Abstract
Introduction
Threat-related information is preferentially processed, facilitating quick and efficient responses. However, the impact of extended sleep deprivation on perception of and response to threatening information is not well known. Sleep loss may increase amygdalar activity and negative mood, potentially facilitating threat processing. However, it also reduces cognitive function, possibly impairing ability to respond. The present study assessed the extent to which extended sleep deprivation modulates threat processing using a threat expectation paradigm.
Methods
Twenty-one participants underwent one baseline night of sleep followed by 62hrs total sleep deprivation (TSD) and one recovery night of sleep (12hrs). Threat expectation task performance was assessed at baseline, at multiple time points during TSD, and following recovery sleep. To control for circadian influence, performance at three 1100 sessions (baseline, 52hrs into TSD, and recovery) were compared. The threat expectation task involved determining whether a presented face was fearful (i.e., signaled threat) or neutral. Faces were presented at three expectation levels: 80%, 50%, and 20% chance of viewing a fearful face.
Results
Overall, responses were faster (F=9.77, p=0.001) and more accurate (F=11.48, p=0.001) when the type of face (fearful or neutral) was expected. Accuracy significantly decreased over TSD (t=7.71, p<0.001) and recovered following subsequent sleep. Fear bias was calculated for accuracy (accuracy for fearful face minus neutral face). Under conditions of high expectation (80%) of viewing a fearful face, fear bias increased across TSD (t=-1.95, p=0.07). Although accuracy to both fearful and neutral faces significantly declined across TSD (both p<0.001), decline for neutral faces was greater, thus increasing fear bias. Importantly, the increased bias toward fear was still evident compared to baseline following a 12-hour recovery sleep opportunity, (t=-1.93, p=0.07).
Conclusion
Extended sleep deprivation, common in operational environments where there is also high expectation of encountering threat, impairs cognitive control and is thought to enhance amygdala activity. These data show that, consequently, cognitive resources become biased toward biologically adaptive behaviors (i.e., threat processing) at the expense of attending and responding more broadly to all stimuli. This behavior is not reversed with a single extended sleep opportunity.
Support (if any)
Department of Defense Military Operational Medicine Research Program (MOMRP)
Driver seat comfort for level 3-4 autonomous vehicles
