Cortical thickness is influenced by bilingual language dominance in typical and pathological cognitive aging

Abstract

AbstractBackgroundPrior research in clinical and non‐clinical populations demonstrates that speaking more than one language can modify gray matter structure (e.g., DeLuca et al., 2020; Torres et al., 2022). However, existing work has largely classified bilingualism dichotomously, diminishing the ability to establish associations between bilingual factors (e.g., language dominance) and regional cortical thickness (CT).MethodWe examined cross‐sectional differences in CT across a large, ethnically diverse sample of bilingual and monolingual older adults from the Health and Aging Brain Study: Health Disparities cohort (n = 1,605; O’Bryant et al., 2021). Differences in CT between language dominance subgroups across three levels of diagnostic phase (cognitively unimpaired, mild cognitive impairment (MCI), and Alzheimer’s dementia (AD)) were evaluated using linear‐mixed effects models. Participants were characterized as English‐dominant or Spanish‐dominant speakers according to responses on language interview questions and the Short Acculturation Scale for Hispanics (Marin et al., 1987).ResultAssociations between CT and language dominance differed across diagnostic phase (Tukey‐adjusted ps <0.05). In the cognitively unimpaired group, English‐dominant speakers exhibited greater CT than Spanish‐dominant speakers in in regions‐of‐interest (ROIs) related to auditory processing (Heschl’s gyrus), semantic processing (temporal pole), and executive functioning (frontal pole). Spanish‐dominant speakers exhibited greater CT than English‐dominant speakers in two ROIs supporting motor (precentral) and spatial processing (entorhinal). These CT differences were attenuated in the MCI phase. Here, English‐dominant speakers exhibited greater CT than Spanish‐dominant patients in the frontal pole (executive processing). Spanish‐dominant patients did not exhibit greater CT than English‐dominant patients. Notably, this pattern of group differences was reversed in AD patients. Here, Spanish‐dominant patients exhibited greater CT than English‐dominant patients in ROIs in Alzheimer’s disease‐related regions (parahippocampal and entorhinal).ConclusionOur results indicate associations between language dominance and regional CT, with distinct patterns by diagnostic phase. Whereas English‐dominant bilinguals demonstrated greater regional CT relative to Spanish‐dominant bilinguals in the cognitively unimpaired and MCI phase, Spanish‐dominant bilinguals had greater CT in disease‐related regions in the AD phase. The presence of greater CT in Spanish‐dominant bilingual AD patients is suggestive of resistance to pathological progression. Future research examining longitudinal patterns of neural decline in larger samples is needed.