Submitted by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-05-23T17:06:36Z No. of bitstreams: 1 DIS_BRUNA_TESSARO_COMPLETO.pdf: 2324970 bytes, checksum: e436d9dc2e7d4836b9be4f4e228cb763 (MD5) / Made available in DSpace on 2017-05-23T17:06:36Z (GMT). No. of bitstreams: 1 DIS_BRUNA_TESSARO_COMPLETO.pdf: 2324970 bytes, checksum: e436d9dc2e7d4836b9be4f4e228cb763 (MD5) Previous issue date: 2017-01-17 / Conselho Nacional de Pesquisa e Desenvolvimento Cient?fico e Tecnol?gico - CNPq / Communicating is one of the first ability that suffer losses in neurodegenerative events associated to aging, such as Alzheimer?s Disease (AD). Since the number of elderly people has grown significantly, given the increase of life expectancy in the last decades, the study of aging, and the cognitive states that follow it, is extremely urgent and relevant. The concern with an accurate and early diagnosis is mounting, in order to differentiate an expected age-associated decline from deficits that point to Mild Cognitive Impairment (MCI), a clinical condition bewtween normal aging and dementia. Semantic memory deficits have increasingly been associated to AD, and, in some cases, to MCI. The discussion on which types of semantic categories are the most affected is quite productive in the literature, however, it is still inconclusive. Even more, low schooling appears to be related to cognitive aspects, including semantic deterioration. Thus, this work investigated semantic processing in MCI and AD as compared to age and socioeconomic paired controls, in a low schooling population. The thesis is organized in two studies, both include individuals aged between 60 and 80 years old, monolinguals and with 2 to 8 years of schooling. The first study included 36 individuals divided in an aging control group (ACG), a AD group (ADG), and a MCI group (MCIG), in which they were administered naming and semantic association assessments. Analyses considered total score, naming error types, and stimuli animacy in both tasks. Study two included a verbal fluency task (VFT) with semantic (animals and clothing), and phonologic (/p/) criteria, administered in 25 controls, 24 MCI and 18 AD participants. Analyses were conducted considering the formation of clusters, intersections and returns in verbal fluency productions. Findings of the first study confirm semantic memory deterioration in ADG, while MCIG did not show any differences from ACG and ADG, revealing inconclusive data. Living items seem to be the most deteriorated due to degeneration caused by AD and MCI. Most frequent error types in the naming task were coordinated, omission, and superordinated, and the ADG presented significantly more errors of these types as compared to ACG. Results from the second study show that the VFT was sensitive in differentiating the ACG from the clinical groups, and the semantic criteria differentiated MCIG from ADG, while the phonologic criteria differentiated ACG from ADG. Whilst the combination of cluster analysis and total score of both VFT showed efficacy in differentiating groups, the semantic category alone was very effective in differentiating the three groups. Reading habits have positively influenced the performance of participants in both studies, they were even more predictive of the performance of participants than scholling was. This study intends to broaden the understanding of the relationship between low schooling, cognitive decline, semantic memory, and animacy, as well as discuss models and theories based on our findings. / A habilidade de comunicar-se ? uma das primeiras habilidades que sofrem perdas em eventos neurodegenerativos associados ao envelhecimento, como a Doen?a de Alzheimer (DA). Uma vez que o n?mero de idosos tem crescido significativamente, dado que a expectativa de vida aumentou nas ?ltimas d?cadas, o estudo do envelhecimento e os estados cognitivos que o acompanham torna-se de extrema urg?ncia e relev?ncia. A preocupa??o com um diagn?stico acurado e precoce ? crescente, a fim de diferenciar um decl?nio esperado devido ? idade, de desempenhos indicativos de Comprometimento Cognitivo Leve (CCL), est?gio possivelmente anterior ao quadro demencial. D?ficits de mem?ria sem?ntica t?m sido cada vez mais associados ? DA e, em alguns casos, ao CCL. A discuss?o acerca de que tipos de categorias sem?nticas seriam mais afetadas ? bastante frut?fera na literatura, por?m, ainda inconclusiva em rela??o aos mecanismos subjacentes dessa dificuldade sem?ntica. Al?m disso, a baixa escolaridade demonstra relacionar-se com aspectos cognitivos, dentre eles a deteriora??o sem?ntica. Desta forma, o presente trabalho investigou o processamento sem?ntico no CCL e na DA, comparado ao processamento por controles emparelhados por idade e perfil socioecon?mico, em popula??es de baixa escolaridade. A disserta??o organiza-se em dois estudos, ambos incluindo indiv?duos com idade entre 60 e 80 anos, monol?ngues e com 2 a 8 anos de escolaridade. Participaram do primeiro estudo 36 indiv?duos divididos igualmente em grupo de idosos controle (GIC), grupo com DA (GDA) e grupo com CCL (GCCL), no qual desempenharam uma tarefa de nomea??o e uma de associa??o sem?ntica. As an?lises consideraram o escore total, os tipos de erros da nomea??o e a animacidade dos est?mulos das duas tarefas. J? o estudo dois contemplou uma tarefa de flu?ncia verbal (TFV) de crit?rio sem?ntico (animais e roupas) e fonol?gico (/p/), desempenhada por 25 participantes controles, 24 CCL e 18 DA. As an?lises reca?ram sobre a forma??o de clusters, intersections e returns na produ??o das flu?ncias. Os resultados do primeiro estudo confirmam uma deteriora??o da mem?ria sem?ntica no GDA enquanto que o GCCL n?o apresenta diferen?as em rela??o ao GIC e ao GDA, revelando dados inconclusivos. Os itens animados parecem ser os mais prejudicados pela degenera??o decorrente da DA e do CCL. Os tipos de erros mais frequentes de todos o grupos na tarefa de nomea??o foram dos tipos coordenado, omiss?o e superordenado, sendo que o GDA apresentou um n?mero significativamente maior de erros destes tipos em rela??o ao GIC. Os resultados do segundo estudo apontam que a TFV foi sens?vel na diferencia??o do GIC dos grupos cl?nicos, sendo que a flu?ncia verbal com crit?rio sem?ntico diferenciou o GCCL do GDA, enquanto que a flu?ncia verbal fonol?gica diferenciou o GIC do GCCL. Enquanto a combina??o da an?lise de clusters e do escore total das duas TFVs demonstrou efic?cia na diferencia??o dos grupos, a categoria sem?ntica animais, sozinha, demonstrou ser sens?vel na diferencia??o dos tr?s grupos. A frequ?ncia dos h?bitos de leitura influenciaram positivamente o desempenho dos participantes nos dois estudos, sendo at? mesmo mais preditivos que os anos de escolaridade para o desempenho dos participantes. O estudo pretende ampliar o entendimento da rela??o entre baixa escolaridade, decl?nio cognitivo, mem?ria sem?ntica e animacidade, bem como discutir modelos e teorias ? luz dos dados apresentados.

Submitted by PPG Letras (letraspg@pucrs.br) on 2018-12-13T10:48:35Z No. of bitstreams: 1 Fernanda_Schneider_VERS?O FINAL 12_12.pdf: 2996710 bytes, checksum: 10f51d3b651bfbe650679cd5e4cd2155 (MD5) / Approved for entry into archive by Sheila Dias (sheila.dias@pucrs.br) on 2018-12-17T11:56:24Z (GMT) No. of bitstreams: 1 Fernanda_Schneider_VERS?O FINAL 12_12.pdf: 2996710 bytes, checksum: 10f51d3b651bfbe650679cd5e4cd2155 (MD5) / Made available in DSpace on 2018-12-17T12:08:18Z (GMT). No. of bitstreams: 1 Fernanda_Schneider_VERS?O FINAL 12_12.pdf: 2996710 bytes, checksum: 10f51d3b651bfbe650679cd5e4cd2155 (MD5) Previous issue date: 2018-10-31 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior - CAPES / This research aims to analyze the micro and macrostructure in the oral production of three narratives, produced from a sequence of pictures, comparing the performance of participants with right hemisphere damage (RHD) and participants with left hemisphere damage (LHD) with the performance of control groups. Furthermore, the linguistic analysis was related to data of memory constructs (working and semantic), reading and writing habits and brain volumetry. The research is divided into Study 1 and Study 2. For Study 1, 32 adult participants who suffered ischemic stroke were selected (with a minimum of four months before collection) to compose the experimental group ? 16 participants in LHD and 16 participants in RHD aged between 50 to 79, with a minimum of two and maximum of 12 years of schooling; and 16 participants within the health control (CG), paired in age and schooling with the clinics. From these, 10 participants of each group ? capable to perform magnetic resonance imaging exams ? were selected to make part of Study 2, which relates data from neuropsychological and linguistic tasks to data from structural neuroimaging ? known as Voxel-based morphometry (VBM). Inferential analysis in Study 1 showed that post-stroke writing and naming variables influenced significantly the performance in the micro and macrostructure, with the clinical group obtaining a statistically inferior result in the microstructure in relation to the CG. In relation to the macrostructure, there was also a significant correlation in the Digit span (working memory) in the three groups, but without statistical difference between the groups. In Study 2, LHD performance was significantly lower in the microstructure compared to RHD and CG. In the microstructure, in Study 2, the brain regions were grouped into four major components (CP). Thus, in the inferential analysis, all the main components of the brain areas (CP1, CP2, CP3 and CP4) showed a correlation with the microstructure variable, as well as significantly correlated post-stroke writing and naming variables. In the macrostructure, in Study 2, the variables that presented a statistically significant effect were GLH, writing habits after stroke, working memory and interaction between GLH variables and one of the main components of brain area (CP3). The parahippocampal area of the RH showed a correlation with CP3, which implies that this region significantly differentiates LHD from RHD in macrostructure processing. Thus, based on the analyzes and discussions proposed in the study, it was intended to contribute to the understanding of micro and macrostructure characterization in the production of oral narratives in right hemisphere and left hemisphere lesions and their correlation with cognitive constructs, more specifically of working memory and semantics and volume of brain areas involved. The work sought to advance in the discussion surrounding hemispheric specialization for the processing of oral discourse, especially of narratives, for the benefit of future research and clinical intervention. / A presente pesquisa teve como objetivo analisar o processamento da micro e da macroestrutura na produ??o oral de tr?s narrativas, produzidas a partir de sequ?ncia de figuras, comparando o desempenho de participantes com les?o de hemisf?rio esquerdo (LHE) e participantes com les?o de hemisf?rio direito (LHD), com a de controles. Outrossim, a an?lise lingu?stica foi relacionada a dados de construtos de mem?ria (de trabalho e sem?ntica), h?bitos de leitura e escrita e de volumetria cerebral. A pesquisa foi dividida em Estudo 1 e Estudo 2. Para o Estudo 1, foram selecionados 32 participantes adultos que sofreram acidente vascular cerebral (AVC) isqu?mico - 16 participantes no GLHE e 16 participantes no GLHD - entre 50 e 79 anos, com m?nimo de 2 e m?ximo 12 anos de escolaridade; 16 participantes saud?veis controles (GC), pareados por idade e escolaridade com os cl?nicos. Desses, 10 participantes de cada grupo foram selecionados para fazerem parte do Estudo 2, em que se relacionaram os dados das tarefas neuropsicol?gicas e lingu?sticas a dados da neuroimagem estrutural de morfometria baseada em voxel (em ingl?s, Voxel-based morphometry - VBM). A an?lise inferencial mostrou, no Estudo 1, que as vari?veis escrita p?s-AVC e nomea??o influenciaram de modo significativo o desempenho na micro e na macroestrutura com o grupo cl?nico obtendo resultado estatisticamente inferior na microestrutura em rela??o ao GC. Em rela??o ? macroestrutura, houve uma correla??o significativa tamb?m no Digit span (mem?ria de trabalho) nos tr?s grupos, por?m sem diferen?a estat?stica entre os grupos. No Estudo 2, a performance do GLHE foi significativamente inferior na microestrutura em compara??o com GLHD e GC. Na microestrutura do Estudo 2, as regi?es cerebrais foram agrupadas em quatro componentes principais (CP). Assim, na an?lise inferencial, todos os componentes principais das ?reas cerebrais (CP1, CP2, CP3 e CP4) apresentaram correla??o com a vari?vel microestrutura, assim como correlacionaram significativamente as vari?veis escrita p?sAVC e nomea??o. Na macroestrutura, do Estudo 2, as vari?veis que apresentaram efeito estatisticamente significativo foram GLHE, h?bitos de escrita posteriores ao AVC, mem?ria de trabalho e a intera??o entre as vari?veis GLHE e um dos componentes principais de ?rea cerebral (CP3). A ?rea parahipocampal do HD demonstrou correla??o com o CP3, o que implica afirmar que essa regi?o diferencia significativamente o GLHE do GLHD quanto ao processamento da macroestrutura. Assim, a partir das an?lises e discuss?es propostas no estudo, pretendeu-se contribuir para a compreens?o da caracteriza??o da micro e da macroestrutura na produ??o de narrativas orais na les?o de HD e HE e sua correla??o com construtos cognitivos, mais especificamente da mem?ria de trabalho e sem?ntica, e volume de ?reas cerebrais implicadas. Procurou-se avan?ar na discuss?o sobre a especializa??o hemisf?rica para o processamento do discurso oral, em especial de narrativas, para benef?cio de pesquisas futuras e de interven??es cl?nicas.