Structural language change in Alzheimer's disease: a study of spontaneous speech tasks across different stages of cognitive impairment
Year: 2023
Type: dissertation
Study design: Observational Study
Abstract: Linguistic characteristics of spontaneous connected speech are showing increasing potential for detecting and tracking cognitive decline in Alzheimer’s disease, and could be particularly suitable to study at-risk populations in early phases of the disease’s pathology. This thesis focuses specifically on linguistic aspects of connected speech reflecting shared features of language and episodic memory. All variables were at the grammatical (rather than lexical) level of linguistic organization, and centered on definiteness/specificity at the level of nominal reference (e.g. the blue curtain vs. something blue), embedded clauses connecting events to each other (e.g. he is falling off the stool while stealing the cookie), spatio-temporal marking in verb phrases (verbal Aspect, non-present tenses, modality) and genericity of clauses (e.g. You can swim on a beach vs. I swam on that beach). Chapter 1 motivates these variables in the context of a general introduction to AD and the study of language therein. Chapter 2 first tests the measures in a pilot study with an English language dataset. We found that nominal definiteness, verbal Aspect, as well as event specificity (measured through adjunct clauses), were sensitive to differences between healthy adults and those with both MCI and AD. Following this, Chapter 3 details the design and implementation of a new spontaneous speech protocol with tasks that, unlike traditional picture descriptions, involve mental time travel. Using this protocol, in-person interviews were conducted at ACE Alzheimer Center (Barcelona) with 119 participants from five diagnostic groups: healthy controls (n = 18), subjective cognitive decline (n = 31), nonamnestic MCI (n = 16), amnestic MCI (n = 23), and Alzheimer’s disease (n = 31). Using this dataset, Chapters 4-6 examine between-group linguistic differences in a picture description task, a past-directed autobiographical narrative, and a scene construction, respectively. Finally, Chapter 7 examines the relation of linguistic variables to viii neuropsychological scores, and explores which task provides language samples which best predict group membership and neuropsychological test performance. Several variables showed consistent between-group effects across tasks. In particular, definite reference decreased in groups with objective cognitive impairment relative to controls, while generic clauses increased towards the AD end of the spectrum, representing a loss of specificity as measured at the clausal level. Genericity correlated strongly with qualitative assessments of episodicity, which confirms that past qualitative measures of episodicity correspond to changes in linguistic variables in the context of an episodic memory impairment. In turn, modality, past tense marking, and prepositional and adverbial adjunct marking, and embedded clauses showed early group differences in specific tasks, potentially reflecting differences between each tasks’ specific cognitive demands. The scene construction task (the most episodic of the three) provided language samples which predicted the most variance in neuropsychological scores and diagnosis group membership (up to 54.4%). While no linguistic variables successfully differenciated healthy adults from those with subjective cognitive decline, they did differentiate controls from participants at the MCI stage. Overall, this thesis demonstrates that linguistic markers adapted to the neuropsychological context of AD distinguish cognitively-impaired groups on the AD spectrum from healthy aging adults, and they offer further evidence for a connection between language and episodic memory. These findings invite further neurocognitive studies targeting the relation between language- and episodic memory-related networks in the brain, and they have methodological implications for clinical settings: tasks which involve mental time travel can offer language samples which most effectively depict group differences.
Author: Chapin, Kayla
Language: English
Subfield: Psychiatry and Mental health
Field: Medicine
Domain: Health Sciences
Keywords: Alzheimer's disease, language change, spontaneous speech, connected speech, episodic memory, genericity, verbal aspect, mental time travel, autobiographical narrative, subjective cognitive decline, mild cognitive impairment, picture description, scene construction, grammatical complexity, neuropsychological performance, cognitive decline detection, English speakers, Spanish speakers, bilingualism
Open Access status: green
OpenAlex ID: W7034172824
