Predicted Brain Age Difference Scores at Age 56 Are Associated with Executive Function Changes Across 12 Years.

Background: Predicted brain age difference (PBAD) scores are novel metrics which compare chronological age to age predicted from neuroimaging data. PBADs are highly relevant to Alzheimer's disease (AD) and cognitive aging. Individuals with AD have predicted brain ages about 10 years older than their chronological age (Franke et al. 2010). Less is known about PBADs in midlife, but recent work on the sample studied here has demonstrated that PBADs are highly heritable (59‐75%), highly stable between middle‐ and early‐old age, and moderated by modifiable lifestyle behaviors earlier in midlife (age 40). The current study examined how midlife PBAD (mean age 56) predicted executive functions concurrently and longitudinally into early old age (mean age 68). Method: We examined 779 non‐demented men in the Vietnam Era Twin Study of Aging (VETSA) who completed cognitive and neuroimaging assessments at up to three assessments (M = 56, 62, and 68 years). PBADs were based on two methods (Liem et al., 2017, Cole et al. 2019) which include information about grey matter or both gray and white matter. Executive function was based on six tasks spanning inhibition, shifting and working memory. Analyses controlled for age, diabetes, hypertension, and APOE‐ε4 dosage. Result: Latent growth models revealed that greater PBAD at age 56 (i.e., greater brain age than chronological age) was associated with worse executive functioning at baseline and greater decline in executive functioning over the following 12 years. The latter association was observed only for PBADs that included white matter (Cole et al., 2019) and persisted even when excluding N = 61 individuals with amnestic mild cognitive impairment (MCI). Subsequent analyses indicated that associations were not moderated by APOE allele status or cognitive reserve (general cognitive ability at mean age 20). Conclusion: Our findings indicate that PBAD is associated with executive functioning in midlife and predicts executive function changes into early old age. They also suggest that PBADs capturing more information about brain health are better predictors of executive function decline across midlife into early‐old age. Finally, the persistence of these effects within cognitively normal subjects suggests PBADs are important predictors of cognitive decline before the onset of MCI/AD. [ABSTRACT FROM AUTHOR]