- Email: [email protected]
Dysgraphia in Korean patients with Alzheimer's disease as a manifestation of bilateral hemispheric dysfunction
Journal of the Neurological Sciences 320 (2012) 72–78
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Dysgraphia in Korean patients with Alzheimer's disease as a manifestation of bilateral hemispheric dysfunction Ji Hye Yoon a, b, HyangHee Kim a, c,⁎, Sang Won Seo b, Juhee Chin b, Jung-Hyun Kim b, Kyung-Han Lee d, Yong Wook Kim c, Eun Sook Park c, Mee Kyung Suh b, Duk L. Na b,⁎⁎ a
Graduate Program in Speech and Language Pathology, Yonsei University, Seoul, Republic of Korea Department of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea Department and Research Institute of Rehabilitation Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea d Department of Nuclear Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea b c
a r t i c l e
i n f o
Article history: Received 2 January 2012 Received in revised form 14 May 2012 Accepted 26 June 2012 Available online 16 July 2012 Keywords: Hangul Writing Dysgraphia Linguistic Nonlinguistic Visuospatial Alzheimer's disease PET
a b s t r a c t Background: In writing, linguistic (i.e., spelling) and nonlinguistic (i.e., arranging strokes or letters) functions are processed by the left and right hemispheres, respectively. The configuration of Korean alphabet, ‘Hangul’ invokes nonlinguistic, visuospatial functions that other writing systems use less extensively. Patients with Alzheimer's disease (AD) have bilateral involvement of temporoparietal–frontal areas that are responsible for processing language and visuospatial functions. Objectives: The aim of this study was to examine the nature of Hangul writing dysfunction, which may be associated with bilateral hemispheric impairments in AD. Methods: A sample of 75 patients with AD and 20 healthy controls (HC) performed a Hangul writing task. Neuroimaging positron emission tomography (PET) data of 22 patients were utilized to measure the regional brain glucose metabolism associated with Hangul writing. Results: The writing performance of the AD group was significantly reduced and different types of errors were observed as the disease got worse. Glucose hypometabolism correlated with Hangul writing impairment was located in the right occipitotemporal lobe and left temporoparietal lobe. Conclusions: The PET findings demonstrate that impairment in Hangul writing performance in Korean AD patients is closely related to a functional decline in both the right and left hemispheres. The study provides a unique contribution to the knowledge of dysgraphia in a non-alphabetical writing system as well as the underlying neuropathology of dysgraphic features in such languages. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Writing is not a unitary process, but requires coordination of linguistic and visuospatial aspects [1,2]. Traditionally, linguistic functions such as spelling are thought to be processed by the left hemisphere (LH), while visuospatial constructions in writing such as arranging strokes or letters are processed by the right hemisphere (RH) [3]. Therefore, each hemisphere may play a significant role in writing processing. In alphabetical written language systems, the letters are written horizontally. Writing is usually related to functional cortical activity, which appears to be predominant in the LH areas. Most previous studies on writing impairments of alphabetical writing systems have been designed to focus on LH lesions [4–8].
⁎ Correspondence to: H. Kim, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul 120‐752, Republic of Korea. Tel.: +82 2 2228 3900. ⁎⁎ Correspondence to: D.L. Na, Department of Neurology, Samsung Medical Center, 50 Ilwon-dong, Kangnam-ku, Seoul 135‐710, Republic of Korea. Tel.: +82 2 3410 3591. E-mail addresses: [email protected] (H. Kim), [email protected] (D.L. Na). 0022-510X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2012.06.020
Such languages as Chinese (i.e., Hanja), Japanese (i.e., Kanji and Kana), and Korean (i.e., Hangul) utilize non-alphabetic writing systems. For example, Chinese and Kanji characters have complex graphic forms that consist of individual strokes, and the chunks of strokes are arranged in a diverse-dimensional square. The brain areas responsible for the non-alphabetic writing process may differ from those for alphabetically written language systems. The non-alphabetic writing language systems may be more related to RH functions than alphabetically written language systems. The Korean alphabet, Hangul (or Hangeul) is unique in its written application with respect to the nonlinguistic, visuospatial as well as linguistic aspects. Linguistically, the Hangul alphabet consists of 24 characters, split between 10 vowel-graphemes (‘mo-eum’) and 14 consonant-graphemes (‘ja-eum’). Each of the 24 Hangul graphemes corresponds to one phoneme. Visuospatially, each grapheme is placed within a square space to form a syllable. Hangul syllables are classified into three subtypes according to their visuospatial constructional patterns: (a) vertical- (written from top-to-bottom, e.g., ), (b) horizontal- (written from left-to-right, e.g., ), and (c)
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
mixed-construction (i.e., combination of horizontal and vertical constructions, e.g., ). Taking this syllabic rule into consideration, a word (e.g., ) could be created (Fig. 1). Thus, the configuration of Hangul syllables invokes visuospatial functions that other writing systems use less extensively. It can be hypothesized that this unique nature of Hangul may compromise Hangul writing functions in patients with bilateral hemispheric lesions differently compared with those with selective damage to a unilateral hemisphere. To date, however, studies on Hangul have focused on stroke patients with unilateral focal lesions [9–12]. The primary errors observed in unilateral stroke patients are linguistic errors such as substitution [10–13]. Stroke patients with RH lesions display prominent errors relevant to the physical form such as stroke omission and addition, and visuospatial errors such as graphemic misposition [13,14]. Alzheimer's disease (AD) is a disease with both visuospatial and language dysfunctions as part of the disease process due to the bilateral involvement of the temporoparietal–frontal lobes [15]. A previous study was conducted to explore the nature of writing dysfunction in Korean patients with early-onset AD [16]. The authors reported that these patients demonstrated linguistic and visuospatial errors in writing. In addition, the severity of dementia and multiple cognitive domains such as attention, language, immediate memory, and frontal executive functions significantly correlated with the Hangul writing performance. The aim of this study was to examine if the nature of Hangul writing dysfunction is associated with bilateral hemispheric impairment. We investigated the neural correlates of Hangul writing using positron emission tomography (PET) imaging analysis. By employing both behavioral and PET data results, we attempted to establish the relationship between the unique features of Korean writing system and bilateral brain lesion found in AD.
2. Methods 2.1. Participants Initially, we recruited 80 consecutive patients diagnosed with AD at the Memory Disorder Clinic at Samsung Medical Center in Seoul, Korea between September 2009 and August 2011. The AD patients fulfilled the criteria for probable AD proposed by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS–ADRDA) [17]. The exclusion criteria were as follows: 1) those with less than six years of education because they might have had difficulty in writing Hangul in the premorbid state; 2) those with abnormal findings on laboratory tests including a complete blood count, blood chemistry, vitamin B12/folate, syphilis serology, and thyroid-function tests; 3) those with territorial cerebral infarctions, brain tumors, and other structural lesions on MRI; 4) those with mild bradykinesia and rigidity that might affect writing ability; 5) those who had refused the experimental testing.
73
Out of 80 patients, four declined to complete the experimental testing. One patient with mixed dementia was also excluded. As a result, 75 patients (male:female = 27:48) were included in this study. The mean age was 73.1 years (range: 46 to 92 years old). A panel consisting of two neurologists (D.L.N. and S.W.S.) and one neuropsychologist (J.C.) with expertise in dementia made the clinical decisions including diagnosis and clinical dementia rating (CDR) after reviewing the patients' clinical data. The patient group included 16 very mild (CDR 0.5), 37 mild (CDR 1), 19 moderate (CDR 2), and 3 severe (CDR 3) patients. All AD subjects were right-handed, spoke standard Korean, and reported no history of premorbid writing disturbances. In order to assess performances on the writing tests, we also recruited 20 age-and education-matched caregivers of the patients as healthy control (HC) group. They had no history of neurological or psychiatric illnesses. The demographic features of the 75 AD patients and 20 HC are presented in Table 1. 2.2. Materials 2.2.1. Hangul writing to dictation task All patients and HC completed a Hangul writing to dictation task. The dictation task comprises 60 single syllables extracted from multisyllabic words with one-to-one grapheme to phoneme correspondence [e.g., ‘ ’ (/ka/) of ‘ ’ (/ka-wi/) (meaning ‘scissors’)]. The selection of these syllables was based on the frequency, configuration, and imageability of the items. The multisyllabic words were selected from a pool of high frequency words in the Learner's dictionary of Korean [18]. In terms of configuration, characters with four graphemes or less were selected to reduce the complexity of the task. All words were concrete words with high imageability. For the Hangul writing to dictation task, the subjects wrote the stimuli on an A4 sheet of paper. A single syllable embedded in words was verbally presented to each patient. The subject was asked to verbally repeat the target stimuli before writing in order to rule out any deficit in auditory input processing that might influence writing to dictation. No time limit was imposed. The stimuli are provided in Appendix A. 2.2.2. Language test All patients completed the Western Aphasia Battery (WAB) [19]. The oral language subtests such as spontaneous speech, auditory comprehension, repetition, and naming were used to determine the type of aphasia and its severity. An Aphasia Quotient (AQ) was derived from the raw scores. The two subtests, reading and writing, were used to assess written language ability. Adding the reading and writing scores, the Language Quotient (LQ) was obtained. The AD patients showed the mean AQ of 82.1 and LQ of 76.5 (maximum score = 100). 2.2.3. Figure copying test Visuospatial ability is also regarded as an important function in writing. In order to investigate the relationship between general visuospatial function and writing, 66 out of 75 AD patients underwent the Rey–Osterrieth Complex Figure Test (RCFT) using a standardized neuropsychological battery; the Seoul Neuropsychological Screening Battery (SNSB) [20]. The data of the nine patients could not be obtained Table 1 Demographic variables for the HC and AD groups (Mean ± SD).
Age (years) Sex: female, N (%) Education (years) MMSE CDR Fig. 1. Letters in English words are written in a linear fashion, while graphemes in Korean words are arranged in a square pattern.
HC (N = 20)
AD (N = 75)
71.5 ± 7.1 12(60%) 10.9 ± 3.7 29.0 ± 1.0 N/A
73.1 ± 9.5 48(64%) 11.1 ± 3.7 17.5 ± 5.8⁎ 1.2 ± 0.6
HC: healthy control, AD: Alzheimer's disease, MMSE: Mini-Mental State Examination, CDR: Clinical Dementia Rating; N/A: not applicable. ⁎ p b .05.
74
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
because they declined to undergo the test or because the time gap between the test and the Hangul writing experiment was over six months. The mean score on the RCFT was 23.1 in the patient group (maximum score= 36). 2.3. Writing error analysis The number of correctly produced Hangul syllables was counted. We regarded a response as correct if all graphemes in a syllable were appropriate with respect to the linguistic and nonlinguisitc aspects. The maximum possible score was 60 and the minimum was 0. We additionally analyzed error patterns based on the criteria proposed by previous studies [10,13,14,21,22]. The linguistic errors consisted of graphemic omission (e.g., ), substitution (e.g., ), and addition (e.g., ). The nonlinguistic errors consisted of awkward shaping and implausible configurations of Korean graphemes or syllables such as stroke omission, stroke addition, and graphemic misposition. By definition, stroke omission errors mean deletion of part(s) of a grapheme (e.g., ). Stroke addition errors occur when redundant strokes are added to an original grapheme (e.g., ). Errors of graphemic misposition (i.e., a slight shift of a grapheme) create a nonexistent form within the boundaries of the Korean syllable (e.g., ). More than one error could be counted per syllable because the errors were examined on each grapheme. In addition to the linguistic and nonlinguistic errors, mixed (linguistic as well as nonlinguistic) and miscellaneous (e.g., unintelligible responses, figure drawing, no response) were also recorded. 2.4. Statistics A t-test was conducted to confirm the differences in the total number of correct or error responses between the AD and HC groups. One-way analysis of variance and post hoc Bonferroni correction were conducted to identify any differences in number of erroneous responses according to disease severity within the AD group (p b .05). To evaluate the relationship between the figure copying or the language results and Hangul writing in the AD group, partial correlation analysis was conducted after adjusting for age, sex, and duration of education. 2.5. Positron emission tomography (PET) imaging analysis Resting state regional brain glucose metabolism was measured using [1F]fluoro-2-deoxy-D-glucose (18FDG). The participants fasted for at least 6 h before the scan. Thirty-minute positron emission tomography (PET) scans were acquired 40 min after the intravenous injection of 4.8 MBq/kg FDG using a GE Advance PET scanner. The protocol for obtaining the PET scan was published elsewhere [23]. Of the 75 patients, 22 underwent FDG-PET. The rest of the patients either did not undergo PET imaging within six months of performing the Hangul writing test or declined to undergo imaging due to personal reasons. In terms of AD severity, there were 3 very mild (CDR 0.5), 8 mild (CDR 1), 9 moderate (CDR 2), and 2 severe (CDR 3) patients. The mean age (73.4 ± 8.8 years), sex ratio (eight males), and education (11.9 ± 3.4 years) of patients with FDG-PET did not differ from those that did not undergo FDG-PET. The 22 patients exhibited the mean AQ of 77.9 and LQ of 73.0 (maximum score = 100). The mean AQ and LQ of patients with FDG-PET did not differ from those without. The PET study was approved by the Institutional Review Board (IRB; no. 2006-03-011) of Samsung Medical Center. The PET images were analyzed using SPM5 (Wellcome Department of Cognitive Neurology, Institute of Neurology, London, UK) and implemented using Matlab 7.0 (MathWorks Inc., Sherborn, MA, USA). Prior to the statistical analysis, all of the images were spatially normalized into the MNI standard template (Montreal Neurological Institute, McGill University, Montreal, Canada) in order to remove inter-subject anatomic variability. Spatially normalized images
were smoothed with a 16 mm FWHM isotropic Gaussian kernel. The count of each voxel was normalized to the average count of the cerebellum with proportional scaling in SPM5. In order to investigate hypometabolic brain areas that were associated with the performance, we performed multiple linear regression analysis within the patient group (corrected p b .05; FDR, false discovery rate) after adjusting for age and educational factors. To visualize the t-score statistics (SPM1 map), the significant voxels were projected onto the 3-D rendered brain or a standard high-resolution MRI template provided by SPM5. 3. Results 3.1. Performance of the Hangul writing task In terms of the mean total number of correct responses between the AD and HC groups, the performance of the AD group (M=41.1, SD = 18.3) was significantly reduced (p b .001) than that of the HC group (M= 58.6, SD = 1.3). When the total erroneous responses between the two groups were compared, the performance of the AD group (M= 32.5, SD=44.5) was significantly worse than that of the HC group (M = 1.3, SD = 1.4) (p b .001). Within the AD group, the number of total erroneous responses significantly differed according to CDR (p b .001). In the post hoc analysis, there were significant differences between CDR 0.5 and CDR 2 (p = .003), CDR 0.5 and CDR 3 (p = .017), CDR 1 and CDR 2 (p = .001), and between CDR 1 and CDR 3 (p = .021) (Table 2). No differences were observed between CDR 0.5 and CDR 1 and between CDR 2 and CDR 3. With respect to the error patterns, CDRs of 0.5 and 1 yielded more linguistic errors, while CDRs of 2 and 3 demonstrated more miscellaneous errors. The distribution of writing errors over different subtypes of error was presented in Table 3. The examples according to the error subtypes are presented in Fig. 3. 3.2. The relationship between the copying figure or the language test and the writing task performance The number of correct responses in the writing task was correlated with the RCFT scores. In addition, the number of correct responses in the writing task was correlated with the scores of all WAB subtests. The results are provided in Table 4. 3.3. SPM analysis of FDG-PET The total number of correct responses (M= 37.1, SD = 20.9) of patients with FDG-PET did not differ from those without. The voxels for which glucose metabolism positively correlated with the number of correct responses were located in the right occipitotemporal lobe and left temporoparietal lobe (Table 5, Fig. 2a). The voxels for which glucose metabolism positively correlated with the RCFT performance were located in the right occipitotemporal lobe (Table 6, Fig. 2b). The voxels for which glucose metabolism positively correlated with AQ were located in the left temporal, parietal, and frontal lobes (Table 7, Fig. 2c). 4. Discussion Due to the bilateral hemispheric compromise of AD, research on writing in this clinical population might provide significant information on the neural correlates of writing. Hangul writing is distinctive in that it has unique linguistic and visuospatial, nonlinguistic features. From the results of the current study, we found that the AD patients made linguistic errors including graphemic omission, graphemic substitution, and graphemic addition. These errors are not unique and are common errors in any written language such as English, Italian, and French [24–31]. In alphabetic writings, transposition errors could be observed between a vowel and a consonant and between a vowel and a
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
75
Table 2 The number of erroneous responses according to CDR (Mean ± SD).
Linguistic Nonlinguistic Mixed Miscellaneous N
HC (N = 20)
CDR 0.5 (N = 16)
CDR 1 (N = 37)
CDR 2 (N = 19)
CDR3 (N = 3)
Total AD (N = 75)
1.3 ± 1.4 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 1.3 ± 1.4
9.6 ± 11.5⁎ 2.1 ± 2.9 0.5 ± 0.9 2.4 ± 4.1 14.8 ± 15.9⁎
10.8 ± 12.9⁎ 2.3 ± 3.1 2.8 ± 11.0 4.0 ± 15.3 20.0 ± 28.6⁎
20.0 ± 19.4⁎ 7.8 ± 12.7⁎†‡ 5.9 ± 17.0 28.9 ± 55.5⁎ 62.6 ± 58.4⁎†‡
22.0 ± 33.0⁎ 8.0 ± 13.0⁎†‡ 0.0 ± 0.0 60.6 ± 94.8⁎†‡ 90.6 ± 81.5⁎†‡
13.2 ± 15.8⁎ 3.8 ± 7.5 2.9 ± 11.5 12.2 ± 36.5⁎ 32.5 ± 44.5⁎
HC: healthy control, AD: Alzheimer's disease, N: number of erroneous responses. ⁎ p b .05 between HC and CDR 0.5, HC and CDR 1, HC and CDR 2, HC and CDR 3, and between HC and AD group. † p b .05 between CDR 0.5 and CDR 2 and between CDR 0.5 and CDR 3. ‡ p b .05 between CDR 1 and CDR 2, and between CDR 1 and CDR 3.
vowel, and consonant and a consonant. Transposition errors of character components such as constituents were also observed in Chinese [22,32]. Nonlinguistic errors of writing in AD were confined to letter formation in previous studies on alphabetic writings [28,33–37], although we found stroke addition and omission as nonlinguistic errors in this study. However, among the nonlinguistic errors, graphemic misposition errors (Fig. 3c) are language-specific in that misposition can result in nonexistent and meaningless forms in Korean syllabic writing. This language-specific error has also been observed in Korean patients with stroke [13,14] and in patients with early onset AD [16]. This is attributed to the fact that each grapheme of Hangul has its own allocated space within a square syllabic form (Fig. 1). This is in contrast to alphabetic writings, which are arranged in a horizontal, linear fashion to form a syllable. A slight horizontal or vertical shift of a grapheme (e.g., ‘r’ in ‘root’) in alphabetical writing would not change the lexical meaning. The visuospatially unique feature of Hangul writing is related to performance of visuoconstructional tasks such as RCFT. In the current study, the performance of the RCFT task was highly associated with that of the Hangul writing task. Generally, figure copying might be accomplished through a pictorial route (e.g., slavish or ‘stroke-by-stroke’ copying), while Hangul writing is generated by a graphemic (spelling and writing) route. Unlike copying in the pictorial route, graphemic Hangul writing can take advantage of graphemic knowledge stored in long-term memory. However, these two tasks call for visuospatial abilities during executing of movement for performance, regardless of the route involved. This finding supports the notion that Hangul writing needs visuospatial functions which are also required in picture drawing/copying tasks. A similar correlation was also found between Hangul writing and WAB performance. This result suggests that AD affects performance on ‘parallel’ verbal and written language tasks [36,38]. Although the result offered information regarding the correlations between Hangul writing performance and visual and language functions, it remains to be seen how much of the impaired writing performance in the patients could be attributed to poor comprehension and maintenance of the stimuli during the writing-to-dictation task, and how much of it would be related to visuospatial disorders. Cleary, more research is needed to illuminate the relative role of multiple variables including working memory on the writing disturbance via regression analysis. The behavioral manifestations observed in the correlation analyses of Hangul writing performance were objectively confirmed via PET
Table 3 The distribution of writing errors over different subtypes of errors. Category
Subtypes
Frequency (%)
Linguistic errors
Graphemic omission Graphemic substitution Graphemic addition Stroke addition Stroke omission Graphemic misposition
14.1% 52.2% 5.4% 6.0% 11.7% 8.6%
Nonlinguistic errors
findings. Significant associations between Hangul writing performance (i.e., the total number of correct responses) and regional metabolic activity were identified in the left temporal lobes, including the middle and inferior temporal areas and the left angular gyrus. It has been suggested that writing ability is usually related to the LH and dysgraphia is attributed to damage in the temporal lobe, angular, and supramarginal gyri of the LH [2,31,39,40]. Recently, Nakamura et al. [41] demonstrated activation in the left posterior inferior temporal area during Japanese writing, replicating the results of the previous studies [2,31,39,40]. Further, the voxels with significant correlations with the total number of correct responses in Hangul writing were also located in the right inferior temporal and occipital lobes. The finding was the expected results considering visuospatial features of Hangul writing. These regions are known to be related with the functions of the visual memories [42], processing of visual materials [43], and orientation discrimination [44]. In Japanese, these areas have been speculated for the visual processing and retrieval of graphic image [45,46]. The areas might also play an important role for Hangul writing, which is likely to be involved in the effective visualization of graphemic forms. With the functional decline in the RH of the AD patients, the visuospatial aspects of Hangul syllables might have been compromised. In addition, a significant positive correlation between Hangul writing performance and glucose metabolism was also found in the right posterior cingulate gyrus. This area which is connected to the occipital lobe may be involved in visuospatial functions [47], and thus its lesion might be also associated with the visuospatial aspects of Hangul syllables. In contrast to our expectations, we did not find any correlation between writing and metabolism in the frontal and parietal lobes except for the left inferior parietal lobule. The whole writing process requires the coordination of linguistic and visuospatial-motor operations. In particular, the parietal lobe is involved in the ability to generate mental
Table 4 Partial correlations between the sub-scores of neuropsychological/language tests and the number of correct responses in Hangul writing. Correct responses in Hangul writing Visuospatial function RCFT: copy(N = 66)
.608**
Western Aphasia Battery (N = 75) Spontaneous speech Auditory comprehension Repetition Naming Reading Writing Aphasia quotient Language quotient
.534** .779** .654** .724** .856** .914** .735** .876**
RCFT: Rey–Osterrieth Complex Figure Test, N: the number of patients who performed the subtests. **p b .01.
76
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
Table 5 Regions of hypometabolism related to the correct response within the AD group (corrected pb .05, FDR, k=200), k: cluster size, FDR: false discovery rate. Regions
MNI coordinates
Rt. inferior temporal gyrus Rt. middle occipital lobe Lt. inferior parietal lobule Lt. inferior temporal gyrus Lt. superior temporal pole Lt. middle temporal gyrus Rt. posterior cingulate gyrus
X
Y
Z
52 40 −56 −56 −46 −58 2
−62 −72 −54 −60 6 2 −24
−8 16 34 6 −20 −16 40
Brodmann area
37 19 39 37 38 21 23
t-Value
6.06 4.65 5.35 5.30 5.34 4.86 4.83
Lt: left, Rt: right.
representations (motor engrams) of hand movements required for writing [2]. The premotor and motor regions in the frontal lobe translate these representations into the corresponding motor programs and appropriate hand movements [2]. We speculate that this unexpected result might reflect a false negative in the correlation analysis. AD is not a focal disease, and therefore, it is likely that each cognitive process under investigation requires the integrity of a much larger network of brain regions than we have observed. Due to the diffuse nature of the disease, we are only able to detect the strongest associations between cerebral metabolism and the writing process. These correlations could support the hypothesis that specific brain areas are involved in the writing process observed in patients with AD, but there are likely disruptions to a larger neural circuitry that underlies these deficits [48]. Another interpretation for the unexpected result is that we measured resting state regional brain glucose metabolism instead of measuring glucose metabolism while the patients are performing a writing-related cognitive task (e.g., imagining of writing words). One of the supportive findings to our hypothesis was that the area most significantly associated with RCFT performance was located in the right occipitotemporal lobe. These regions were partially overlapped with the regions of Hangul writing. This pattern supported the concept that visuospatial abilities in Hangul writing are related to drawing and copying and that Hangul writing may be processed in the RH [48,49]. When examining the correlation between the performance of AQ and cerebral metabolism, we observed an association with the left temporal lobe. This result is consistent with that of existing literature [48,50] and suggests that this area may be sensitive to both oral and written languages. In terms of the patterns of association, Hangul writing was related more with inferior regional metabolism, and AQ
Table 6 Regions of hypometabolism related to RCFT performance within the AD group (uncorrected pb .001, k=200), k: cluster size. Regions
Rt. inferior temporal gyrus Rt. middle occipital lobe
MNI coordinates X
Y
Z
50 42
−58 −76
−4 18
Brodmann area
t-Value
37 19
5.65 4.20
Rt: right.
had a higher relative metabolism in the middle region. This pattern is consistent with a previous finding that activation in the inferior temporal lobe may be specific to lexical orthographic processing [38]. In the previous study, the task of writing–naming contrast showed a focus of activation associated with writing, but not naming, located in this region [38]. Although identifying the pattern linking AQ and metabolism within the left temporal lobe was our main focus, associations with the right temporal lobe were also observed. This finding may reflect a functional role of this region in language and is consistent with previous work in AD [51,52].
5. Conclusions Our study is noteworthy because it is the first attempt to investigate the neural correlates of dysgraphia in Korean AD patients and yielded empirical evidence describing this correlation. The findings of this study provide objective support for the view that Hangul has both linguistic and visuospatial, nonlinguistic aspects and impairment in Hangul writing performance might be closely related to a functional decline in both the RH and LH. The study provides a unique contribution to the knowledge of dysgraphia in a non-alphabetical writing system as well as the underlying neuropathology of dysgraphic features in such languages. However, this study has not without limitations in that it mostly focused on the peripheral processes of writing and did not consider the central (e.g., lexical and phonological routes) or intermediate process (e.g., graphemic buffer) that could be involved in this patient population. Further studies will be needed to clarify lexical–semantic aspects and graphemic buffer deficits in the writing process. In addition to the underlying mechanisms involved between specific brain regions and Hangul writing, future studies are warranted to explore the specific brain regions related to each of these linguistic and nonlinguistic components in a larger number of patients.
Fig. 2. Regions of glucose metabolism (red color) related to (a) correct responses, (b) RCFT scores, and (c) AQ within the AD group.
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78 Table 7 Regions of hypometabolism related to the AQ scores within the AD group (corrected p b .05, FDR, k = 200), k: cluster size, FDR: false discovery rate. Regions
MNI coordinates
Lt. inferior parietal lobule Lt. middle temporal gyrus Rt. middle temporal gyrus Lt. posterior cingulate gyrus Lt. precentral gyrus Lt. middle frontal cortex
X
Y
Z
−56 −50 42 −4 −32 −34
−54 6 −68 −32 10 8
34 −20 14 38 44 36
Brodmann area
t-Value
39 21 21 23 6 44
8.21 5.25 7.73 5.35 3.60 3.40
Lt: left, Rt: right, AQ: aphasia quotient.
Fig. 3. Examples of (a) miscellaneous errors, (b) graphemic omission in the linguistic errors, and (c) graphemic misposition in the nonlinguistic errors.
Conflict of interest There is no conflict of interest. Appendix A The stimuli of the Hangul writing to dictation task. Category
Stimuli
Horizontal construction Vertical construction Mixed construction
References [1] Ellis AW. Spelling and writing (and reading and speaking). In: Ellis AW, editor. Normality and pathology in cognitive functions. London: Academic Press; 1982. p. 113–46. [2] Roeltgen DP. Agraphia. In: Heilman KM, Valenstein E, editors. Clinical neuropsychology. New York: University Press; 2003. p. 126–45. [3] McNeil MR, Tseng CH. Acquired neurogenic dysgraphias. In: LaPointe LL, editor. Aphasia and related neurogenic language disorders. New York: Thieme Medical Publishers; 1990. p. 147–76. [4] Rapcsak SZ, Beeson PM. The role of left posterior inferior temporal cortex in spelling. Neurology 2004;62:2221–9. [5] Alexander MP, Friedman RB, Loverso F, Fischer RS. Lesion location of phonological dysgraphia. Brain Lang 1992;43:83–95. [6] Cantagallo A, Bonazzi S. Acquired dysgraphia with selective damage to the buffer: a single case report. Ital J Neurol Sci 1996;17:249–54. [7] Rapp B, Caramazza A. From graphemes to abstract letter shapes: levels of representation in written spelling. J Exp Psychol 1997;23:1130–52.
77
[8] Hillis AE, Chang S, Breese E, Heidler J. The crucial role of posterior frontal regions in modality specific components of the spelling process. Neurocase 2004;10:175–87. [9] Kim H, Na DL. Dissociation of pure Korean words and Chinese-derivative words in phonological dysgraphia. Brain Lang 2000;74:134–7. [10] Kwon JC, Lee HJ, Chin J, Lee YM, Kim H, Na DL. Hanja alexia with dysgraphia after left posterior inferior temporal lobe infarction: a case study. J Korean Med Sci 2002;17: 91–5. [11] Kim HJ, Chu K, Lee KM, Kim DW, Park SH. Phonological dysgraphia after superior temporal gyrus infarction. Arch Neurol 2002;59:1314–6. [12] Kim H, Na DL, Park ES. Intransient vowel–consonant position in Korean dysgraphia: evidence of spatial–constructive representation. Behav Neurol 2007;18:91–7. [13] Yoon JH, Shin JC, Kim DY, Suh MK, Kim H. Korean dysgraphia subsequent to right hemispheric lesion. Speech Sci 2006;13:121–32. [14] Yoon JH, Suh MK, Kim H. Language specific dysgraphia in Korean stroke patients. Cogn Behav Neurol 2010;23:247–55. [15] Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 1991;82:239–59. [16] Yoon JH, Suh MK, Jeong Y, Ahn HJ, Moon SY, Chin J, et al. Agraphia in Korean patients with early onset Alzheimer's disease. Int Psychogeriatr 2011;23:1317–26. [17] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Report of the NINCDS–ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 1984;34:939. [18] Seo SK. Learner's dictionary of Korean. Seoul, Korea: Shinwon Prime; 2006. [19] Kim H, Na DL. Normative data on the Korean version of the Western Aphasia Battery. J Clin Exp Neuropsychol 2004;26:1011–20. [20] Kang Y, Na DL. Seoul neuropsychological screening battery. Incheon, Korea: Human Brain Research & Consulting Co; 2003. [21] Maeshima S, Sekiguchi E, Kakishita K, Okada H, Okita R, Ozaki F, et al. Dysgraphia with abnormal writing stroke sequences due to cerebral infarction. Brain Inj 2003;17:339–45. [22] Law SP. Writing error of a Cantonese dysgraphic patient and their theoretical implications. Neurocase 2004;10:132–40. [23] Seo SW, Cho SS, Park A, Chin J, Na DL. Subcortical vascular versus amnestic mild cognitive impairment: comparison of cerebral glucose metabolism. J Neuroimaging 2009;19:213–9. [24] Carthery MT, Mattos-Pimenta-Parente MA, Nitrini R, Bahia VS, Caramelli P. Spelling tasks and Alzheimer's disease staging. Eur J Neurol 2005;12:907–11. [25] Croisile B, Cannoi T, Adeleine P, Trillet M. Spelling in Alzheimer's disease. Behav Neurol 1995;8:135–43. [26] Neils J, Roeltgen DP, Greer A. Spelling and attention in early Alzheimer's disease: evidence for impairment of the graphemic buffer. Brain Lang 1995;49:241–62. [27] Lambert J, Eustache F, Viader F, Dary M, Rioux P, Lechevalier B. Agraphia in Alzheimer's disease: an independent lexical impairment. Brain Lang 1996;53:222–33. [28] Forbes KE, Shanks MF, Venneri A. The evolution of dysgraphia in Alzheimer's disease. Brain Res Bull 2004;63:19–24. [29] Croisile B, Brabant MJ, Carmoi T, Lepage Y, Aimard G, Trillet M. Comparison between oral and written spelling in Alzheimer's disease. Brain Lang 1996;54:361–87. [30] Rapcsak SZ, Arthur SA, Bliklen DA, Rubens AB. Lexical agraphia in Alzheimer's disease. Arch Neurol 1989;46:65–8. [31] Penniello MJ, Lambert J, Eustache F, Petit-Taboue MC, Barre L, Viader F, et al. A PET study of the functional neuroanatomy of writing impairment in Alzheimer's disease: the role of the left supramarginal and left angular gyri. Brain 1995;118:697–706. [32] Law SP, Leung MT. Structural representations of characters in Chinese writing: evidence from a case of acquired dysgraphia. Psychologia 2000;43:67–83. [33] Lambert J, Giffard B, Nore F, de la Sayette V, Pasquier F, Eustache F. Central and peripheral agraphia in Alzheimer's disease: from the case of Auguste D. to a cognitive neuropsychology approach. Cortex 2007;43:935–51. [34] Luzzatti C, Laiacona M, Agazzi D. Multiple patterns of writing disorders in dementia of the Alzheimer type and their evolution. Neuropsychologia 2003;41: 759–72. [35] Neils-Strunjas J, Shuren J, Roeltgen D, Brown C. Perseverative writing errors in a patient with Alzheimer's disease. Brain Lang 1998;63:303–20. [36] Hughes J, Graham N, Patterson K, Hodges JR. Agraphia in mild dementia of Alzheimer's type. Neuropsychologia 1997;35:533–45. [37] Platel H, Lambert J, Eustache F, Cadet B, Dary M, Viader F, et al. Characteristics and evolution of writing impairment in Alzheimer's disease. Neuropsychologia 1993;31: 1147–58. [38] Kavrie S, Neils-Strunjas J. Dysgraphia in Alzheimer's disease with mild cognitive impairment. J Med Speech Lang Pathol 2002;10:73–85. [39] Hillis AE, Kane A, Tuffiash E, Beauchamp NJ, Barker PB, Jacobs MA, et al. Neural substrates of the cognitive processes underlying spelling: evidence from MR diffusion and perfusion imaging. Aphasiology 2002;16:425–38. [40] Beeson PM, Rapcsak SZ, Plante E, Chargualaf J, Chung A, Johnson SC, et al. The neural substrates of writing: a functional magnetic resonance imaging study. Aphasiology 2003;17:647–66. [41] Nakamura K, Honda M, Okada T, Hanakawa T, Toma K, Fukuyama H, et al. Participation of the left posterior inferior temporal cortex in writing and mental recall of kanji orthography: a functional MRI study. Brain 2000;123:954–6. [42] Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends Neurosci 1992;15:20–5. [43] Martin A, Wiggs CL, Weisberg J. Modulation of human medial temporal lobe activity by form, meaning, and experience. Hippocampus 1997;7:587–93. [44] Macaluso E, Frith C. Interhemispheric differences in extrastriate areas during visuo-spatial selective attention. Neuroimage 2000;12:485–94. [45] Benson DF. Alexia. In: Vinken PJ, Bruyn GW, Klawans HL, editors. Handbook of clinical neurology. Amsterdam: Elsevier; 1985. p. 433–55.
78
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
[46] Coltheart M, Patterson K, Marshall JC, editors. Deep dyslexia. London: Routledge and Kegan Paul; 1987. p. 326–80. [47] Vogt B, Vogt L, Laureys S. Cytology and functionally correlated circuits of human posterior cingulate areas. Neuroimage 2006;29:452–66. [48] Teipel SJ, Willoch F, Ishii K, Bürger A, Drzezga R, Engel P, et al. Resting state glucose utilization and the CERAD cognitive battery in patients with Alzheimer's disease. Neurobiol Aging 2006;27:681–90. [49] Ober BA, Jagust WJ, Koss E, Delis DC, Friedland RP. Visuoconstructive performance and regional cerebral glucose metabolism in Alzheimer's disease. J Clin Exp Neuropsychol 1991;13:752–72.
[50] Hirono N, Mori E, Ishii K, Imamura T, Tanimukai S, Kazui H, et al. Neuronal substrates for semantic memory: a positron emission tomography study in Alzheimer's disease. Dement Geriatr Cogn Disord 2001;12:15–21. [51] Apostolova LG, Lu P, Rogers S, Dutton RA, Hayashi KM, Toga AW, et al. 3D mapping of language networks in clinical and pre-clinical Alzheimer's disease. Brain Lang 2008;104:33–41. [52] Melrose RJ, Campa OM, Harwood DG, Osato S, Mandelkern MA, Sultzer DL. The neural correlates of naming and fluency deficits in Alzheimer's disease: and FDG-PET study. Int J Geriatr Psychiatry 2009;24:885–93.
Contents lists available at SciVerse ScienceDirect
Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns
Dysgraphia in Korean patients with Alzheimer's disease as a manifestation of bilateral hemispheric dysfunction Ji Hye Yoon a, b, HyangHee Kim a, c,⁎, Sang Won Seo b, Juhee Chin b, Jung-Hyun Kim b, Kyung-Han Lee d, Yong Wook Kim c, Eun Sook Park c, Mee Kyung Suh b, Duk L. Na b,⁎⁎ a
Graduate Program in Speech and Language Pathology, Yonsei University, Seoul, Republic of Korea Department of Neurology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea Department and Research Institute of Rehabilitation Medicine, Yonsei University College of Medicine, Seoul, Republic of Korea d Department of Nuclear Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Republic of Korea b c
a r t i c l e
i n f o
Article history: Received 2 January 2012 Received in revised form 14 May 2012 Accepted 26 June 2012 Available online 16 July 2012 Keywords: Hangul Writing Dysgraphia Linguistic Nonlinguistic Visuospatial Alzheimer's disease PET
a b s t r a c t Background: In writing, linguistic (i.e., spelling) and nonlinguistic (i.e., arranging strokes or letters) functions are processed by the left and right hemispheres, respectively. The configuration of Korean alphabet, ‘Hangul’ invokes nonlinguistic, visuospatial functions that other writing systems use less extensively. Patients with Alzheimer's disease (AD) have bilateral involvement of temporoparietal–frontal areas that are responsible for processing language and visuospatial functions. Objectives: The aim of this study was to examine the nature of Hangul writing dysfunction, which may be associated with bilateral hemispheric impairments in AD. Methods: A sample of 75 patients with AD and 20 healthy controls (HC) performed a Hangul writing task. Neuroimaging positron emission tomography (PET) data of 22 patients were utilized to measure the regional brain glucose metabolism associated with Hangul writing. Results: The writing performance of the AD group was significantly reduced and different types of errors were observed as the disease got worse. Glucose hypometabolism correlated with Hangul writing impairment was located in the right occipitotemporal lobe and left temporoparietal lobe. Conclusions: The PET findings demonstrate that impairment in Hangul writing performance in Korean AD patients is closely related to a functional decline in both the right and left hemispheres. The study provides a unique contribution to the knowledge of dysgraphia in a non-alphabetical writing system as well as the underlying neuropathology of dysgraphic features in such languages. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Writing is not a unitary process, but requires coordination of linguistic and visuospatial aspects [1,2]. Traditionally, linguistic functions such as spelling are thought to be processed by the left hemisphere (LH), while visuospatial constructions in writing such as arranging strokes or letters are processed by the right hemisphere (RH) [3]. Therefore, each hemisphere may play a significant role in writing processing. In alphabetical written language systems, the letters are written horizontally. Writing is usually related to functional cortical activity, which appears to be predominant in the LH areas. Most previous studies on writing impairments of alphabetical writing systems have been designed to focus on LH lesions [4–8].
⁎ Correspondence to: H. Kim, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul 120‐752, Republic of Korea. Tel.: +82 2 2228 3900. ⁎⁎ Correspondence to: D.L. Na, Department of Neurology, Samsung Medical Center, 50 Ilwon-dong, Kangnam-ku, Seoul 135‐710, Republic of Korea. Tel.: +82 2 3410 3591. E-mail addresses: [email protected] (H. Kim), [email protected] (D.L. Na). 0022-510X/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2012.06.020
Such languages as Chinese (i.e., Hanja), Japanese (i.e., Kanji and Kana), and Korean (i.e., Hangul) utilize non-alphabetic writing systems. For example, Chinese and Kanji characters have complex graphic forms that consist of individual strokes, and the chunks of strokes are arranged in a diverse-dimensional square. The brain areas responsible for the non-alphabetic writing process may differ from those for alphabetically written language systems. The non-alphabetic writing language systems may be more related to RH functions than alphabetically written language systems. The Korean alphabet, Hangul (or Hangeul) is unique in its written application with respect to the nonlinguistic, visuospatial as well as linguistic aspects. Linguistically, the Hangul alphabet consists of 24 characters, split between 10 vowel-graphemes (‘mo-eum’) and 14 consonant-graphemes (‘ja-eum’). Each of the 24 Hangul graphemes corresponds to one phoneme. Visuospatially, each grapheme is placed within a square space to form a syllable. Hangul syllables are classified into three subtypes according to their visuospatial constructional patterns: (a) vertical- (written from top-to-bottom, e.g., ), (b) horizontal- (written from left-to-right, e.g., ), and (c)
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
mixed-construction (i.e., combination of horizontal and vertical constructions, e.g., ). Taking this syllabic rule into consideration, a word (e.g., ) could be created (Fig. 1). Thus, the configuration of Hangul syllables invokes visuospatial functions that other writing systems use less extensively. It can be hypothesized that this unique nature of Hangul may compromise Hangul writing functions in patients with bilateral hemispheric lesions differently compared with those with selective damage to a unilateral hemisphere. To date, however, studies on Hangul have focused on stroke patients with unilateral focal lesions [9–12]. The primary errors observed in unilateral stroke patients are linguistic errors such as substitution [10–13]. Stroke patients with RH lesions display prominent errors relevant to the physical form such as stroke omission and addition, and visuospatial errors such as graphemic misposition [13,14]. Alzheimer's disease (AD) is a disease with both visuospatial and language dysfunctions as part of the disease process due to the bilateral involvement of the temporoparietal–frontal lobes [15]. A previous study was conducted to explore the nature of writing dysfunction in Korean patients with early-onset AD [16]. The authors reported that these patients demonstrated linguistic and visuospatial errors in writing. In addition, the severity of dementia and multiple cognitive domains such as attention, language, immediate memory, and frontal executive functions significantly correlated with the Hangul writing performance. The aim of this study was to examine if the nature of Hangul writing dysfunction is associated with bilateral hemispheric impairment. We investigated the neural correlates of Hangul writing using positron emission tomography (PET) imaging analysis. By employing both behavioral and PET data results, we attempted to establish the relationship between the unique features of Korean writing system and bilateral brain lesion found in AD.
2. Methods 2.1. Participants Initially, we recruited 80 consecutive patients diagnosed with AD at the Memory Disorder Clinic at Samsung Medical Center in Seoul, Korea between September 2009 and August 2011. The AD patients fulfilled the criteria for probable AD proposed by the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS–ADRDA) [17]. The exclusion criteria were as follows: 1) those with less than six years of education because they might have had difficulty in writing Hangul in the premorbid state; 2) those with abnormal findings on laboratory tests including a complete blood count, blood chemistry, vitamin B12/folate, syphilis serology, and thyroid-function tests; 3) those with territorial cerebral infarctions, brain tumors, and other structural lesions on MRI; 4) those with mild bradykinesia and rigidity that might affect writing ability; 5) those who had refused the experimental testing.
73
Out of 80 patients, four declined to complete the experimental testing. One patient with mixed dementia was also excluded. As a result, 75 patients (male:female = 27:48) were included in this study. The mean age was 73.1 years (range: 46 to 92 years old). A panel consisting of two neurologists (D.L.N. and S.W.S.) and one neuropsychologist (J.C.) with expertise in dementia made the clinical decisions including diagnosis and clinical dementia rating (CDR) after reviewing the patients' clinical data. The patient group included 16 very mild (CDR 0.5), 37 mild (CDR 1), 19 moderate (CDR 2), and 3 severe (CDR 3) patients. All AD subjects were right-handed, spoke standard Korean, and reported no history of premorbid writing disturbances. In order to assess performances on the writing tests, we also recruited 20 age-and education-matched caregivers of the patients as healthy control (HC) group. They had no history of neurological or psychiatric illnesses. The demographic features of the 75 AD patients and 20 HC are presented in Table 1. 2.2. Materials 2.2.1. Hangul writing to dictation task All patients and HC completed a Hangul writing to dictation task. The dictation task comprises 60 single syllables extracted from multisyllabic words with one-to-one grapheme to phoneme correspondence [e.g., ‘ ’ (/ka/) of ‘ ’ (/ka-wi/) (meaning ‘scissors’)]. The selection of these syllables was based on the frequency, configuration, and imageability of the items. The multisyllabic words were selected from a pool of high frequency words in the Learner's dictionary of Korean [18]. In terms of configuration, characters with four graphemes or less were selected to reduce the complexity of the task. All words were concrete words with high imageability. For the Hangul writing to dictation task, the subjects wrote the stimuli on an A4 sheet of paper. A single syllable embedded in words was verbally presented to each patient. The subject was asked to verbally repeat the target stimuli before writing in order to rule out any deficit in auditory input processing that might influence writing to dictation. No time limit was imposed. The stimuli are provided in Appendix A. 2.2.2. Language test All patients completed the Western Aphasia Battery (WAB) [19]. The oral language subtests such as spontaneous speech, auditory comprehension, repetition, and naming were used to determine the type of aphasia and its severity. An Aphasia Quotient (AQ) was derived from the raw scores. The two subtests, reading and writing, were used to assess written language ability. Adding the reading and writing scores, the Language Quotient (LQ) was obtained. The AD patients showed the mean AQ of 82.1 and LQ of 76.5 (maximum score = 100). 2.2.3. Figure copying test Visuospatial ability is also regarded as an important function in writing. In order to investigate the relationship between general visuospatial function and writing, 66 out of 75 AD patients underwent the Rey–Osterrieth Complex Figure Test (RCFT) using a standardized neuropsychological battery; the Seoul Neuropsychological Screening Battery (SNSB) [20]. The data of the nine patients could not be obtained Table 1 Demographic variables for the HC and AD groups (Mean ± SD).
Age (years) Sex: female, N (%) Education (years) MMSE CDR Fig. 1. Letters in English words are written in a linear fashion, while graphemes in Korean words are arranged in a square pattern.
HC (N = 20)
AD (N = 75)
71.5 ± 7.1 12(60%) 10.9 ± 3.7 29.0 ± 1.0 N/A
73.1 ± 9.5 48(64%) 11.1 ± 3.7 17.5 ± 5.8⁎ 1.2 ± 0.6
HC: healthy control, AD: Alzheimer's disease, MMSE: Mini-Mental State Examination, CDR: Clinical Dementia Rating; N/A: not applicable. ⁎ p b .05.
74
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
because they declined to undergo the test or because the time gap between the test and the Hangul writing experiment was over six months. The mean score on the RCFT was 23.1 in the patient group (maximum score= 36). 2.3. Writing error analysis The number of correctly produced Hangul syllables was counted. We regarded a response as correct if all graphemes in a syllable were appropriate with respect to the linguistic and nonlinguisitc aspects. The maximum possible score was 60 and the minimum was 0. We additionally analyzed error patterns based on the criteria proposed by previous studies [10,13,14,21,22]. The linguistic errors consisted of graphemic omission (e.g., ), substitution (e.g., ), and addition (e.g., ). The nonlinguistic errors consisted of awkward shaping and implausible configurations of Korean graphemes or syllables such as stroke omission, stroke addition, and graphemic misposition. By definition, stroke omission errors mean deletion of part(s) of a grapheme (e.g., ). Stroke addition errors occur when redundant strokes are added to an original grapheme (e.g., ). Errors of graphemic misposition (i.e., a slight shift of a grapheme) create a nonexistent form within the boundaries of the Korean syllable (e.g., ). More than one error could be counted per syllable because the errors were examined on each grapheme. In addition to the linguistic and nonlinguistic errors, mixed (linguistic as well as nonlinguistic) and miscellaneous (e.g., unintelligible responses, figure drawing, no response) were also recorded. 2.4. Statistics A t-test was conducted to confirm the differences in the total number of correct or error responses between the AD and HC groups. One-way analysis of variance and post hoc Bonferroni correction were conducted to identify any differences in number of erroneous responses according to disease severity within the AD group (p b .05). To evaluate the relationship between the figure copying or the language results and Hangul writing in the AD group, partial correlation analysis was conducted after adjusting for age, sex, and duration of education. 2.5. Positron emission tomography (PET) imaging analysis Resting state regional brain glucose metabolism was measured using [1F]fluoro-2-deoxy-D-glucose (18FDG). The participants fasted for at least 6 h before the scan. Thirty-minute positron emission tomography (PET) scans were acquired 40 min after the intravenous injection of 4.8 MBq/kg FDG using a GE Advance PET scanner. The protocol for obtaining the PET scan was published elsewhere [23]. Of the 75 patients, 22 underwent FDG-PET. The rest of the patients either did not undergo PET imaging within six months of performing the Hangul writing test or declined to undergo imaging due to personal reasons. In terms of AD severity, there were 3 very mild (CDR 0.5), 8 mild (CDR 1), 9 moderate (CDR 2), and 2 severe (CDR 3) patients. The mean age (73.4 ± 8.8 years), sex ratio (eight males), and education (11.9 ± 3.4 years) of patients with FDG-PET did not differ from those that did not undergo FDG-PET. The 22 patients exhibited the mean AQ of 77.9 and LQ of 73.0 (maximum score = 100). The mean AQ and LQ of patients with FDG-PET did not differ from those without. The PET study was approved by the Institutional Review Board (IRB; no. 2006-03-011) of Samsung Medical Center. The PET images were analyzed using SPM5 (Wellcome Department of Cognitive Neurology, Institute of Neurology, London, UK) and implemented using Matlab 7.0 (MathWorks Inc., Sherborn, MA, USA). Prior to the statistical analysis, all of the images were spatially normalized into the MNI standard template (Montreal Neurological Institute, McGill University, Montreal, Canada) in order to remove inter-subject anatomic variability. Spatially normalized images
were smoothed with a 16 mm FWHM isotropic Gaussian kernel. The count of each voxel was normalized to the average count of the cerebellum with proportional scaling in SPM5. In order to investigate hypometabolic brain areas that were associated with the performance, we performed multiple linear regression analysis within the patient group (corrected p b .05; FDR, false discovery rate) after adjusting for age and educational factors. To visualize the t-score statistics (SPM1 map), the significant voxels were projected onto the 3-D rendered brain or a standard high-resolution MRI template provided by SPM5. 3. Results 3.1. Performance of the Hangul writing task In terms of the mean total number of correct responses between the AD and HC groups, the performance of the AD group (M=41.1, SD = 18.3) was significantly reduced (p b .001) than that of the HC group (M= 58.6, SD = 1.3). When the total erroneous responses between the two groups were compared, the performance of the AD group (M= 32.5, SD=44.5) was significantly worse than that of the HC group (M = 1.3, SD = 1.4) (p b .001). Within the AD group, the number of total erroneous responses significantly differed according to CDR (p b .001). In the post hoc analysis, there were significant differences between CDR 0.5 and CDR 2 (p = .003), CDR 0.5 and CDR 3 (p = .017), CDR 1 and CDR 2 (p = .001), and between CDR 1 and CDR 3 (p = .021) (Table 2). No differences were observed between CDR 0.5 and CDR 1 and between CDR 2 and CDR 3. With respect to the error patterns, CDRs of 0.5 and 1 yielded more linguistic errors, while CDRs of 2 and 3 demonstrated more miscellaneous errors. The distribution of writing errors over different subtypes of error was presented in Table 3. The examples according to the error subtypes are presented in Fig. 3. 3.2. The relationship between the copying figure or the language test and the writing task performance The number of correct responses in the writing task was correlated with the RCFT scores. In addition, the number of correct responses in the writing task was correlated with the scores of all WAB subtests. The results are provided in Table 4. 3.3. SPM analysis of FDG-PET The total number of correct responses (M= 37.1, SD = 20.9) of patients with FDG-PET did not differ from those without. The voxels for which glucose metabolism positively correlated with the number of correct responses were located in the right occipitotemporal lobe and left temporoparietal lobe (Table 5, Fig. 2a). The voxels for which glucose metabolism positively correlated with the RCFT performance were located in the right occipitotemporal lobe (Table 6, Fig. 2b). The voxels for which glucose metabolism positively correlated with AQ were located in the left temporal, parietal, and frontal lobes (Table 7, Fig. 2c). 4. Discussion Due to the bilateral hemispheric compromise of AD, research on writing in this clinical population might provide significant information on the neural correlates of writing. Hangul writing is distinctive in that it has unique linguistic and visuospatial, nonlinguistic features. From the results of the current study, we found that the AD patients made linguistic errors including graphemic omission, graphemic substitution, and graphemic addition. These errors are not unique and are common errors in any written language such as English, Italian, and French [24–31]. In alphabetic writings, transposition errors could be observed between a vowel and a consonant and between a vowel and a
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
75
Table 2 The number of erroneous responses according to CDR (Mean ± SD).
Linguistic Nonlinguistic Mixed Miscellaneous N
HC (N = 20)
CDR 0.5 (N = 16)
CDR 1 (N = 37)
CDR 2 (N = 19)
CDR3 (N = 3)
Total AD (N = 75)
1.3 ± 1.4 0.0 ± 0.0 0.0 ± 0.0 0.0 ± 0.0 1.3 ± 1.4
9.6 ± 11.5⁎ 2.1 ± 2.9 0.5 ± 0.9 2.4 ± 4.1 14.8 ± 15.9⁎
10.8 ± 12.9⁎ 2.3 ± 3.1 2.8 ± 11.0 4.0 ± 15.3 20.0 ± 28.6⁎
20.0 ± 19.4⁎ 7.8 ± 12.7⁎†‡ 5.9 ± 17.0 28.9 ± 55.5⁎ 62.6 ± 58.4⁎†‡
22.0 ± 33.0⁎ 8.0 ± 13.0⁎†‡ 0.0 ± 0.0 60.6 ± 94.8⁎†‡ 90.6 ± 81.5⁎†‡
13.2 ± 15.8⁎ 3.8 ± 7.5 2.9 ± 11.5 12.2 ± 36.5⁎ 32.5 ± 44.5⁎
HC: healthy control, AD: Alzheimer's disease, N: number of erroneous responses. ⁎ p b .05 between HC and CDR 0.5, HC and CDR 1, HC and CDR 2, HC and CDR 3, and between HC and AD group. † p b .05 between CDR 0.5 and CDR 2 and between CDR 0.5 and CDR 3. ‡ p b .05 between CDR 1 and CDR 2, and between CDR 1 and CDR 3.
vowel, and consonant and a consonant. Transposition errors of character components such as constituents were also observed in Chinese [22,32]. Nonlinguistic errors of writing in AD were confined to letter formation in previous studies on alphabetic writings [28,33–37], although we found stroke addition and omission as nonlinguistic errors in this study. However, among the nonlinguistic errors, graphemic misposition errors (Fig. 3c) are language-specific in that misposition can result in nonexistent and meaningless forms in Korean syllabic writing. This language-specific error has also been observed in Korean patients with stroke [13,14] and in patients with early onset AD [16]. This is attributed to the fact that each grapheme of Hangul has its own allocated space within a square syllabic form (Fig. 1). This is in contrast to alphabetic writings, which are arranged in a horizontal, linear fashion to form a syllable. A slight horizontal or vertical shift of a grapheme (e.g., ‘r’ in ‘root’) in alphabetical writing would not change the lexical meaning. The visuospatially unique feature of Hangul writing is related to performance of visuoconstructional tasks such as RCFT. In the current study, the performance of the RCFT task was highly associated with that of the Hangul writing task. Generally, figure copying might be accomplished through a pictorial route (e.g., slavish or ‘stroke-by-stroke’ copying), while Hangul writing is generated by a graphemic (spelling and writing) route. Unlike copying in the pictorial route, graphemic Hangul writing can take advantage of graphemic knowledge stored in long-term memory. However, these two tasks call for visuospatial abilities during executing of movement for performance, regardless of the route involved. This finding supports the notion that Hangul writing needs visuospatial functions which are also required in picture drawing/copying tasks. A similar correlation was also found between Hangul writing and WAB performance. This result suggests that AD affects performance on ‘parallel’ verbal and written language tasks [36,38]. Although the result offered information regarding the correlations between Hangul writing performance and visual and language functions, it remains to be seen how much of the impaired writing performance in the patients could be attributed to poor comprehension and maintenance of the stimuli during the writing-to-dictation task, and how much of it would be related to visuospatial disorders. Cleary, more research is needed to illuminate the relative role of multiple variables including working memory on the writing disturbance via regression analysis. The behavioral manifestations observed in the correlation analyses of Hangul writing performance were objectively confirmed via PET
Table 3 The distribution of writing errors over different subtypes of errors. Category
Subtypes
Frequency (%)
Linguistic errors
Graphemic omission Graphemic substitution Graphemic addition Stroke addition Stroke omission Graphemic misposition
14.1% 52.2% 5.4% 6.0% 11.7% 8.6%
Nonlinguistic errors
findings. Significant associations between Hangul writing performance (i.e., the total number of correct responses) and regional metabolic activity were identified in the left temporal lobes, including the middle and inferior temporal areas and the left angular gyrus. It has been suggested that writing ability is usually related to the LH and dysgraphia is attributed to damage in the temporal lobe, angular, and supramarginal gyri of the LH [2,31,39,40]. Recently, Nakamura et al. [41] demonstrated activation in the left posterior inferior temporal area during Japanese writing, replicating the results of the previous studies [2,31,39,40]. Further, the voxels with significant correlations with the total number of correct responses in Hangul writing were also located in the right inferior temporal and occipital lobes. The finding was the expected results considering visuospatial features of Hangul writing. These regions are known to be related with the functions of the visual memories [42], processing of visual materials [43], and orientation discrimination [44]. In Japanese, these areas have been speculated for the visual processing and retrieval of graphic image [45,46]. The areas might also play an important role for Hangul writing, which is likely to be involved in the effective visualization of graphemic forms. With the functional decline in the RH of the AD patients, the visuospatial aspects of Hangul syllables might have been compromised. In addition, a significant positive correlation between Hangul writing performance and glucose metabolism was also found in the right posterior cingulate gyrus. This area which is connected to the occipital lobe may be involved in visuospatial functions [47], and thus its lesion might be also associated with the visuospatial aspects of Hangul syllables. In contrast to our expectations, we did not find any correlation between writing and metabolism in the frontal and parietal lobes except for the left inferior parietal lobule. The whole writing process requires the coordination of linguistic and visuospatial-motor operations. In particular, the parietal lobe is involved in the ability to generate mental
Table 4 Partial correlations between the sub-scores of neuropsychological/language tests and the number of correct responses in Hangul writing. Correct responses in Hangul writing Visuospatial function RCFT: copy(N = 66)
.608**
Western Aphasia Battery (N = 75) Spontaneous speech Auditory comprehension Repetition Naming Reading Writing Aphasia quotient Language quotient
.534** .779** .654** .724** .856** .914** .735** .876**
RCFT: Rey–Osterrieth Complex Figure Test, N: the number of patients who performed the subtests. **p b .01.
76
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
Table 5 Regions of hypometabolism related to the correct response within the AD group (corrected pb .05, FDR, k=200), k: cluster size, FDR: false discovery rate. Regions
MNI coordinates
Rt. inferior temporal gyrus Rt. middle occipital lobe Lt. inferior parietal lobule Lt. inferior temporal gyrus Lt. superior temporal pole Lt. middle temporal gyrus Rt. posterior cingulate gyrus
X
Y
Z
52 40 −56 −56 −46 −58 2
−62 −72 −54 −60 6 2 −24
−8 16 34 6 −20 −16 40
Brodmann area
37 19 39 37 38 21 23
t-Value
6.06 4.65 5.35 5.30 5.34 4.86 4.83
Lt: left, Rt: right.
representations (motor engrams) of hand movements required for writing [2]. The premotor and motor regions in the frontal lobe translate these representations into the corresponding motor programs and appropriate hand movements [2]. We speculate that this unexpected result might reflect a false negative in the correlation analysis. AD is not a focal disease, and therefore, it is likely that each cognitive process under investigation requires the integrity of a much larger network of brain regions than we have observed. Due to the diffuse nature of the disease, we are only able to detect the strongest associations between cerebral metabolism and the writing process. These correlations could support the hypothesis that specific brain areas are involved in the writing process observed in patients with AD, but there are likely disruptions to a larger neural circuitry that underlies these deficits [48]. Another interpretation for the unexpected result is that we measured resting state regional brain glucose metabolism instead of measuring glucose metabolism while the patients are performing a writing-related cognitive task (e.g., imagining of writing words). One of the supportive findings to our hypothesis was that the area most significantly associated with RCFT performance was located in the right occipitotemporal lobe. These regions were partially overlapped with the regions of Hangul writing. This pattern supported the concept that visuospatial abilities in Hangul writing are related to drawing and copying and that Hangul writing may be processed in the RH [48,49]. When examining the correlation between the performance of AQ and cerebral metabolism, we observed an association with the left temporal lobe. This result is consistent with that of existing literature [48,50] and suggests that this area may be sensitive to both oral and written languages. In terms of the patterns of association, Hangul writing was related more with inferior regional metabolism, and AQ
Table 6 Regions of hypometabolism related to RCFT performance within the AD group (uncorrected pb .001, k=200), k: cluster size. Regions
Rt. inferior temporal gyrus Rt. middle occipital lobe
MNI coordinates X
Y
Z
50 42
−58 −76
−4 18
Brodmann area
t-Value
37 19
5.65 4.20
Rt: right.
had a higher relative metabolism in the middle region. This pattern is consistent with a previous finding that activation in the inferior temporal lobe may be specific to lexical orthographic processing [38]. In the previous study, the task of writing–naming contrast showed a focus of activation associated with writing, but not naming, located in this region [38]. Although identifying the pattern linking AQ and metabolism within the left temporal lobe was our main focus, associations with the right temporal lobe were also observed. This finding may reflect a functional role of this region in language and is consistent with previous work in AD [51,52].
5. Conclusions Our study is noteworthy because it is the first attempt to investigate the neural correlates of dysgraphia in Korean AD patients and yielded empirical evidence describing this correlation. The findings of this study provide objective support for the view that Hangul has both linguistic and visuospatial, nonlinguistic aspects and impairment in Hangul writing performance might be closely related to a functional decline in both the RH and LH. The study provides a unique contribution to the knowledge of dysgraphia in a non-alphabetical writing system as well as the underlying neuropathology of dysgraphic features in such languages. However, this study has not without limitations in that it mostly focused on the peripheral processes of writing and did not consider the central (e.g., lexical and phonological routes) or intermediate process (e.g., graphemic buffer) that could be involved in this patient population. Further studies will be needed to clarify lexical–semantic aspects and graphemic buffer deficits in the writing process. In addition to the underlying mechanisms involved between specific brain regions and Hangul writing, future studies are warranted to explore the specific brain regions related to each of these linguistic and nonlinguistic components in a larger number of patients.
Fig. 2. Regions of glucose metabolism (red color) related to (a) correct responses, (b) RCFT scores, and (c) AQ within the AD group.
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78 Table 7 Regions of hypometabolism related to the AQ scores within the AD group (corrected p b .05, FDR, k = 200), k: cluster size, FDR: false discovery rate. Regions
MNI coordinates
Lt. inferior parietal lobule Lt. middle temporal gyrus Rt. middle temporal gyrus Lt. posterior cingulate gyrus Lt. precentral gyrus Lt. middle frontal cortex
X
Y
Z
−56 −50 42 −4 −32 −34
−54 6 −68 −32 10 8
34 −20 14 38 44 36
Brodmann area
t-Value
39 21 21 23 6 44
8.21 5.25 7.73 5.35 3.60 3.40
Lt: left, Rt: right, AQ: aphasia quotient.
Fig. 3. Examples of (a) miscellaneous errors, (b) graphemic omission in the linguistic errors, and (c) graphemic misposition in the nonlinguistic errors.
Conflict of interest There is no conflict of interest. Appendix A The stimuli of the Hangul writing to dictation task. Category
Stimuli
Horizontal construction Vertical construction Mixed construction
References [1] Ellis AW. Spelling and writing (and reading and speaking). In: Ellis AW, editor. Normality and pathology in cognitive functions. London: Academic Press; 1982. p. 113–46. [2] Roeltgen DP. Agraphia. In: Heilman KM, Valenstein E, editors. Clinical neuropsychology. New York: University Press; 2003. p. 126–45. [3] McNeil MR, Tseng CH. Acquired neurogenic dysgraphias. In: LaPointe LL, editor. Aphasia and related neurogenic language disorders. New York: Thieme Medical Publishers; 1990. p. 147–76. [4] Rapcsak SZ, Beeson PM. The role of left posterior inferior temporal cortex in spelling. Neurology 2004;62:2221–9. [5] Alexander MP, Friedman RB, Loverso F, Fischer RS. Lesion location of phonological dysgraphia. Brain Lang 1992;43:83–95. [6] Cantagallo A, Bonazzi S. Acquired dysgraphia with selective damage to the buffer: a single case report. Ital J Neurol Sci 1996;17:249–54. [7] Rapp B, Caramazza A. From graphemes to abstract letter shapes: levels of representation in written spelling. J Exp Psychol 1997;23:1130–52.
77
[8] Hillis AE, Chang S, Breese E, Heidler J. The crucial role of posterior frontal regions in modality specific components of the spelling process. Neurocase 2004;10:175–87. [9] Kim H, Na DL. Dissociation of pure Korean words and Chinese-derivative words in phonological dysgraphia. Brain Lang 2000;74:134–7. [10] Kwon JC, Lee HJ, Chin J, Lee YM, Kim H, Na DL. Hanja alexia with dysgraphia after left posterior inferior temporal lobe infarction: a case study. J Korean Med Sci 2002;17: 91–5. [11] Kim HJ, Chu K, Lee KM, Kim DW, Park SH. Phonological dysgraphia after superior temporal gyrus infarction. Arch Neurol 2002;59:1314–6. [12] Kim H, Na DL, Park ES. Intransient vowel–consonant position in Korean dysgraphia: evidence of spatial–constructive representation. Behav Neurol 2007;18:91–7. [13] Yoon JH, Shin JC, Kim DY, Suh MK, Kim H. Korean dysgraphia subsequent to right hemispheric lesion. Speech Sci 2006;13:121–32. [14] Yoon JH, Suh MK, Kim H. Language specific dysgraphia in Korean stroke patients. Cogn Behav Neurol 2010;23:247–55. [15] Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 1991;82:239–59. [16] Yoon JH, Suh MK, Jeong Y, Ahn HJ, Moon SY, Chin J, et al. Agraphia in Korean patients with early onset Alzheimer's disease. Int Psychogeriatr 2011;23:1317–26. [17] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Report of the NINCDS–ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology 1984;34:939. [18] Seo SK. Learner's dictionary of Korean. Seoul, Korea: Shinwon Prime; 2006. [19] Kim H, Na DL. Normative data on the Korean version of the Western Aphasia Battery. J Clin Exp Neuropsychol 2004;26:1011–20. [20] Kang Y, Na DL. Seoul neuropsychological screening battery. Incheon, Korea: Human Brain Research & Consulting Co; 2003. [21] Maeshima S, Sekiguchi E, Kakishita K, Okada H, Okita R, Ozaki F, et al. Dysgraphia with abnormal writing stroke sequences due to cerebral infarction. Brain Inj 2003;17:339–45. [22] Law SP. Writing error of a Cantonese dysgraphic patient and their theoretical implications. Neurocase 2004;10:132–40. [23] Seo SW, Cho SS, Park A, Chin J, Na DL. Subcortical vascular versus amnestic mild cognitive impairment: comparison of cerebral glucose metabolism. J Neuroimaging 2009;19:213–9. [24] Carthery MT, Mattos-Pimenta-Parente MA, Nitrini R, Bahia VS, Caramelli P. Spelling tasks and Alzheimer's disease staging. Eur J Neurol 2005;12:907–11. [25] Croisile B, Cannoi T, Adeleine P, Trillet M. Spelling in Alzheimer's disease. Behav Neurol 1995;8:135–43. [26] Neils J, Roeltgen DP, Greer A. Spelling and attention in early Alzheimer's disease: evidence for impairment of the graphemic buffer. Brain Lang 1995;49:241–62. [27] Lambert J, Eustache F, Viader F, Dary M, Rioux P, Lechevalier B. Agraphia in Alzheimer's disease: an independent lexical impairment. Brain Lang 1996;53:222–33. [28] Forbes KE, Shanks MF, Venneri A. The evolution of dysgraphia in Alzheimer's disease. Brain Res Bull 2004;63:19–24. [29] Croisile B, Brabant MJ, Carmoi T, Lepage Y, Aimard G, Trillet M. Comparison between oral and written spelling in Alzheimer's disease. Brain Lang 1996;54:361–87. [30] Rapcsak SZ, Arthur SA, Bliklen DA, Rubens AB. Lexical agraphia in Alzheimer's disease. Arch Neurol 1989;46:65–8. [31] Penniello MJ, Lambert J, Eustache F, Petit-Taboue MC, Barre L, Viader F, et al. A PET study of the functional neuroanatomy of writing impairment in Alzheimer's disease: the role of the left supramarginal and left angular gyri. Brain 1995;118:697–706. [32] Law SP, Leung MT. Structural representations of characters in Chinese writing: evidence from a case of acquired dysgraphia. Psychologia 2000;43:67–83. [33] Lambert J, Giffard B, Nore F, de la Sayette V, Pasquier F, Eustache F. Central and peripheral agraphia in Alzheimer's disease: from the case of Auguste D. to a cognitive neuropsychology approach. Cortex 2007;43:935–51. [34] Luzzatti C, Laiacona M, Agazzi D. Multiple patterns of writing disorders in dementia of the Alzheimer type and their evolution. Neuropsychologia 2003;41: 759–72. [35] Neils-Strunjas J, Shuren J, Roeltgen D, Brown C. Perseverative writing errors in a patient with Alzheimer's disease. Brain Lang 1998;63:303–20. [36] Hughes J, Graham N, Patterson K, Hodges JR. Agraphia in mild dementia of Alzheimer's type. Neuropsychologia 1997;35:533–45. [37] Platel H, Lambert J, Eustache F, Cadet B, Dary M, Viader F, et al. Characteristics and evolution of writing impairment in Alzheimer's disease. Neuropsychologia 1993;31: 1147–58. [38] Kavrie S, Neils-Strunjas J. Dysgraphia in Alzheimer's disease with mild cognitive impairment. J Med Speech Lang Pathol 2002;10:73–85. [39] Hillis AE, Kane A, Tuffiash E, Beauchamp NJ, Barker PB, Jacobs MA, et al. Neural substrates of the cognitive processes underlying spelling: evidence from MR diffusion and perfusion imaging. Aphasiology 2002;16:425–38. [40] Beeson PM, Rapcsak SZ, Plante E, Chargualaf J, Chung A, Johnson SC, et al. The neural substrates of writing: a functional magnetic resonance imaging study. Aphasiology 2003;17:647–66. [41] Nakamura K, Honda M, Okada T, Hanakawa T, Toma K, Fukuyama H, et al. Participation of the left posterior inferior temporal cortex in writing and mental recall of kanji orthography: a functional MRI study. Brain 2000;123:954–6. [42] Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends Neurosci 1992;15:20–5. [43] Martin A, Wiggs CL, Weisberg J. Modulation of human medial temporal lobe activity by form, meaning, and experience. Hippocampus 1997;7:587–93. [44] Macaluso E, Frith C. Interhemispheric differences in extrastriate areas during visuo-spatial selective attention. Neuroimage 2000;12:485–94. [45] Benson DF. Alexia. In: Vinken PJ, Bruyn GW, Klawans HL, editors. Handbook of clinical neurology. Amsterdam: Elsevier; 1985. p. 433–55.
78
J.H. Yoon et al. / Journal of the Neurological Sciences 320 (2012) 72–78
[46] Coltheart M, Patterson K, Marshall JC, editors. Deep dyslexia. London: Routledge and Kegan Paul; 1987. p. 326–80. [47] Vogt B, Vogt L, Laureys S. Cytology and functionally correlated circuits of human posterior cingulate areas. Neuroimage 2006;29:452–66. [48] Teipel SJ, Willoch F, Ishii K, Bürger A, Drzezga R, Engel P, et al. Resting state glucose utilization and the CERAD cognitive battery in patients with Alzheimer's disease. Neurobiol Aging 2006;27:681–90. [49] Ober BA, Jagust WJ, Koss E, Delis DC, Friedland RP. Visuoconstructive performance and regional cerebral glucose metabolism in Alzheimer's disease. J Clin Exp Neuropsychol 1991;13:752–72.
[50] Hirono N, Mori E, Ishii K, Imamura T, Tanimukai S, Kazui H, et al. Neuronal substrates for semantic memory: a positron emission tomography study in Alzheimer's disease. Dement Geriatr Cogn Disord 2001;12:15–21. [51] Apostolova LG, Lu P, Rogers S, Dutton RA, Hayashi KM, Toga AW, et al. 3D mapping of language networks in clinical and pre-clinical Alzheimer's disease. Brain Lang 2008;104:33–41. [52] Melrose RJ, Campa OM, Harwood DG, Osato S, Mandelkern MA, Sultzer DL. The neural correlates of naming and fluency deficits in Alzheimer's disease: and FDG-PET study. Int J Geriatr Psychiatry 2009;24:885–93.