Project Summary/Abstract Spontaneous speech undergoes subtle but significant changes years, or even decades, before the onset of dementia. Signs of change in speech, or speech markers, include the use of syntactically simpler structures, reduced idea density, and more filler words, nonspecific nouns, and higher frequency words (Berisha et al., 2015; Snowdon, 1996). Such changes are potentially useful clinically for early diagnosis of dementia and are more broadly interesting for understanding normal and impaired language production. The proposed studies will investigate speech markers of Alzheimer’s Disease (AD) prior to diagnosis (i.e., in prodromal AD) in Spanish-English bilinguals. Speech markers have not been investigated in bilinguals even though use of two languages might magnify the cognitive mechanisms underlying speech production and sensitivity to prodromal AD. Bilinguals may exhibit unique speech markers not found in monolinguals, and some of the same markers as monolinguals but with differential sensitivity across their two languages. When bilinguals speak, they must choose which language to speak and must avoid interference from the language they are not using, tasks that may be especially difficult when speaking in the relatively less proficient, or non-dominant, language (L2). By contrast, speech in the dominant language, or L1, is relatively more automatic, at least in some aspects, and is also the language in which bilinguals may have their highest level of semantic knowledge and linguistic skill. These key differences between L1 and L2 lead to different expectations about which aspects of speech in each language should be most sensitive to cognitive decline (Aim 1) and most strongly correlated with other domains of cognitive functioning (Aim 2). To test these hypotheses, in Aim 1, we will examine archival data of spontaneous speech in bilinguals who were cognitively healthy at the time of testing but years later developed a diagnosis of Probable AD to identify linguistic markers of prodromal AD in L1 and L2. In Aim 2, we will prospectively examine relationships between L1 and L2 spontaneous speech markers and general cognitive functioning (measured by a battery of neuropsychological tests) and will test the hypothesis that time- pressured picture-naming, a recently identified speech marker of dementia risk (Stiver et al., 2021) may be a better speech marker of prodromal AD than naming with unlimited time because of greater demand it places on executive control. By examining the intersection of speech production and cognitive decline in bilinguals the proposed studies will inform models of bilingualism and will shed unique light on the role of executive control in supporting speech production more generally. The intersection between these fields of study may also lead to improved methods for early identification of AD in a diverse demographic. Lastly, this research project will enhance the applicant’s graduate school training and...