Tired and wired: sleep deprivation prevents emotional adaptation to prolonged and ambiguous threat
A new study combining virtual reality, heart rate monitoring, and multidimensional experience sampling reveals that sleep deprivation blocks the brain’s natural ability to downregulate fear and anxiety in the face of sustained, unclear threats. The findings offer a mechanistic explanation for why poor sleep is a well-established risk factor for anxiety disorders.
Anyone who has lain awake at 3 a.m. with a vague sense of dread about tomorrow knows that a tired brain is an anxious brain. Now, researchers at the University of York have put that intuition to a rigorous experimental test, and the results are striking: after a single night of total sleep deprivation, healthy young adults lost the ability to emotionally adapt to an environment saturated with ambiguous threat.
The study, posted June 25 as a preprint on bioRxiv, is the first to use immersive virtual reality to probe how sleep loss affects emotional responses to prolonged and uncertain danger. The findings suggest that the same mechanism that normally lets the brain recalibrate its threat response over time collapses under sleep deprivation, leaving individuals stuck in a maladaptive state of heightened arousal.
What they found
The research team, led by Sullivan E.C., McCall C., and Cairney S., recruited healthy young adults and randomly assigned them to either a full night of restful sleep or total sleep deprivation. The next morning, participants donned virtual reality headsets and navigated an immersive virtual world that alternated between ambiguously threatening and non-threatening contexts.
The experiment combined three measurement layers: physiological arousal via heart rate monitoring, subjective experience via repeated in-task sampling, and a behavioral measure of goal-oriented emotional control. This multidimensional approach allowed the researchers to track how threat responses evolved over time, not just at a single snapshot.
Sleep-restored participants showed a clear pattern of emotional adaptation. As they spent more time in the threatening virtual environment, their physiological arousal and subjective fear ratings gradually declined. Their brains were doing exactly what healthy brains are supposed to do in the face of sustained, uncertain threat: renormalize.
Sleep-deprived participants showed the opposite pattern. Despite initial increases in emotional arousal comparable to the rested group, they could not overturn those elevated responses. Arousal stayed high and in some cases escalated further. Worse, sleep-deprived individuals exhibited a breakdown of goal-oriented emotional control, the capacity to intentionally steer attention away from threat and toward task-relevant goals.
Intriguingly, the study also identified a potential protective factor. Resting heart rate variability, a physiological marker of the brain’s inhibitory control capacity, predicted better regulation of arousal responses to ambiguous threat, even among sleep-deprived participants. Individuals with higher resting heart rate variability were partially buffered against the emotional dysregulation caused by sleep loss, suggesting that baseline autonomic flexibility acts as a resilience factor.
Why it matters
Anxiety disorders affect an estimated 300 million people worldwide, and poor sleep is one of the most consistent epidemiological risk factors. But the mechanistic link has remained elusive. This study provides a clear candidate mechanism: sleep deprivation prevents the brain from performing adaptive renormalization of emotional arousal in contexts where threat is prolonged and uncertain, the very kind of context that defines real-world anxiety.
Everyday life is not a series of discrete, unambiguous threats like a predator appearing in your path. It is a fog of ambiguous uncertainties: a delayed text from a partner, a concerning medical test result, a tense work meeting that keeps getting rescheduled. The brain must somehow downregulate its alarm system in the face of this lingering ambiguity, or anxiety becomes chronic. This study shows that sleep is essential for that downregulation process to work.
The finding also has implications for clinical populations and shift workers. If one night of total sleep deprivation can produce this effect in healthy young adults, the cumulative impact of chronic partial sleep restriction, the kind experienced by millions of people every night, may be even more profound.
Limitations
As a preprint, this study has not yet undergone peer review, and the results should be interpreted with appropriate caution.
The sample consisted of healthy young adults, and it is unclear how the findings generalize to older populations, clinical groups, or individuals with pre-existing sleep or anxiety disorders. The study also used total sleep deprivation, an extreme manipulation that does not fully mirror the chronic partial sleep restriction that is more common in the general population. The virtual reality paradigm, while more ecologically valid than traditional lab-based threat paradigms, is still an artificial environment, and it remains to be seen whether the same patterns hold in real-world contexts.
Additionally, the protective effect of heart rate variability was observed at rest, not during the task itself, leaving open questions about the dynamic interplay between autonomic regulation and emotional control in real time.
Bottom line
Sleep deprivation blocks the brain’s ability to emotionally adapt to prolonged and ambiguous threat, trapping individuals in a state of elevated arousal and impairing goal-oriented emotional control. The findings point to a specific mechanistic pathway linking poor sleep to anxiety. Higher resting heart rate variability may offer some protection, but the clearest prescription remains the simplest one: get enough sleep.
Source
Sullivan EC, McCall C, Croissant M, Henderson L, Schofield G, Cairney S. “Sleep deprivation prevents emotional adaptation to prolonged and ambiguous threat.” bioRxiv, posted June 25, 2026. DOI: 10.64898/2026.06.21.733648.
*This article reports on a preprint that has not yet been peer-reviewed. The findings should be considered preliminary until publication in a peer-reviewed journal.*
*This content is reproduced from the preprint under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC-BY-NC-ND 4.0).*