Sleep deprivation alters functioning within the neural network underlying the covert orienting of attention

Abstract

One function of spatial attention is to enable goal-directed interactions with the environment through the allocation of neural resources to motivationally relevant parts of space. Studies have shown that responses are enhanced when spatial attention is predictively biased towards locations where significant events are expected to occur. Previous studies suggest that the ability to bias attention predictively is related to posterior cingulate cortex (PCC) activation [Small, D.M., et al., 2003. The posterior cingulate and medial prefrontal cortex mediate the anticipatory allocation of spatial attention. Neuroimage 18, 633-41]. Sleep deprivation (SD) impairs selective attention and reduces PCC activity [Thomas, M., et al., 2000. Neural basis of alertness and cognitive performance impairments during sleepiness. I. Effects of 24 h of sleep deprivation on waking human regional brain activity. J. Sleep Res. 9, 335-352]. Based on these findings, we hypothesized that SD would affect PCC function and alter the ability to predictively allocate spatial attention. Seven healthy, young adults underwent functional magnetic resonance imaging (fMRI) following normal rest and 34-36 h of SD while performing a task in which attention was shifted in response to peripheral targets preceded by spatially informative (valid), misleading (invalid), or uninformative (neutral) cues. When rested, but not when sleep-deprived, subjects responded more quickly to targets that followed valid cues than those after neutral or invalid cues. Brain activity during validly cued trials with a reaction time benefit was compared to activity in trials with no benefit. PCC activation was greater during trials with a reaction time benefit following normal rest. In contrast, following SD, reaction time benefits were associated with activation in the left intraparietal sulcus, a region associated with receptivity to stimuli at unexpected locations. These changes may render sleep-deprived individuals less able to anticipate the locations of upcoming events, and more susceptible to distraction by stimuli at irrelevant locations.

Figure 1

Schematic representation of the Posner task, and the timing parameters of a single trial. Subjects performed three runs in the scanner in each session. An event-related design was used. Each trial consisted of fixation followed by a cue presented for 200, 400, or 800 msecs. Following the cue, the target was presented for 100 msecs. The intertrial interval varied as a function of the SOA length to maintain an overall trial length of 2100 msecs, matching TR length.

Figure 2

Activation related to the presence of cue benefit (V⁺ − V⁻). A) PCC activation in the rested state greater than the sleep-deprived state. B) IPS activation in the sleep-deprived state greater than the rested state. All peaks are significant at p < 0.05 corrected at the cluster level. The color bar identifies the t-values.

Figure 3

Peri-stimulus time histograms for V⁺ (diamonds) and V⁻ (squares) trials in Rested and Sleep Deprived states for the maxima from the PCC (xyz: −6 −57 21) and IPS (xyz: −12 −69 51) clusters. (A) Mean group BOLD responses in the rested state within the PCC for V⁺ (solid line) and V⁻ (dashed line) trials. (B) Mean group BOLD response in the sleep-deprived state within the PCC for V⁺ and V⁻ trials. (C) Mean group BOLD response in the rested state within the IPS for V⁺ and V⁻ trials. (D) Mean group BOLD response in the sleep-deprived state within the IPS for V⁺ and V⁻ trials. Values are group means ± standard error of the mean. The graphs demonstrate the greatest cue benefits (difference between V⁺ and V⁻) in the PCC when subjects are rested (3A) and the IPS when they are sleep deprived (3D).