Electrocortical indices of cognitive control in working memory : exploring the effects of proactive interference, cognitive load, and aging

摘要

Cognitive control involves the ability to flexibly adjust cognitive processing inorder to resist interference and promote goal-directed behaviour. Although frontal cortexis considered to be broadly involved in cognitive control, the mechanisms by whichfrontal brain areas implement control functions are unclear. Furthermore, aging isassociated with reductions in the ability to implement control functions and questionsremain as to whether unique cortical responses serve a compensatory role in maintainingmaximal performance in later years. Described here are three studies in whichelectrophysiological data were recorded while participants performed modified versionsof the standard Sternberg task. The goal was to determine how top-down control isimplemented in younger adults and altered in aging. In study I, the effects of frequentstimulus repetition on the interference-related N450 were investigated in a Sternbergtask with a small stimulus set (requiring extensive stimulus resampling) and a task witha large stimulus set (requiring no stimulus resampling).The data indicated that constantstimulus res amp ling required by employing small stimulus sets can undercut the effectof proactive interference on the N450. In study 2, younger and older adults were testedin a standard version of the Sternberg task to determine whether the unique frontalpositivity, previously shown to predict memory impairment in older adults during aproactive interference task, would be associated with the improved performance whenmemory recognition could be aided by unambiguous stimulus familiarity. Here, resultsindicated that the frontal positivity was associated with poorer memory performance,replicating the effect observed in a more cognitively demanding task, and showing thatstimulus familiarity does not mediate compensatory cortical activations in older adults. Although the frontal positivity could be interpreted to reflect maladaptive corticalactivation, it may also reflect attempts at compensation that fail to fully ameliorate agerelateddecline. Furthermore, the frontal positivity may be the result of older adults'reliance on late occurring, controlled processing in contrast to younger adults' ability toidentify stimuli at very early stages of processing. In the final study, working memoryload was manipulated in the proactive interference Sternberg task in order to investigatewhether the N450 reflects simple interference detection, with little need for cognitiveresources, or an active conflict resolution mechanism that requires executive resourcesto implement. Independent component analysis was used to isolate the effect ofinterference revealing that the canonical N450 was based on two dissociable cognitivecontrol mechanisms: a left frontal negativity that reflects active interference resolution,,but requires executive resources to implement, and a right frontal negativity that reflectsglobal response inhibition that can be relied on when executive resources are minimalbut at the cost of a slowed response. Collectively, these studies advance understandingof the factors that influence younger and older adults' ability to satisfy goal-directedbehavioural requirements in the face of interference and the effects of age-relatedcognitive decline.