UC Santa Barbara

When accomplishing goal-directed behavior in naturalistic settings our globalphysiological state can vary dramatically, oscillating between periods of wakefulness, emotional stress, and physical activity. Fluctuations in global state, in turn, induce a cascade of neuromodulatory changes that affect how the brain processes sensory information from the external environment. Despite the inextricable link between global state and brain function, goal-directed behavior has predominantly been studied when the body is stationary and at rest. Thus, it is unclear as to whether perturbations in global state modulate cognitive processes dependent on this sensory information, such as working memory (WM). The current body of work aims to determine how changes in global state induced by an acute bout of aerobic exercise modulate WM and its underlying neural correlates. Study 1 investigated the relationship between acute exercise and cognition, which revealed that aerobic exercise induces a small enhancement in general task-performance. Moderator analyses indicated that time-dependent measures of cognition were especially improved by exercise-induced perturbations in global state. Importantly, executive functions, such as inhibitory control and WM, were influenced by engaging in physical activity. Building on these meta-analytic results, Study 2 investigated whether the fidelity of spatial WM representations is impacted during an instance of aerobic exercise. Participants completed a delayed change detection task both at rest and during a bout of low-intensity cycling while neural activity was concurrently recorded using electroencephalography (EEG). An inverted encoding modeling technique was employed to estimate location-selective channel response functions from topographical patterns of alpha-band (8-12 Hz) activity. Importantly, robust spatially selective responses were reconstructed both at rest and during exercise throughout the stimulus encoding and retention period, demonstrating for the first time that the fidelity of spatial WM representations could be tracked in a physiologically active state. The selectivity of these responses was degraded during exercise relative to rest, suggesting that the fidelity of location representations may be diminished. Study 3 further investigated the impact of exercise on WM encoding and maintenance abilities. Participants completed a delayed change detection task that consisted of varying set sizes. Importantly, on some of the trials participants were required to encode target stimuli while simultaneously ignoring distractors, thus enabling the evaluation of WM filtering efficiency. Analyses of an event-related potential known as contralateral delay activity (CDA), which tracks the number of items stored in WM, indicated that there was no difference in WM load between rest and exercise conditions. Decoding analyses revealed that patterns of voltage potentials across the scalp tracked WM load both at rest and during exercise. These results suggest that WM filtering efficiency and the number of items that can be actively stored are robust to perturbations in global state caused by light intensity exercise. Together, this collection of studies illuminates the selective impact of exercise on WM processes, and highlights the importance of considering global state when developing theoretical frameworks of cognition.