Impact of Mid‐ and Late‐Life Diabetes in the Trajectories of Cognitive Aging in a Bi‐ethnic Cohort of Mexican Americans and Non‐Hispanic White Older Adults

Abstract

AbstractBackgroundMidlife vascular risk factors, including diabetes, have been associated with an increased risk of Alzheimer’s disease (AD) and related dementias (ADRD). Hispanics experience both a higher burden of cardiometabolic diseases and AD/ADRD compared with non‐Hispanic White older adults. However, there is a lack of diversity in AD/ADRD research, and additional investigations are needed to understand the impact of cardiometabolic diseases in AD/ADRD disparities. We aimed to investigate the association of mid‐ and late‐life diabetes with cognitive decline in Mexican American (MA) and Non‐Hispanic White (NHW) adults.MethodWe included participants from the San Antonio Heart Study (SAHS), initiated in 1979 to examine ethnic differences in diabetes and cardiovascular disease between MA and NHW. In 1992, SAHS participants were invited to the San Antonio Longitudinal Study of Aging (SALSA), in which cognitive function was assessed using the MMSE at four time points. The latent variable mixture model implemented in Mplus was used to identify statistically distinct classes in cognitive trajectories based on repeated MMSE scores. The Bolck‐Croon‐Hagenaars method was used to test covariate differences across classes. Participants with missing data in 2 or more MMSE scores or covariates were excluded from the analysis.ResultOur analytical sample included 337 participants (52.2% MA, 60.8% women, Table 1). Two distinct classes were identified (Shannon’s entropy = 0.813). Participants within class 1 had a steeper cognitive decline (0.53‐point drop per follow‐up), whereas those in class 2 appeared more stable (0.17‐point drop per follow‐up). Compared to class 2, participants in class 1were 2.1 and 2.5 times more likely to have diabetes during midlife (p = 0.019) and late‐life (p = 0.002) respectively. In contrast, ethnic‐stratified analysis revealed significant findings only among MA (Table 2), where participants in class 1 were 2.4 times more likely to have diabetes in midlife (p = 0.023) and 1.8 times in late life (p = 0.005) compared to class 2.ConclusionDiabetes was related to a steeper cognitive decline, particularly among MAs with prevalent diabetes in midlife. These findings provide new insights into the role of vascular burden in cognitive decline disparities among MAs. Further analysis exploring additional vascular factors are underway in this unique population.