The nature of brain-behavior covariations with increasing age is poorly understood. In the current study, we used a multivariate approach to investigate the covariation between behavioral-health variables and brain features across adulthood. We recruited healthy adults aged 20–73 years-old (29 younger, mean age = 25.6 years; 30 older, mean age = 62.5 years), and collected structural and functional MRI (s/fMRI) during a resting-state and three tasks. From the sMRI, we extracted cortical thickness and subcortical volumes; from the fMRI, we extracted activation peaks and functional network connectivity (FNC) for each task. We conducted canonical correlation analyses between behavioral-health variables and the sMRI, or the fMRI variables, across all participants. We found significant covariations for both types of neuroimaging phenotypes ( ps = 0.0004) across all individuals, with cognitive capacity and age being the largest opposite contributors. We further identified different variables contributing to the models across phenotypes and age groups. Particularly, we found behavior was associated with different neuroimaging patterns between the younger and older groups. Higher cognitive capacity was supported by activation and FNC within the executive networks in the younger adults, while it was supported by the visual networks’ FNC in the older adults. This study highlights how the brain-behavior covariations vary across adulthood and provides further support that cognitive performance relies on regional recruitment that differs between older and younger individuals.

IntroductionHealthy aging is typically associated with cognitive decline and lower negative affect. Previous studies have reported a significant and opposite role of the amygdala in relation to cognitive and affective processing in early adulthood. However, it remains unclear how aging impacts such relationships.MethodsSeventy-seven healthy participants including 40 young (mean age = 26.1 years) and 37 older (mean age = 61.8 years) adults completed a functional MRI Affective Stroop (AS) paradigm, a cognitive battery, and the state-trait anxiety inventory. The AS fMRI paradigm included “task trials,” where participants saw a positively, negatively or neutrally valenced distractor image, followed by a numerical display, followed by another distractor image. We extracted signal in both amygdalas during the AS Task and compared it across all conditions and age group. We further conducted moderation analyses to investigate the impact of aging on the relationship between amygdala activation and anxiety or cognitive variables, respectively.ResultsAt the behavioral level, older participants showed lower trait anxiety than the younger adults (p = 0.002). While overall slower during the AS task, older adults achieved comparable accuracy during the AS task, relative to the younger adults. At the brain level, we revealed a significant interaction between age group and trial types in amygdala activation (F = 4.9, p = 0.03), with the older group showing stronger activation during the most complex trials compared to the passive view trials. We further found that age significantly modulated the relationship between anxiety and the left amygdala activation during negative stimuli, where the younger adults showed a positive association while the older adults showed a negative association. Age also significantly modulated the relationship between verbal fluency and left amygdala activation during incongruent versus view trials, with the younger adults showing a negative association and the older adults showing a positive association.DiscussionThe current study suggests that the role of the amygdala on both emotional processing and cognitive traits changes between early and late adulthood.

No abstract

No abstract