Parametric modulation of practice-related changes over time in neural activity during a spatial working memory task
MetadataShow full item record
Working Memory (WM) has been studied extensively due to its predictive validity and impairment in mental disorders. Neuroimaging studies of the WM network generally have found involvement of the prefrontal cortex, parietal cortex, cerebellum, and occasionally the striatum. Neuroimaging studies have also demonstrated practice-related changes in neural activity during WM tasks. Practice effects have demonstrated between-group differences, as well as within-subject variability in neural activation between scans taken in different times (Sadek, 2001; Weissman, Woldorff, Hazlett, & Mangun, 2002.) These differences appear to correspond with both reductions and increases in neural activity, which are interpreted as a decrease in cognitive effort. Several studies have demonstrated effects of task domain on practice effects (Kelly & Garavan, 2005), with more basic sensorimotor tasks associated with increased activation over time, and higher-order tasks associated with decreased activation over time. Few experiments have specifically tested practice effects across runs of a single scanning session (Landau, 2005). This experiment aimed to demonstrate the effects of practice and task domain on neural activity over time. A series of experiments investigated intra-run changes in neural activity during a spatial 3-back, a working memory task, and 0-back, a visual attention task. We hypothesized increased activation in strategy-specific brain regions (i.e., posterior parietal cortex) and decreased activation in task non-specific regions (i.e., dorsolateral prefrontal cortex), with more pronounced decreases during the more demanding WM task, compared with the 0-back task. Results from the WM task revealed a general decrease in neural activity in both the dorsolateral prefrontal cortex and the posterior parietal cortex. Results from the visual attention task revealed an increase in dorsolateral prefrontal cortex. Behavioral results failed to demonstrate significant practice effects on performance measures. These results appear to reflect a practice-related decrease in cognitive effort during the higher-order task (i.e., WM), and an increased need for attentional control, possibly due to fatigue, during the less demanding task (i.e., visual attention.) Results of this study demonstrate the importance of understanding the temporal pattern of the WM network in describing WM abilities. More generally, they demonstrate the effects of uncontrolled variables on replicability in fMRI studies.