EP 83. Neural correlates of age-related changes in cognitive action control

Abstract

Introduction Healthy aging has been associated with a decrease in cognitive action control, but the underlying neural mechanisms are still elusive. The spatial stimulus-response compatibility (SRC) task captures this aspect of cognitive control, as it requires overcoming predominant but incorrect action tendencies to enable correct responding. Aim To identify brain areas that change their action-control-related activity with age. Methods Using fMRI we measured brain activity in a large sample of adults recruited from the population-based 1000BRAINS study (n = 234; mean age 52.5 years; 130 males) while performing an SRC task. Participants were instructed to give a speeded response to lateralized visual stimuli with the ipsilateral (spatially compatible) or contralateral (spatially incompatible) hand. Effects of incompatibility and age on performance [reaction time (RT) and error rate (ER)] were tested via an analysis of covariance (ANCOVA) with compatibility as within-subject factor and age as covariate. Effects on brain activity were also tested via an ANCOVA, additionally including as covariate the behavioral incompatibility effect (ICE; i.e., the difference between mean RT in incompatible and compatible trials). Results RT and ER globally increased under incompatibility, and the increase in RT became larger with advancing age as indicated by a significant interaction effect. Neurally, we observed incompatibility-related activity bilaterally in dorsolateral prefrontal cortex (DLPFC), premotor cortex (PMC), intraparietal sulcus (IPS) and adjacent inferior and superior parietal lobule (IPL/SPL), anterior insula, (pre-)supplementary motor area [(pre)SMA] extending into midcingulate cortex (MCC), and cerebellum. Age-related hyper-activity in this network was restricted to bilateral DLPFC and cerebellum. In turn, a significant association with the behavioral ICE was found in the left IPS and SPL. Thus, while older adults showed a stronger recruitment of the cerebellum and DLPFC, high-ICE performers showed a stronger recruitment of the IPS/SPL. Conclusion We here replicated the significant increase of incompatibility costs with age ( Langner et al., 2014 ) in a substantially larger sample. Moreover, our results showing an age-related increase of task-dependent activation within the DLPFC provide an important complement to its previously demonstrated reduced resting-state connectivity. We thus argue that reduced connectivity during rest and increased recruitment during active tasks may represent complementary mechanisms of age-related network changes. In contrast, activity in the left IPS correlates with behavioral incompatibility costs, rather than age, highlighting a critical distinction between these. Our behavioral and neural data jointly corroborate a significant influence of age on cognitive action control, which is neurobiologically differentiable from behavioral performance.