Tracking the content of spatial working memory during a bout of acute aerobic exercise.

Abstract

Exercise changes human performance in a range of cognitive domains and these changes often co-occur with fluctuations of neural activity. Electrophysiological studies have shown that brief bouts of low-intensity exercise can increase the visual evoked response, consistent with the notion that exercise induced arousal results in a gain in sensory processing. More recently, studies using encoding models have shown that feature-selective responses reconstructed from patterns of neural activity are also modulated during bouts of exercise. Here, we investigated whether the selectivity of spatial memories are modulated by brief bouts of exercise. Participants (n=4) performed a simple delayed estimation task involving the presentation of a circular stimulus (250ms) at one of eight equally spaced locations around fixation and the subsequent recall of the stimulus location following a brief retention period (1750ms). This task was performed while participants were seated on a stationary bike under two conditions: rest and while cycling at a low-intensity (50 W at 50 RPM). Sixty-four channel electroencephalography was recorded during the task (and during exercise) and gaze-contingent eye-tracking ensured fixation was maintained throughout each trial. We used an inverted encoding modeling technique to estimate location-selective tuning functions (TFs) from spatially distributed alpha activity measured across the scalp during the target and retention period. In both rest and low intensity exercise, the peak amplitude of the TFs was significantly different than a permuted control (p< .05). When the conditions were compared directly, the mean amplitude at 0 offset (i.e., the remembered location) during rest (M = .41 μV, SEM = 0.055 μV) was similar to the response during exercise (M = .38 μV, SEM = 0.050 μV). These results show that the spatial selectivity of visual working memory can be reconstructed during exercise and these reconstructions may have the same fidelity as when at rest.

Acknowledgement: This work was generously supported by US Army Research Office Grant W911NF-09-0001.