عوامل موثر در اختلال معنایی عمل شی برای افراد چینی زبان با یا بدون بیماری آلزایمر

The present study aimed to elucidate the nature of action-object semantic impairment, as revealed in a category fluency task and a picture naming task completed by Chinese-speaking persons with or without Alzheimer's disease (AD). Also, the predictive power of semantic variables on the severity of dementia was addressed. Speech samples were collected from twenty AD persons, twenty control seniors and twenty control adults in Taiwan. Each participant individually completed two tasks: a category fluency task and a picture naming task. Results of the category fluency task indicated that the content of information in Chinese-speaking AD participants was seriously deteriorated, thus producing significantly smaller number of semantic-lexical items. Category Effect was only significant for healthy controls in the semantic fluency task; no significant Category Effect was found in AD participants. Additionally, results of Multiple regression analyses demonstrated that action fluency, action naming and frequency of pictures could reliably predict the severity of dementia. Findings in the present study helped characterize the nature of semantic disorders in Chinese-speaking AD participants and contributed to the predictive power of semantic variables and of pictorial variables on the severity of dementia.

Over the past decades, persons with Alzheimer's disease (AD) have been characterized by multiple cognitive deficits, the most important of which is the semantic-lexical disorder (Gainotti, Di Betta, & Silveri, 1996). Much research on AD patients' semantic-lexical impairment has been conducted concerning three issues: (a) the nature of deficit, (b) semantic category effects, and (c) pictorial attribute effects. First of all, the nature of semantic-lexical deficit for AD individuals remains under debate. Some researchers (cf. Abeysinghe, Bayles, & Trosset, 1990; Grober, Buschke, Kawas, & Fuld, 1985; Hodges, Salmon, & Butters, 1992) argue that semantic-lexical disorders result primarily from a loss of information in semantic representations. The content of information stored in AD individuals is reported to be seriously deteriorated. Semantic fluency task is often utilized to fully explore the knowledge of a certain semantic target. Gainotti et al. (1996), for example, asked sixteen AD persons and eleven matched controls to generate in 1 min as many words as possible that were related to a stimulus target. In comparison to normal controls, AD participants produced significantly less amount of information about semantic-lexical items. The “loss of information” hypothesis was empirically supported with a selective loss of specific associates in AD utterances. Other researchers claim that semantic information is intact for AD individuals who are mainly impaired in the conscious access to that information (cf. Nebes, 1989; Nebes & Brady, 1988, 1990; Nebes, Martin, & Horn, 1984; Nicholas, Olber, Albert, & Goodglass, 1985). Nebes et al. conducted a series of studies on semantic processing. Nebes and Brady (1990), for instance, investigated the organization of semantic attributes in Alzheimer's disease, revealing that attribute dominance had a greater impact on AD participants' performance than on normal controls' performance. It was concluded that knowledge of relative importance was retained and that basic organization of semantic attributes was preserved in AD participants. Accounts of naming impairment in normal aging also supported this argument. Nicholas, Olber, Albert, and Goodglass (1985) addressed the issue of lexical retrieval for nouns and verbs in healthy aging with picture naming tests. Results indicated that the ability to name both word types declined with age and that more errors were made on object names especially for older participants. Analysis of response type difference reflected greater difficulties for healthy elders in quantitative ways, not in qualitative ways. This finding led to the conclusion that naming difficulty for normal aging was at the label retrieval stage. The second issue regarding AD persons' semantic-lexical impairment addresses semantic category effects. Semantic categories have been known to highly correlate with semantic memory deficits. In the past decades, neuropsychological studies have been conducted to examine whether AD patients perform differently in different semantic categories and to elucidate the mechanisms of brain functions and linguistic concept formation. A recent direction on semantic category effects on AD individuals has addressed object/noun naming versus action/verb naming, but findings have not gone undisputed, as reviewed in Vigliocco, Vinson, Druks, Barber, and Cappa (2011). Some studies (Bowles, Obler, & Albert, 1987; Irigaray, 1973; Robinson, Rossor, & Cipolotti, 1999; Williamson, Adair, Raymer, & Heilman, 1998) indicate that the production of verbs is better preserved than that of nouns. Others (Cappa et al., 1998; Druks et al., 2006; Kim & Thompson, 2004; Lee et al., 1998; Mätzig, Druks, Masterson, & Vigliocco, 2009; Robinson, Grossman, White-Devine, & D'Esposito, 1996) argue that action naming is more impaired than object naming in the memory tests. Still others (Almor et al., 2009; Wang, 2010) claim that graceful degradation is found for both nouns and verbs and feature-based semantic representations help account for the impairment. Empirical support for the object impairment with relative preservation in verbs in AD persons comes primarily from systematic naming studies (Bowles et al., 1987; Irigaray, 1973; Robinson et al., 1999; Williamson et al., 1998; Yang et al., 2006). Investigating the spontaneous speech of dementia individuals, Irigaray (1973) found that more verbs were retained than nouns. This result was similar to the pattern observed for aphasic patients with temporal lobe lesions. Early Alzheimer's disease is centered on the temporal lobe, and is assumed to have a greater impact on noun processing than verb processing. Using the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 1983) and the Action Naming Test (Obler, Albert, & Lozowick, 1986), Bowles et al. (1987) and Williamson et al. (1998) revealed that AD participants performed less well overall than the controls. These AD participants were disproportionably more impaired at naming the object pictures. This claim was also supported in Robinson et al. (1999), in which a probable AD participant with very severe dementia (MMSE score of 7/30) was examined. Results were indicative of a large verb advantage (object naming, 7/40 correct and action naming, 32/40 correct). Recently, Yang et al. (2006) examined semantic memory in low-educated Chinese-speaking AD persons with different degrees of dementia severity. AD persons and normal controls completed semantic memory tasks including the Object Naming Test, the Remote Memory Test and the Semantic Association of Verbal Fluency Test. Results indicated that AD participants with moderate to severe dementia showed significantly impaired performances on the Object Naming Test. Pignatti, Ceriani, Bertella, Mori, and Semenza (2006), examining the naming abilities in spontaneous speech in Parkinson and Alzheimer's disease, also revealed that verb vocabulary for AD persons in early stages was less rich. Other studies (Cappa et al., 1998; Druks et al., 2006; Kim & Thompson, 2004; Lee et al., 1998; Masterson et al., 2007; Robinson et al., 1996; Wang, 2010), however, have challenged the selective preservation of verbs in AD individuals. Cappa et al. (1998), for example, compared the object and action naming performance of nineteen Italian AD participants. These patients were somewhat better at naming objects. The distinction, nonetheless, was not significant. Lee et al. (1998) conducted a similar study in Chinese. Again, no difference was found between object naming and action naming. Graceful degradation for both categories (i.e. nouns and verbs) and no category advantage were also identified in Wang's (2010) study on Chinese-speaking AD persons. Kim and Thompson (2004) examined fourteen persons with the probable AD by using thirty-six object and thirty-six action pictures with the matched verbal labels for frequency and length. Results indicated that the AD group performed significantly better in object pictures (96.6%) than in action ones (87.1%). Recently, Druks et al. (2006) compared the object and action performance of nineteen participants with mild to moderate Alzheimer's disease with that of nineteen healthy age matched participants. Carefully-matched object and action items were adopted as stimuli. Both groups made faster responses and fewer errors on the object pictures than on the action ones. It was concluded that object and action naming posed different demands for the language system. Greater processing demands posed by verbs may play an essential role in AD individuals who showed poorer performance in verb than in noun naming. Similar conclusion was reached in Mätzig et al. (2009), who investigated noun and verb differences in picture naming for nine aphasic patients and nine age-matched unimpaired participants. These patients performed faster and more accurately in naming the object pictures than the action ones, implying that action naming placed more and different demands on their cognitive processes than object naming. Concerning the AD semantic-lexical impairment, the third issue in debate is about pictorial attribute effects. During the past decades, a picture-naming task has been frequently adopted to investigate how AD persons process conceptual-semantic and lexical information for a certain item (Barry, Morrison, & Ellis, 1997; Gainotti et al., 1996; Hodges, Patterson, Graham, & Dawson, 1996; Kohnert, Hernandez, & Bates, 1998; Nicholas, Obler, Au, & Albert, 1996; Whatmough et al., 2003). Picture naming helps examine whether the content of information remains intact and unveil the process of which word the subject is endeavoring to produce. In picture naming, three serially organized stages are often involved: object recognition, object comprehension, and lexicalization (Warren & Morton, 1982). Also, different attributes of pictures, for instance, frequency, familiarity, age of acquisition, imageability, and visual complexity (Druks & Masterson, 2000; Lachman & Lachman, 1980; Mätzig et al., 2009), may help account for naming latency and retrieval processes. Earlier reports of picture naming concerned the development of cognitive functions for adults and for children. Cycowicz, Friedman, Rothstein, and Snodgrass (1997), for instance, collected normative data of line drawings viewed by five- and six-year-old children. Three pictorial variables, including name agreement, familiarity and visual complexity, were investigated. Results showed that children differed from adults in both the name most frequently assigned and the number of alternative names. Distinction between these two groups was also observed in their rating of picture familiarity and, to a lesser extent, in visual complexity. Exploring stimulus familiarity effects on AD participants, Gainotti et al. (1996) implemented a picture naming task, contrasting living with non-living categories. The rating of familiarity value followed Funnel and Sheridan (1992). It was found that performance in AD participants and controls was significantly influenced by the familiarity value. The higher the familiarity, the more accurate responses the participants gave. Compared with the healthy controls, AD participants seemed more sensitive to the familiarity effect. It was finally concluded that defective performance in AD participants could be mainly due to a low familiarity factor. A recent study regarding pictorial attribute effects on aphasic patients' naming abilities was done in Mätzig et al. (2009), which compared nine aphasic patients' and nine age-matched unimpaired individuals' performance in naming noun and verb pictures. Pictorial attributes such as frequency, age of acquisition, imageability, visual complexity and familiarity were discussed in a regression analysis for the control group. Significant predictors for the control participants were age of acquisition, imageability, and visual complexity. Frequency and familiarity, however, were not significant predictors. For the regression analysis on the accuracy data for each aphasic patient, only four pictorial attributes (i.e. phoneme length, age of acquisition, imageability and grammatical class) were entered as predictor variables. Results indicated that imageability was significantly associated with accuracy for two patients, and age of acquisition for one patient. None of the other predictors that entered the analysis turned out to be significant determiners. Despite the growing research interest in nominal and verbal patterns of AD, there are still open questions. So far, few studies have addressed Chinese-speaking AD individuals. Little attention has been drawn to their mechanism for verb fluency and noun fluency so as to examine the content of information stored in their brains. Whether they have the selective preservation of nouns or verbs in semantic fluency tasks remains unknown. Also, whether factors in semantic disorders influence them in the same way as speakers in the western context is worthy of investigation. Hence, the present study aims: (a) to elucidate the nature of semantic impairment, as revealed in a category fluency task completed by AD persons and by healthy controls in a Chinese-speaking society; (b) to examine possible semantic category effects, as revealed in a picture naming task completed by AD persons and by healthy controls in a Chinese-speaking society; (c) to address the factors in semantic disorders and to elucidate the predictive power of semantic variables on the severity of dementia.

The current investigation revealed factors in action-object picture naming produced by Chinese-speaking persons with or without Alzheimer's disease. First of all, the present analysis addressed the nature of semantic impairment in a category fluency task. A significant group effect was identified and AD participants offered significantly less amount of information, thus supporting the hypothesis of information loss. Secondly, the category effect (action vs. object) was examined in a category fluency task and a picture-naming task. Despite no selective impairment in the AD group, a significant object advantage was revealed for the normal controls in the fluency task. Thirdly, different semantic processes for AD participants and normal controls were identified in the error types and error frequency in naming pictures. Errors made by healthy controls were semantic-related, shape-related responses, or responses of picture parts. AD participants, however, tended to offer no responses or circumlocutionary comments. The current study was additionally contributive in the effect of age and in the significant predictors of semantic and pictorial variables. The effect of age clearly demonstrated that healthy elders encountered greater difficulty than the healthy adults in doing semantic tasks in a quantitative way, not in a qualitative way. As for the significant predictors, “action naming” and “action fluency” were two statistically significant semantic variables, which helped reliably predict the severity of dementia for Chinese-speaking AD participants in the present investigation. “Frequency” was a statistically significant pictorial variable. These findings contributed to the pictorial effects on AD individuals, which have been neglected in the preceding literature. Issues addressing how these significant predictors helped in the clinical treatment for AD persons, as suggested in Ally, Gold, and Budson (2009), and why the less frequent concepts were especially prone to semantic impairments for AD persons were worthy of future investigation.