Challenges to attention: a continuous arterial spin labeling (ASL) study of the effects of distraction on sustained attention

Abstract

Maintaining attention and performance over time is an essential part of many activities, and effortful cognitive control is required to avoid vigilance decrements and interference from distraction. Regions at or near right middle frontal gyrus (Brodmann's area (BA) 9), as well as in other prefrontal and parietal areas, are often activated in studies of sustained attention (e.g., Cabeza and Nyberg, 2000; Kim et al., 2006; Lim et al., 2010). This activation has often been interpreted as representing the engagement of cognitive control processes. However, such studies are typically implemented at one level of task difficulty, without an experimental manipulation of control demands. The present study used the distractor condition sustained attention task (dSAT), which has been used extensively in animals to determine the role of neuromodulator systems in attentional performance, to test the hypotheses that BA 9 is sensitive to changes in the demand for cognitive control and that this sensitivity reflects an increased engagement of attentional effort. Continuous arterial spin labeling (ASL) was used to measure neural activity in sixteen healthy, young adults performing a sustained attention task under standard conditions and under a distraction condition that provided an experimental manipulation of demands on cognitive control. The distractor impaired behavioral performance and increased activation in right middle frontal gyrus. Larger increases in right middle frontal gyrus activity were associated with greater behavioral vulnerability to the distractor. These findings indicate that while right middle frontal gyrus regions are sensitive to demands for attentional effort and control, they may not be sufficient to maintain performance under challenge. In addition, they demonstrate the sensitivity of ASL methods to variations in task demands, and suggest that the dSAT may be a useful tool for translational cross-species and clinical research.

Figure 1. Sustained Attention Task (SAT)

Participants completed 140 and 160 s blocks of the SAT and distractor condition SAT (dSAT). After 1, 2 or 3 s a short signal appeared (signal trials) or did not appear (nonsignal trials). Signal and nonsignal trials were pseudo-randomized and equally presented. After a short, constant delay, participants heard a low frequency buzzer (response cue). Participants then made a button-press response to indicate whether a signal had or had not occurred on that trial. Correct responses (both hits and correct rejections) generated a high frequency feedback tone signaling a monetary reward; incorrect responses and omissions did not receive any feedback. During dSAT blocks, participants performed the SAT in the presence of a visual distractor, the screen flashing silver to black at 10 Hz.

Figure 2. Distraction impairs task performance

The bars show the mean SAT score (see text for calculation) collapsed across SAT (black bars) and dSAT (white bars) task blocks. Error bars represent between-subjects standard error around the mean. Chance performance is a SAT score of zero. While duration did not lead to strong effects within the SAT, the presence of distraction decreased performance in a duration-dependent manner, with the biggest deficits evident on the shortest signal duration condition.

Figure 3. Region of interest analyses in frontal and occipital cortex

Bars depict mean and between-subjects standard error of difference scores for the contrast values on the parameter estimates for dSAT – dFIX and dSAT – SAT. In right MFG and right medial frontal gyrus, dSAT activity is greater than dFIX activity and than SAT activity. However, in right cuneus, dSAT activity is greater than SAT, but not greater than dFIX activity. This suggests that while activation in visual regions during dSAT blocks is largely driven by the flashing screen, the visual stimulation does not fully account for the activity during dSAT blocks in frontal regions.

Figure 4. Activation in right frontal regions during SAT performance increases in the presence of distraction

SAT performance (A) elicited activation in right dorsolateral prefrontal cortex as well as bilateral motor, cingulate and insular cortex regions. The presence of distraction (dSAT blocks, B) activated regions in frontal and parietal cortex. These regions were strongly right lateralized after controlling for the visual distractor stimulus (C). Compared to the SAT blocks, dSAT performance resulted in increased activation in parts of right dorsolateral prefrontal cortex (BA 9, D). Color bar indicates Z scores ranging from 3 to 5. Anatomical image represents the average of each subject’s normalized structural scan. Axial slices shown at z = 36, saggittal slices at x = 44, MNI coordinates.

Figure 5. Right frontal activation during distraction correlates with behavioral performance decrements

(A) Right MFG (BA 9, ROI centered on MNI coordinates (36, 10, 34)) activity was negatively correlated with SAT scores during distraction. Scatterplots depict participants contrast on the parameter estimate values for the dSAT – SAT contrast versus their mean SAT score for each signal duration during the task blocks with distraction. The contrast values were negatively correlated with the 50 and 29 ms SAT scores, with a trend for a negative correlation seen on the 17 ms SAT scores. (B) The contrast values for right MFG were also positively correlated with the distractor effect, or the difference for each participant between their mean SAT scores on blocks without distraction and their mean scores on blocks with distraction. (C) This pattern was not seen in visual regions, as no correlations were observed between contrast values in right cuneus (BA 7, ROI centered on MNI coordinates (10, −68, 32)). These data support the idea that the increased activity in right MFG during distraction is related to the increased attentional control demands of the dSAT condition.