37 Bilingualism does not modify the association between stroke and cognitive performance in Mexican American older adults

Objective:The Latinx population is rapidly aging and growing in the US and is at increased risk for stroke and dementia. We examined whether bilingualism confers cognitive resilience following stroke in a community-based sample of Mexican American (MA) older adults.Participants and Methods:Participants included predominantly urban, non-immigrant MAs aged 65+ from the Brain Attack Surveillance in Corpus Christi- Cognitive study. Participants were recruited using a two-stage area probability sample with door-to-door recruitment until the onset of the COVID-19 pandemic; sampling and recruitment were then completed via telephone. Cognition was assessed with the Montreal Cognitive Assessment (MoCA; 30-item in-person, 22-item via telephone) in English or Spanish. Bilingualism was assessed via a questionnaire and degree of bilingualism was calculated (range 0%-100% bilingual). Stroke history was collected via self-report. We harmonized the 22-item to the 30-item MoCA using published equipercentile equating. We conducted a series of regressions with the harmonized MoCA score as the dependent variable, stroke history and degree of bilingualism as independent variables, and age, sex/gender, education, assessment language, assessment mode (in-person vs. phone), and self-reported vascular risk factors (hypertension, diabetes, heart disease) as covariates. We included a stroke history by bilingualism interaction to examine whether bilingualism modifies the association between stroke history and MoCA performance.Results:Participants included 841 MA older adults (59% women; age M(SE) = 73.5(0.2); 44% less than high school education). Most (77%) of the sample completed the MoCA in English. 93 of 841 participants reported a history of stroke. In an unadjusted model, degree of bilingualism (b = 3.41, p < .0001) and stroke history (b = -1.98, p = .003) were associated with MoCA performance. In a fully adjusted model, stroke history (b = -1.79, p = .0007) but not bilingualism (b = 0.78, p = .21) was associated with MoCA performance. When an interaction term was added to the fully adjusted model, the interaction between stroke history and bilingualism was not significant (b= -0.47, p = .78).Conclusions:Degree of bilingualism does not modify the association between stroke history and MoCA performance in Mexican American older adults. These results should be replicated in samples of validated strokes, more comprehensive bilingualism and cognitive assessments, and in other biling