it has been argued that there may be hemisphere differences when it comes to transitive vs. intransitive gesture execution. For example, some authors have suggested that the left hemisphere may control transitive gestures while both hemispheres may be involved in the control of intransitive gestures (Buxbaum, Kyle, Grossman, & Coslett, 2007; Haaland & Flaherty, 1984; Mozaz et al., 2002; Rapcsak, Ochipa, Beeson, & Rubens, 1993). More recent evidence from neuroimaging studies, however, indicates that for both transitive and intransitive gesture execution activates a common left hemisphere network involving frontal, parietal and temporal regions (Kroliczak & Frey, 2009). Given the contradictory evidence, together with the relatively few studies examining performance of both transitive and intransitive gestures in both pantomime and imitation tasks, we considered it important to directly compare the performance of both pantomime and imitation in transitive and intransitive gestures, to examine performance differences in both hemisphere groups and to report the frequencies of the various patterns of deficits in each group of patients for both gesture types. Given the greater role of the left hemisphere in the performance of pantomime as opposed to imitation and of transitive gestures as opposed to intransitive, we predicted that pantomime tasks and transitive gestures would show greater impairment in the LHD patients. We also hypothesized that performance on transitive gestures would be less accurate than intransitive gestures and that pantomime would be less accurate than imitation performance. Finally, based on Roy et al. (2000) and Heath et al. (2001), we also predicted that patterns of deficits with selective imitation in pantomime or imitation, would be equally likely after LHD or RHD, but the pattern with deficits in both pantomime and imitation will be more common after LHD. 2. Methods 2.1. Participants Eighty right-handed participants with a single unilateral hemispheric stroke (confirmed through both clinical examination and neuroradiology report), 35 women and 45 men, 42 LHD and 38 RHD, were included in the study with a mean age of 64 (SD = 13.9) years. In order to provide a more detailed overview of the lesion of the patients all patients’ charts were reviewed. According to records, 17 patients had suffered from a left, while 21 had suffered from right MCA stroke. Two patients had suffered from left and four from right Posterior Cerebral Artery Stroke. Two patients had suffered from left while one from right Anterior Cerebral Artery Stroke. Nine patients had suffered from left and one from right strokes restricted to subcortical regions (such as external capsule, corona radiata, basal ganglia strokes or thalamus). Finally, nine patients had suffered from left and nine from right ‘cortical’ strokes, but the specific artery affected was not available in their charts and thus they were put in a separate ‘cortical’ category. In that last category, among the LHD, three had occipitoparietal stroke, one patient had parietal, one occipital, two basal ganglia and frontal strokes and one posterotemporal and angular gyrus stroke. Among the RHD patients with cortical strokes, four patients had frontal strokes, two had frontotemporal and two had frontal + basal ganglia stroke and one had frontoparietal stroke. We could not find details about the specific locations of their stroke (aside from hemisphere affected) in three left and one right patient, because their charts were no longer available at the time of the retrospective review. Patients were recruited from Sunnybrook Health Sciences Centre in Toronto, Ontario, Canada. Consent to participate in the study was obtained from all participants and the study was approved by the Research Ethics Board at Sunnybrook Health Sciences Centre and at the University of Waterloo. The sample consisted of patients assessed at different stages post-stroke and given past research has suggested that apraxia recovers somewhat over the first 3 months post-stroke (Basso, Burgio, Paulin, & Prandoni, 2000; Foundas, Raymer, Maher, GonzalezRothi, & Heilman, 1993), we expected that chronic patients may perform better than acute patients. Thus, we also categorized patients based on the time elapsed between their stroke and the apraxia assessment. Patients assessed within 3 months post-stroke were categorized as ‘‘acute–subacute”, while patients assessed over 3 months post-stroke were categorized as ‘‘chronic”. This resulted in 42 patients categorized as acute–subacute and 38 patients categorized as chronic. See Table 1 for a summary of the number of patients per group and a summary of the age, years of education, MMSE scores and days since stroke onset for each group. In the acute–subacute group, the time since stroke onset ranged from 3 to 84 days, while in the chronic group the time since onset ranged from 103 to 5753 days. Western Aphasia Battery scores were available for only 23 of the LHD patients, whose mean WAB Aphasia Quotient was 72 (SD = 25). Comprehension WAB scores were available on 23 LHD patients and the average score was 8.3 (SD = 1.5), which indicates good comprehension level (Kertesz & Poole, 1974). Unfortunately, the WAB data was not collected consistently and, therefore, we cannot be sure that the patients who were not tested did not have any comprehension deficits. However, all patients included in the study were at a language comprehension level that allowed them to understand verbal instructions and their responses on the pantomime tasks indicated that they understood what they were asked to do. Patients were excluded from the study if they presented with multiple strokes, had a history of any other neurological disorders, peripheral motor disorders or disorders that could affect their ability to perform gestures such as severe arthritis. 485 V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 Table 1 Sample characteristics per patient group. Acute–subacute Mean (SD) Age Years of education MMSE Days since stroke Chronic Mean (SD) LHD RHD LHD RHD 64 14 20 17 65 13 26 23 67 (9) 14 (3) 25 (6) 1679 (1811) 59 (15) 15 (3) 27 (2) 1697 (1583) (15) (3) (8) (16) (15) (3) (7) (19) In addition, performance of 27 age-matched [mean age = 67.3, SD = 8.7, t(1, 80) = .59, p = .58] normal control community volunteers, with no history of neurological diseases, were assessed on all tasks with each hand. 2.2. Gestural tasks and performance scoring Patients were asked to perform four tasks: Pantomime to Verbal Command of transitive gestures, Pantomime to Verbal Command of intransitive gestures, Concurrent Imitation of transitive gestures and Concurrent Imitation of intransitive gestures. The pantomime conditions were always performed first in order to avoid giving any cues as to how the gesture was performed. Transitive and intransitive gestures were performed in separate blocks. The transitive gestures included a comb, spatula, hammer, fork, knife, watering can, toothbrush and tweezers. The intransitive gestures included waving good-bye, saluting, making okay sign, putting cream on one’s face, beckoning, holding one’s nose as if there were a bad smell, making the okay sign, scratching one’s ear and hailing a cab. From here on the Pantomime to Verbal Command tasks will be referred to as simply Pantomime, while the Concurrent Imitation tasks will be referred to as Imitation. In pantomime, the patients were given a verbal instruction to perform gestures. For transitive gestures, they were asked, for example: ‘‘Show me how you would use a hammer to pound a nail.” The patient was instructed to pretend to hold the object in their hand and to perform the gesture. For intransitive gestures, patients were again given a verbal instruction and asked, for example, ‘‘Show me how you would wave good-bye.” In imitation, the patient imitated the gesture presented by the examiner. The examiner continued the gesture presentation until the patient performed the imitation. The patients were videotaped while performing the gesture and were scored on five performance dimensions: location, posture, action, plane and orientation. Location referred to the location in space of the arm relative to the body. Posture was the hand posture of the participant. Action referred to the movement characteristics of the gesture. Orientation was the orientation of the palm. Finally, plane referred to the plane of movement of the arm in threedimensional space. Each dimension was scored on a 3-point scale: 2 (correct), 1 (distorted) and 0 (incorrect). Performance on each dimension was then expressed by calculating the percentage of the maximum score achieved across the eight gestures. A composite score for each task was calculated by taking the average of the percentage scores of the five dimensions. Performance was scored using procedures with high interrater reliability (Roy, Black, Blair, & Dimeck, 1998). Control Mean (SD) F-value p-Value 67 (9) 15 (3) 29 (2) – 1.2 1.02 5.2 13.6 0.3 0.39 <.01 <.001 hand performance of the control group. Performance for each stroke patient was converted to a Z-score with reference to the mean and standard deviation of the performance of the controls (see Fig. 2 for a summary of the performance mean accuracy of control participants relative to patients on each task). These Z scores allow us to compare performance of the two hemisphere groups taking into account each group’s performance relative to the controls. These Z-scores were used in two analyses. One was an ANOVA comparing performance among the four groups of patients, LHD and RHD stroke in each of two chronicity (acute–subacute vs. chronic) groups, while the other compared the frequency of apraxia among the four groups of patients. In this analysis, Zscores falling 2 SDs below the mean of the controls were considered to be within the impaired or apraxic range. Z-scores above 2 SDs below the mean of the controls were categorized as nonapraxic. 4. Results 4.1. Sample characteristics In order to rule out any differences in age and education among the patient groups, a 2 (chronicity)  2 (hemisphere) MANOVA was run to compare the patients. No main effects or interactions were observed. In addition, in order to compare the patient groups relative to controls, an ANOVA comparison between the four patient subgroups and the controls revealed no group differences on age or education (see Table 1). In addition, a 2 (chronicity)  2 (hemisphere) MANOVA was run to compare the patients’ performance on the MMSE (only 19 LHD acute, 15 RHD acute, 15 LHD chronic and 19 RHD chronic had MMSE data). The analysis showed a significant main effect of chronicity F(3, 64) = 5.3, p < .05 showing acute– subacute patients obtained significantly lower MMSE scores (Mean = 21, SD = 9) than chronic patients (Mean = 26, SD = 4). In addition, a main effect of hemisphere showed that LHD patients performed significantly worse (Mean = 21, SD = 9) than RHD patients (Mean = 26, SD = 5) on MMSE, [F(1, 64) = 6.5, p < .05)]. There was no interaction between the hemisphere and the chronicity factor. While the MMSE was included as a general cognitive measure indicator, the reader should be aware that the lower scores in the LHD group could also be caused by language deficits. As noted above, the overall comprehension level of the LHD patients who had undergone WAB assessments was relatively preserved, but WAB data was not available on all LHD patients. Given the involvement of the left hemisphere in language, it is possible that language deficits contributed to the lower MMSE scores in the LHD group. 3. Analysis All patients were assessed with their ipsilesional hand. Our previous work has shown that there were no hand differences in gesture performance in the controls (Roy, Square-Storer, Hogg, & Adams, 1991) and, hence, ipsilesional hand performance in the stroke groups was examined with reference to the average of the 4.2. Task modality and gesture type comparisons A 2 (chronicity: acute–subacute vs. chronic)  2 (hemisphere: LHD vs. RHD)  2 (task modality: pantomime vs. imitation)  2 (gesture type: transitive vs. intransitive) mixed multivariate anal- 486 V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 ysis of variance (MANOVA) was used to compare the performance of the stroke patients on the apraxia assessment. Analyses (Fig. 1) revealed significant main effects for hemisphere, F(1, 76) = 13.57, p < .001, indicating LHD patients were more impaired (Mean = 6.7) than RHD patients (Mean = 1.4). Significant effect of chronicity was also found F(1, 76) = 4.5, p = .05, showing that acute–subacute patients (Mean = 5.5) performed worse than chronic patients (Mean = 2.6). A significant effect for task modality was also found, F(1, 76) = 11.2, p < .005 indicating performance on imitation (Mean = 5.4) was more impaired relative to controls than performance on pantomime (Mean = 2.7) (see Fig. 1). A main effect of gesture type was also found F(1, 76) = 4.3, p < .05, showing performance of transitive gestures was more impaired (Mean = 4.5) than performance of intransitive gestures (Mean = 3.6). There were no significant interactions. The two interactions hemisphere and gesture and hemisphere and task, that were expected were observed but were not significant (task modality by hemisphere [F(1, 76) = 2.8, p = .07] and gesture type by hemisphere [F(1, 76) = 2.5, p = .1], showing the effects of gesture type and task modality to be slightly more pronounced in the LHD group. 4.3. Frequency of apraxia A chi-square analysis was used to compare the frequency of apraxic vs. non-apraxic patients in each of the four groups of patients. A separate chi-square analysis was run for each of the four tasks. The analysis showed that there was a significant difference among the four groups in the number of patients falling into the apraxic vs. non-apraxic groups. See Table 2 for the number of patients in each category for all four tasks. In addition, chi-square analyses examining the frequency of apraxia in the LHD and the RHD groups irrespective of whether patients were acute–subacute or chronic revealed a higher number of apraxic patients with LHD in each of the four tasks. These differences were significant for all four tasks, except Pantomime of Intransitive Gestures, where it was found to be marginally significant at p = .055 (see Table 2 for number of cases and frequencies of apraxia categories per hemisphere group). 4.4. Patterns of apraxia in transitive and intransitive gestures All patients were classified as apraxic or non-apraxic across tasks. Similarly to the classification described in Roy et al. (2000) and Heath et al. (2001), patients were categorized into four patterns of performance for transitive gestures and four patterns of performance for intransitive gestures based on whether they were impaired on pan- Pantomime Transitive Pantomime Intransitive Imitaiton Transitive Imitation Intransitive Z-Score Relative to Controls 0 -2 -4 -6 Table 2 Number of cases and frequency of occurrence non-apraxics vs. apraxic patients in each hemisphere and chronicity group. Normal Frequency Apraxic v2 p-Value % Frequency % Pantomime transitive Acute LHD 7 Chronic LHD 9 Acute RHD 15 Chronic RHD 13 LHD 16 RHD 28 30 47 79 68 38 74 16 10 4 6 26 10 70 53 21 32 62 26 11.8 <.01 10.2 <.005 Pantomime intransitive Acute LHD 10 Chronic LHD 14 Acute RHD 14 Chronic RHD 16 LHD 24 RHD 30 43 74 74 84 57 79 13 5 5 3 18 8 57 26 26 16 43 21 9.1 <.05 4.3 0.055 Imitation transitive Acute LHD 7 Chronic LHD 7 Acute RHD 9 Chronic RHD 14 LHD 14 RHD 23 30 37 47 74 33 61 16 12 10 5 28 15 70 63 53 26 67 39 8.8 <.05 5.9 <.05 Imitation intransitive Acute LHD 6 Chronic LHD 10 Acute RHD 14 Chronic RHD 12 LHD 16 RHD 26 26 53 74 63 38 68 17 9 5 7 26 12 74 47 26 37 62 32 10.7 <.05 7.4 <.05 tomime and/or imitation. The patterns were both pantomime and imitation not impaired or non-apraxic (NA). (PNA INA) impaired pantomime but preserved imitation (PA INA) impaired imitation, but preserved pantomime (PNA IA) and impaired or apraxic performance in both pantomime and imitation (PA IA). These patterns were coded for each patient separately for transitive and intransitive gestures. Chi-square analyses were run to compare the frequency of each pattern in each of the four groups of participants. The analyses were significant for both gesture types (see Table 3). Looking first at transitive gestures for the LHD patients the most common pattern was an impairment in both tasks and this was the case for both acute and chronic patients. For RHD patients, the most common pattern was no impairment on either task in both acute and chronic patients. For intransitive gestures, in LHD patients the most common pattern in acute patients was impairment in both pantomime and imitation, while in chronic patients the most common pattern was ‘nonapraxic’ on either task. In RHD patients, again for both acute and chronic the most common pattern was that of no impairment on either task. Other interesting observations were that for transitive gestures selective impairments in either pantomime or imitation seemed to be slightly more frequent after RHD stroke. For intransitive gestures, selective impairments in pantomime or in imitation were relatively equally often seen in both hemisphere stroke groups with somewhat higher prevalence of patients with a selective deficit in imitation, as opposed to a selective deficit in pantomime. -8 -10 LHD Acute-Subacute -12 RHD Acute-Subacute LHD Chronic -14 RHD Chronic -16 Fig. 1. Average performance in Z-scores for each of the four patient groups on each task modality. Error bars represent standard errors. 4.5. Case analysis of concurrent transitive and intransitive gestures: associations and dissociations Given our aim was to observe relative performance of transitive and intransitive gestures within the same sample of patients, we investigated associations and dissociations between impairments on transitive and intransitive gestures on both pantomime and imitation. Table 4 summarizes the number of cases for each pat- 487 V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 Table 3 Patterns of apraxia for each group in transitive and intransitive gestures. Transitive gestures, v2 = 24.56, p < .005 Acute Chronic LHD PNA INA PA INA PNA IA PA I A Intransitive gestures, v2 = 17.0, p < .05 RHD Acute LHD RHD Chronic LHD RHD LHD RHD n % n % n % n % n % n % n % n % 5 2 2 14 22 9 9 61 8 1 7 3 42 5 37 16 7 0 2 10 37 0 11 53 10 4 3 2 53 21 16 11 4 2 6 11 17 9 26 48 11 3 3 2 58 16 16 11 8 2 6 3 42 11 32 16 11 1 5 2 58 5 26 11 P = pantome, I = imitation, NA = non-apraxic, and A = apraxic. tern of impairment concurrently for transitive and intransitive gestures. The majority of patients (n = 24) were not impaired on any of the four task modalities (PT+ IT+ PI+ II+, where ‘PT’ = pantomime of transitive gestures, ‘IT’ = imitation of transitive gestures, ‘PI’ = pantomime of intransitive gestures and ‘II’ = imitation of intransitive gestures; ‘+’ indicates preserved, while ‘ ’ indicates impaired performance). In addition, if patients showed any deficits, they were most likely to be impaired on all four task modalities (PT IT PI II ; n = 14). In addition, while there was one patient who was impaired on the two transitive tasks and not impaired on the intransitive tasks (PT IT PI+ II+), there were no cases of patients impaired on the two intransitive tasks but not on the transitive tasks (PT+ IT+ PI II ). There were three patients who were impaired on both intransitive tasks, but impaired only on pantomime of transitive gestures (PT IT+ PI II ). One case was impaired on the two intransitive tasks, and only on imitation of transitive gestures (PT+ IT PI II ). Some patients had selective intransitive impairments (four were impaired only on imitation (PT+ IT+ PI+ II ) and two only on pantomime of intransitive gestures (PT+ IT+ PI II+). Selective deficits in imitation (both for transitive and intransitive gestures) were observed in five cases, while selective deficits in pantomime for both gestures types were not observed in any of the cases. Interestingly, there were double dissociations between pantomime of transitive and intransitive gestures. Five cases were unable to pantomime intransitive gestures, but their pantomime of transitive gestures was preserved. The opposite, preserved pantomime of intransitive gestures, but impaired pantomime of transitive gestures, was observed in a total of 13 cases. 4.6. WAB-apraxia relationships among LHD patients For the subsample of patients on whom we had WAB AQ data (n = 23), we ran a Pearson r correlation between the four gesture performance tasks and the WAB AQ score. Only pantomime of transitive gestures was significantly correlated with WAB AQ (r = .43, p < .05). This correlation is thought to reflect the close association of anatomical regions controlling apraxia(Kertesz, Ferro, & Shewan, 1984) and aphasia and is not thought to be responsible for the low scores on pantomime of transitive gestures. If deficits in pantomime were due to aphasia, we would have expected to see the same association with pantomime of intransitive gestures, given both pantomime tasks rely on good language comprehension. 5. Discussion The current study aimed to systematically examine performance differences on pantomime and imitation of transitive and intransitive gestures in LHD vs. RHD stroke patients. The goal was to extend the work of Roy et al. (2000) and Heath et al. (2001) by directly comparing the performance of transitive and intransitive gestures within the same sample and to examine performance on both gesture types concurrently within the same cases. It was thought that this approach would allow us to look for associations and dissociations of apraxia deficits between the two gesture types. In addition, we included a new variable in our comparison, that of chronicity, because the time post-stroke varied considerably among patients and we expected that chronic patients may perform better than acute–subacute patients presumably because the longer the time since stroke the greater the likelihood of some recovery. 5.1. Limb apraxia performance analyses Table 4 Summary of cases of patients with impairments classified for each of four task modalities. Pattern Cases (n) LHD RHD PT + IT + PI + II+ PT + IT + PI + II PT + IT + PI II+ PT + IT PI + II+ PT + IT PI + II PT + IT PI II+ PT + IT PI II PT IT + PI + II+ PT IT + PI + II PT IT + PI II PT IT PI + II+ PT IT PI + II PT IT PI II+ PT IT PI II 24 4 2 7 5 1 1 2 2 3 1 9 5 14 11 2 0 1 3 0 0 2 1 1 0 7 4 13 13 2 2 6 2 1 1 0 1 2 1 2 1 1 PT = pantomime transitive, IT = imitation transitive, PI = pantomime intransitive, and II = imitation intransitive. Similar to many other studies (De Renzi, Motti, & Nichelli, 1980; Heath et al., 2001; Rothi & Heilman, 1997; Roy et al., 1998, 2000), LHD patients were more impaired than RHD patients irrespective of gesture modality or gesture type. The greater impairment in LHD patients is consistent with the theory that pantomime, as well as performance of tool-related gestures are left hemisphere dominant (Buxbaum et al., 2007; Haaland & Flaherty, 1984; Mozaz et al., 2002; Rapcsak et al., 1993). In addition, performance of transitive gestures was significantly more impaired than that on intransitive gestures, again supporting past literature (Almeida, Black, & Roy, 2002; Goodglass & Kaplan, 1963; Roy et al., 1993). In apparent contrast to our expectations and to past studies reporting pantomime accuracy to be lower than imitation (Alexander, Baker, Naeser, Kaplan, & Palumbo, 1992; Heath et al., 2001; Roy et al., 2000; Schnider et al., 1997), imitation Z-scores were lower than pantomime Z-scores. At first glance, this finding seemed at odds with past studies showing lower accuracy 488 V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 100 Accuracy (%) 90 80 70 60 Controls 50 LHD RHD 40 30 20 10 0 Pantomime Transitive Pantomime Intransitive Imitaiton Transitive Imitation Intransitive Fig. 2. Mean and standard deviations for controls (n = 27), LHD (n = 42) and RHD (n = 49) on each of the four tasks. Error bars represent standard errors. in pantomime than imitation, but in fact, it is not. Our analysis was based on Z-scores that standardized the performance of patients relative to that of controls. The actual percentage scores (see Fig. 2) for each group, revealed that, consistent with past literature, patients in fact performed with lower accuracy in pantomime than imitation; however, relative to control participants, their imitation performance was more severely affected. While patients were less accurate on pantomime than imitation in terms of their percentage scores, the greater negative Z-scores for imitation reflects a greater impairment in the stroke patients due to the greater accuracy and smaller standard deviation for imitation in the control group. In addition, while no significant interactions were observed between task modality and hemisphere group and gesture type and hemisphere group, the effects of task modality and gesture type were somewhat more pronounced in the LHD group. 5.2. Chronic patients were more accurate than acute–subacute patients Our study also included a comparison between acute–subacute and chronic patients. This classification of the patients was included because past studies, while few and conducted only in left hemisphere stroke, have suggested that there is a significant recovery from apraxia, especially in the first three months post-stroke (Basso et al., 2000; Foundas et al., 1993). Given our sample of patients included a wide range of times post-stroke, we expected that the acute–subacute patients would perform worse than chronic patients, particularly for the LHD patients. We expected that chronic LHD patients would have undergone some recovery of praxis function and hence may be less severely apraxic than acute patients. Our expectations were largely supported in that the acute–subacute patients were found to be more impaired than chronic patients. Both acute–subacute and chronic LHD patients had mean Z-scores below 2SDs (acute LHD Mean = 9.4, while chronic LHD patients Mean = 4.1) suggesting that both groups were impaired relative to the control participants. In contrast, the RHD groups’ mean scores (acute RHD Mean = 2.0 and chronic RHD Mean = 1.2) fell at or above the impairment cutoff suggesting better performance than that of LHD patients. Therefore, the better performance in the chronic groups may be seen as some support for recovery of apraxia. A longitudinal study would be required to fully and more reliably address this question. 5.3. Patterns of apraxia In the present study, we observed all four possible patterns of deficits for both transitive and intransitive gestures. In most cases, all four patterns were represented in each of the four groups of par- ticipants, with the exception of the selective deficit in pantomime, which was not observed in chronic LHD patients. First, examining the frequency of patients who did not show deficits on either pantomime or imitation, we noted that in both transitive and intransitive gestures there was always a greater number of RHD patients who fall into this pattern. Again, this confirms the greater role of the left hemisphere in the control of pantomime and imitation tasks. In addition, in all cases the chronic groups had a higher number of patients who were not impaired on either task, supporting the notion that the chronic group was less affected by apraxia, presumably because these patients had recovered to some extent. Second, we turn to the pattern of performance representing a selective impairment in pantomime. Such a pattern of performance, according to Roy (1996), suggests that patients have deficits in the conceptual knowledge of gestures and tools, or may have that knowledge disconnected from the centers responsible for organizing movement. In addition, these patients may have deficits in the early stage of the production system involving response selection and/or image generation. It has been widely accepted that the conceptual knowledge related to gestures and tools is stored in the left hemisphere (Buxbaum, 2001; Heilman, Maher, Greenwald, & Rothi, 1997; Tranel, Kemmerer, Adolphs, Damasio, & Damasio, 2003). So at least for the cases with damage to the conceptual system, with preserved functioning of the early stages of the production system, we would expect that this pattern of performance would be more frequent after LHD. The results in this study indicate that this pattern is equally likely after damage to either hemisphere, and in some cases, it is even more prevalent after right hemisphere damage (Among the chronic group this pattern was not observed after LHD, but it was highly prevalent after RHD.). Given, we had no measure of conceptual knowledge included in this study, the only way we can surely conclude that the conceptual system has been damaged is if we examine performance on both gesture types concurrently. If patients are able to pantomime either one of the two gesture types, we could conclude that the early stages of the production system are intact and thus the pantomime deficit must be associated with a disruption to the conceptual system. We examine this in the next section. For intransitive gestures, selective deficits in pantomime were more prevalent after RHD among acute patients, but the reverse was observed among chronic patients. The third pattern of performance, selective deficit in imitation, was more frequently observed than selective impairment in pantomime. For transitive gestures, the pattern was more prevalent after RHD in both acute and chronic patients. For intransitive gestures, the opposite was observed, a slightly greater number of LHD patients, especially among acute patients, presented with this pattern of apraxia. Selective deficits in imitation may arise from deficits in processing visuospatial information, or from deficits translating this movement into action (Roy, 1996). It is possible, therefore, that the acute RHD patients may have had greater deficits in visuospatial processing, given the role of the right hemisphere in such tasks. Intransitive gestures, may be easier to process visuospatially, given they are performed in a more natural context (as opposed to pantomime where the patient pretends to hold the tool in their hand), which may have rendered them easier to identify. Finally, the fourth pattern, impairment in both pantomime and imitation could reflect impairments in the various stages of the production system (given here we did not include a conceptual task it is hard to determine if the deficits in gesture production is more global (affecting both conceptual and production system) or affecting only the final stages of the production system with preserved conceptual system). This pattern was consistently more frequent after LHD. In all cases, it was close to four times more frequent after LHD, except among the chronic group’s performance V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 on intransitive gestures, where the frequency among LHD was 16% while after RHD was 11%. This finding confirms the greater role of the left hemisphere in the final stages of the control of movement and suggests that the left hemisphere is critical for the proper functioning of the production system. 5.4. Associations and dissociations between transitive vs. intransitive gesture impairments Most apraxic patients were impaired in both transitive and intransitive gestures, supporting the notion that transitive and intransitive gestures share a common neural network (Kroliczak & Frey, 2009). In addition, the fact that most patients with impairments in all four tasks were LHD, supports the idea that these common networks are situated in the left hemisphere. In fact, no patients were observed to be impaired in intransitive gestures but not in transitive gestures (or PT+ IT+ PI II ). We did observe the reverse situation in one case (or PT IT PI+ II+). Overall, complete double dissociation between transitive and intransitive gestures was not found. We did, however, observe cases of selective impairment in imitation (PT+ IT+ PI+ II ) and a selective deficit in pantomime (PT+ IT+ PI II+) of intransitive gestures with preserved performance on transitive gestures. A very recent paper (Kroliczak & Frey, 2009) reported that cases with impairments in intransitive gestures but not transitive are practically non-existent in the literature. Here, we report on six cases such cases. In addition, while the pattern PT+ IT+ PI+ II was equally likely to occur after damage to either hemisphere (2 LHD and 2 RHD patients had this pattern), the pattern PT+ IT+ PI II+ was observed only in two RHD patients. Deficits in pantomime could occur in patients who have either damage to the conceptual action knowledge or to the early stages of gesture production. Given pantomime was preserved in the performance of transitive gestures, we could assume that the early stages of gesture production were intact. This is because these early stages are thought to be common to both gesture types. This leaves out deficits in the conceptual knowledge or loss of access to the conceptual system of intransitive gestures as the cause of pantomime deficits in these patients. Hence, contrary to the evidence of a common conceptual network for transitive and intransitive in the left hemisphere (as most of our cases support), these two patients lend support for the notion that conceptual knowledge of intransitive gestures could also be represented in the right hemisphere. In effect, we report here a double dissociation between pantomime of transitive and pantomime of intransitive gestures. Deficits in pantomime of intransitive, but not transitive gestures, were associated with RHD (all four patients had RHD). Deficits in pantomime of transitive, but not intransitive gestures, were associated with LHD (10 out of 14 patients had LHD). Note that four of these cases with deficits in pantomime of transitive but not intransitive gestures had RHD suggesting that in some cases the right hemisphere could be critical in controlling pantomime of transitive gestures. Earlier in our examination of limb apraxia patterns in transitive gestures, we reported that selective deficits in pantomime are equally likely to occur after damage to either hemisphere. Some of these patients had deficits in intransitive gestures while others did not. Therefore, transitive conceptual knowledge may be more bilaterally represented than previously thought, suggesting that damage to critical regions in the right hemisphere, at least in some cases was sufficient to disrupt pantomime of tool use. Overall, damage to the right hemisphere was generally more likely to result in a pattern of deficits in which pantomime of transitive gestures was preserved. Out of the 20 patients with preserved pantomime of transitive gestures, 14 were RHD. Many of these cases had selective deficits in imitation of either transitive 489 or intransitive gestures. Such deficits could likely be caused by the high visuospatial processing demands placed on the brain during imitation tasks. It has been proposed that direct visuomotor transformations, such as those used in imitation of meaningless gestures, may be bilaterally represented (Buxbaum et al., 2007). If patients were forced to use the direct route due to an inability to visually recognize a gesture, this may explain selective deficits in imitation after RHD. In contrast, 28 out of the 36 patients with deficits in pantomime of transitive gestures were LHD. While, as stated above, we do have several cases that support the notion that the conceptual representation of some meaningful gestures could be stored in the right hemisphere, most cases in our sample provide support for a left hemisphere conceptual system for transitive gestures. Consistent with both patient studies (Goldenberg, Hermsdorfer, Glindemann, Rorden, & Karnath, 2007; Goldenberg & Spatt, 2009; Hanna-Pladdy et al., 2001; Heilman et al., 1997; Tranel et al., 2003) as well as functional neuroimaging studies (Grezes & Decety, 2001; Kroliczak & Frey, 2009; Villarreal et al., 2008). While the left hemisphere is thought to be dominant, neuroimaging studies have shown bilateral activation during preparation of pantomime performance (Kroliczak & Frey, 2009), and during execution and observation of actions (Grezes & Decety, 2001). Given the involvement of both hemispheres in the control of movement, it is not surprising that in some cases damage to the RHD hemisphere could also produce deficits. 5.5. Study limitations and overall conclusion Finally, we would like to point out one limitation of our study with regard to examining the effect of time post-stroke. While we make some inferences from this effect with respect to recovery, these conclusions were purely exploratory in nature. We believe that the only true way to assess recovery of praxis is through a longitudinal study examining apraxia over time post-stroke within each subject. Here we took advantage of the fact that we had both acute and chronic patients in our sample and wanted to examine any potential differences between the two stroke groups in recovery. While we did confirm our expectations that chronic patients with LHD performed more accurately than acute patients, the reader should remember that these are separate samples of patients and there are many other variables that could explain the difference in performance. Aside from confirming some hemisphere and gesture type findings of past studies, we have provided evidence that selective deficits in pantomime, but not imitation could result after damage to either hemisphere in both gesture types. In addition, after looking at both gesture types concurrently, we have described cases of patients who suffered deficits in pantomime of intransitive gestures with preserved performance on transitive gestures. Such cases show that the right hemisphere may be in some cases critical for the successful pantomime of intransitive gestures and the neural networks subserving them may be distinct. This is further supported by our finding that while the WAB scores were significantly correlated with pantomime of transitive gestures scores, they did not correlate with the performance on pantomime of intransitive gestures. In addition, these findings suggest that clinically it may be wise to always assess both transitive and intransitive gestures, even though overall in group analysis transitive gestures tend to be more severely affected. Cases such as the ones reported here with selective deficits in intransitive gestures may go undiagnosed if only transitive gestures were assessed. Finally, the study also provides some evidence that performance on pantomime and imitation may be predictable based on the time that has elapsed since the patient’s stroke. Chronic patients performed better than acute–subacute patients after damage to either 490 V. Stamenova et al. / Brain and Cognition 72 (2010) 483–490 the left or right hemisphere. Therefore, possibly for the first time we provide evidence of potential recovery of limb apraxia after RHD. A longitudinal study should further examine recovery of apraxia directly in both LHD and RHD patients. Acknowledgments We would like to take this opportunity to thank Dr. William McIlroy for reviewing the paper and giving us feedback. We would also like to thank all the research assistants who have helped collect data over the years: Kira Barbour, Anish Joshi, Dr. Quincy Almeida, Dr. Jennifer Salter, Anastasia Aranvitidis and Mark Gravely. We are grateful to funding from the Heart and Stroke Foundation for this research. VS acknowledges training support resources from the Natural Sciences and Engineering Research Council of Canada, Toronto Rehabilitation Institute Scholarship Fund, Margaret & Howard Gamble Research Grant and University of Toronto Fellowships. SEB acknowledges support from the Sunnybrook Research Institute, Brain Sciences Program, the Brill Chair of Neurology, and the Departments of Medicine at Sunnybrook Health Sciences, Centre and University of Toronto. References Alexander, M. P., Baker, E., Naeser, M. A., Kaplan, E., & Palumbo, C. (1992). Neuropsychological and neuroanatomical dimensions of ideomotor apraxia. Brain, 115, 87–107. Almeida, Q. J., Black, S. E., & Roy, E. A. (2002). Screening for apraxia: A short assessment for stroke patients. Brain and Cognition, 48, 253–258. Basso, A., Burgio, F., Paulin, M., & Prandoni, P. (2000). Long-term follow-up of ideomotor apraxia. Neuropsychological Rehabilitation, 10, 1–13. Buxbaum, L. J. (2001). Ideomotor apraxia: A call to action. Neurocase, 7, 445–458. Buxbaum, L. J., Kyle, K., Grossman, M., & Coslett, H. B. (2007). Left inferior parietal representations for skilled hand-object interactions: Evidence from stroke and corticobasal degeneration. Cortex, 43, 411–423. Carmo, J. C., & Rumiati, R. I. (2009). Imitation of transitive and intransitive actions in healthy individuals. Brain and Cognition, 69, 460–464. De Renzi, E., Motti, F., & Nichelli, P. (1980). Imitating gestures. A quantitative approach to ideomotor apraxia. Archives of Neurology, 37, 6–10. Donkervoort, M., Dekker, J., van den Ende, E., Stehmann-Saris, J. C., & Deelman, B. G. (2000). Prevalence of apraxia among patients with a first left hemisphere stroke in rehabilitation centres and nursing homes. Clinical Rehabilitation, 14, 130–136. Foundas, A. L., Raymer, A. M., Maher, L. M., Gonzalez-Rothi, L., & Heilman, K. M. (1993). Recovery in ideomotor apraxia. Journal of Clinical Experimental Neuropsychology, 14, 44. Goldenberg, G., Hermsdorfer, J., Glindemann, R., Rorden, C., & Karnath, H. O. (2007). Pantomime of tool use depends on integrity of left inferior frontal cortex. Cerebral Cortex, 17, 2769–2776. Goldenberg, G., & Spatt, J. (2009). The neural basis of tool use. Brain, 132, 1645–1655. Gonzalez-Rothi, L., Mack, L., Verfaellie, M., Brown, P., & Heilman, K. M. (1988). Ideomotor apraxia: Error pattern analysis. Aphasiology, 2, 381–387. Goodglass, H., & Kaplan, E. (1963). Disturbance of gesture and pantomime in aphasia. Brain, 86, 703–720. Grezes, J., & Decety, J. (2001). Functional anatomy of execution, mental simulation, observation, and verb generation of actions: A meta-analysis. Human Brain Mapping, 12, 1. Haaland, K. Y., & Flaherty, D. (1984). The different types of limb apraxia errors made by patients with left vs. right hemisphere damage. Brain and Cognition, 3, 370–384. Haaland, K. Y., Harrington, D. L., & Knight, R. T. (2000). Neural representations of skilled movement. Brain, 123, 2306–2313. Hanna-Pladdy, B., Daniels, S. K., Fieselman, M. A., Thompson, K., Vasterling, J. J., Heilman, K. M., et al. (2001). Praxis lateralization: Errors in right and left hemisphere stroke. Cortex, 37, 219–230. Heath, M., Roy, E. A., Westwood, D., & Black, S. E. (2001). Patterns of apraxia associated with the production of intransitive limb gestures following left and right hemisphere stroke. Brain and Cognition, 46, 165–169. Heilman, K. M., Maher, L. M., Greenwald, M. L., & Rothi, L. J. (1997). Conceptual apraxia from lateralized lesions. Neurology, 49, 457–464. Heilman, K. M., & Rothi, L. J. (1993). Apraxia. In Clinical neuropsychology (pp. 141–163). New York, NY: Oxford University Press. Kertesz, A., Ferro, J. M., & Shewan, C. M. (1984). Apraxia and aphasia: The functional–anatomical basis for their dissociation. Neurology, 34, 40–47. Kertesz, A., & Poole, E. (1974). The aphasia quotient: The taxonomic approach to measurement of aphasic disability. Canadian Journal of Neurological Sciences, 1, 7–16. Kroliczak, G., & Frey, S. H. (2009). A common network in the left cerebral hemisphere represents planning of tool use pantomimes and familiar intransitive gestures at the hand-independent level. Cerebral Cortex, 19, 2396–2410. Mozaz, M., Rothi, L. J., Anderson, J. M., Crucian, G. P., & Heilman, K. M. (2002). Postural knowledge of transitive pantomimes and intransitive gestures. Journal of the International Neuropsychological Society, 8, 958–962. Rapcsak, S. Z., Ochipa, C., Beeson, P. M., & Rubens, A. B. (1993). Praxis and the right hemisphere. Brain and Cognition, 23, 181–202. Rothi, L. J. G., & Heilman, K. (1997). Apraxia: The neuropsychology of action. Hove, UK: Psychology Press. Roy, E. A. (1996). Hand preference, manual asymmetries, and limb apraxia. In D. Elliot (Ed.), Manual asymmetries in motor control (p. 215). Boca Raton, FL: CRC Press. Roy, E. A., Black, S. E., Blair, N., & Dimeck, P. T. (1998). Analyses of deficits in gestural pantomime. Journal of Clinical and Experimental Neuropsychology, 20, 628–643. Roy, E. A., Brown, L., Winchester, T., Square, P., Hall, C., & Black, S. (1993). Memory processes and gestural performance in apraxia. Adapted Physical Activity Quarterly, 10, 293–311. Roy, E. A., Heath, M., Westwood, D., Schweizer, T. A., Dixon, M. J., Black, S. E., et al. (2000). Task demands and limb apraxia in stroke. Brain and Cognition, 44, 253–279. Roy, E. A., Square-Storer, P., Hogg, S., & Adams, S. (1991). Analysis of task demands in apraxia. International Journal of Neuroscience, 56, 177–186. Rumiati, R. I., & Tessari, A. (2002). Imitation of novel and well-known actions: The role of short-term memory. Experimental Brain Research, 142, 425–433. Schnider, A., Hanlon, R. E., Alexander, D. N., & Benson, D. F. (1997). Ideomotor apraxia: Behavioral dimensions and neuroanatomical basis. Brain and Language, 58, 125–136. Tranel, D., Kemmerer, D., Adolphs, R., Damasio, H., & Damasio, A. R. (2003). Neural correlates of conceptual knowledge for actions. Cognitive Neuropsychology, 20, 409–432. Villarreal, M., Fridman, E. A., Amengual, A., Falasco, G., Gerscovich, E. R., Ulloa, E. R., et al. (2008). The neural substrate of gesture recognition. Neuropsychologia, 46, 2371–2382.