Association between lifestyle at different life periods and brain integrity in older adults

Background Increasing evidence suggests that lifestyle factors are related to cerebral markers of aging and Alzheimer’s disease (AD) in older individuals. However, most studies addressed the association between current lifestyle and neuroimaging, providing little information on the relative effect of lifestyle at different life periods on older adults’ brain integrity. We proposed to determine the life period(s) at which lifestyle factors have the greatest influence on older adults’ brain health. Method We included baseline data of 135 cognitively unimpaired older adults (aged>65) from the Age‐Well study (NCT02977819). They underwent multimodal neuroimaging providing measures of Gray Matter volume (GMv), glucose metabolism, perfusion (early‐Florbetapir‐PET) and beta‐amyloid burden. The Lifestyle Experiences Questionnaires (LEQ) was used to assess complex mental activities during early‐ (13‐30y), mid‐ (30‐65y) and late‐life (>65y). To evidence the effect of activities at each life period controlling for the others, multiple regressions were conducted including the three LEQ sub‐scores to predict each neuroimaging modality, both 1) focusing on AD‐signature regions and 2) voxel‐wise. In the presence of significant associations, complementary analyses were realized to directly compare the effect of life periods between them. All analyses were controlled for age and sex. Results LEQ scores were not associated with AD‐signature regions. Voxel‐wise analyses revealed that higher LEQ score at midlife were positively associated with GMv, including the anterior cingulate cortex (Fig.1a), and negatively associated with amyloid burden, notably in the precuneus (Fig.1b). When compared with the other periods, these associations were statistically stronger for midlife. On the other hand, late‐life activities were positively associated with perfusion (Fig.2a) and glucose metabolism (Fig.2b), mainly in medial frontal regions, and these associations were stronger for late‐life than for any other life period. Conclusion Our study suggests that complex mental activities at midlife are more strongly associated with structural and molecular markers of brain integrity in older adults, while late‐life (i.e. current) activities are more strongly correlated with brain function. Interestingly, these associations were not found on AD‐signature regions but in brain regions relevant to reserve and aging. Lifestyle at different life periods might have an influence on distinct markers of