The sleep-deprived human brain

Abstract

How does a lack of sleep affect our brains? In contrast to the benefits of sleep, frameworks exploring the impact of sleep loss are relatively lacking. Importantly, the effects of sleep deprivation (SD) do not simply reflect the absence of sleep and the benefits attributed to it; rather, they reflect the consequences of several additional factors, including extended wakefulness. With a focus on neuroimaging studies, we review the consequences of SD on attention and working memory, positive and negative emotion, and hippocampal learning. We explore how this evidence informs our mechanistic understanding of the known changes in cognition and emotion associated with SD, and the insights it provides regarding clinical conditions associated with sleep disruption.

Figure 1. Sleep loss, attention and working memory

a | Brain regions and networks associated with attention and working memory (frontoparietal network (FPN); red), arousal (thalamus; green) and the default mode network (DMN; blue) are affected by sleep deprivation. The DMN is a collection of brain areas, including midline frontoparietal regions, that often disengage when an individual performs an externally driven, goal-directed task and then re-engage when an individual stops performing that task. Suppression of the DMN is necessary to mobilize appropriate on-task brain networks to achieve successful goal-directed behaviour, unless that task requires access to internally stored relevant information such as autobiographical memories or previously learned predictive cues^,,. Brain regions across multiple brain networks, including the dorsolateral prefrontal cortex (DLPFC), intraparietal sulcus (IPS), thalamus, medial prefrontal cortex (mPFC) and posterior cingulate cortex (PCC), are differentially altered by sleep loss. b | In the sleep-rested state, there is reciprocal inhibition between task-related FPN activity and DMN activity, supported by sustained ascending arousal input from the thalamus. This leads to consistent attentional and working-memory performance. In the sleep-deprived state, there is unstable reciprocal inhibition between task-related FPN activity and DMN activity, and erratic ascending arousal activity influencing thalamic activity. As a result, there is reduced task-related FPN activity and intermittent intrusions of DMN activity during task engagement. The behavioural consequence is variable and/or impaired attention and working-memory performance, worsening with lowered thalamic activity and improving with higher thalamic activity.

Figure 2. Sleep loss and incentive processing

a | Reward-relevant brain regions that are affected by sleep deprivation (SD) include cortical regions (blue) such as the medial prefrontal cortex (mPFC), insula and orbitofrontal cortex (OFC), and the subcortical region of the striatum (red). b | Increased adenosine load (blue circles) associated with SD triggers downregulation of dopamine (DA) D2 and D3 receptors (D2/3Rs), resulting in decreased receptor membrane expression within the striatum (internalized receptors; grey). Consequently, there is a greater ratio of D1R to D2/3R availability, and the relative increase in striatal D1R activation by DA (green circles) under SD conditions is proposed to increase risk-related and reward-related approach behaviours, shown in the see-saw imbalance. c | SD further disrupts incentive processing within PFC regions and thus is proposed to result in a lower signal-to-noise ratio (SNR) under SD conditions (represented by a narrow dynamic range of PFC sensitivity). This consequentially impairs the ability to accurately map and update changing value or reward probability (coloured dots) over time — such as that involved in reinforcement learning, ultimately contributing to non-optimal incentive-based decision making.

Figure 3. Sleep loss and aversive processing

a | Sleep deprivation (SD) amplifies amygdala reactivity (red) in response to negative emotional stimuli and decreases associated amygdala–medial prefrontal cortex (mPFC) connectivity (blue). b | SD alters sensitivity of the salience-detection network (the amygdala, anterior cingulate cortex (ACC) and anterior insula (AI)) to varying levels of emotional stimuli that range in valence strength (x axis in the two graphs) from negative (red) to neutral (blue) to positive (red). Under sleep-rested conditions (left graph), there is a wide and dynamic range of salience-detection sensitivity, resulting in the ability to accurately discriminate between degrees of emotional salience (tall vertical difference arrow; discerning emotional (red vertical line) from neutral stimuli (blue vertical line)). By contrast, SD (right graph) is proposed to trigger a narrowing and thus more nonspecific detection sensitivity range, impairing the ability to accurately discriminate between degrees of emotional salience (short vertical difference arrow). This loss of ‘net neutrality’ results in over-generalized emotional sensitivity, wherein an otherwise largely neutral stimulus (blue vertical line) is inappropriately registered as ‘emotional’ by the salience-detection network; further observed in biased emotional ratings of neutral stimuli. c | A downstream behavioural consequence of these central brain changes, in combination with disrupted peripheral autonomic nervous system feedback of visceral body information, could lead to inaccurate and even absent outward expression of emotions. This is supported by experimental evidence demonstrating that individuals deprived of sleep fail to register emotional faces shown to them on a screen and consequently are unable to accurately mimic the emotional face expressions of these target faces themselves. Part a is adapted with permission from REF. , Elsevier.

Figure 4. Sleep loss and hippocampal memory encoding

Sleep deprivation (SD) decreases encoding-related activity within the hippocampus (light blue), compared with normal sleep-rested conditions. Moreover, in sleep-deprived individuals, hippocampal connectivity with encoding-relevant cortical regions of the intraparietal sulcus (IPS), posterior cingulate cortex (PCC) and primary visual cortex (V1) (purple regions) during encoding is lower than in sleep-rested individuals. By contrast, hippocampal encoding-related connectivity with subcortical arousal regions of the thalamus and brainstem (red) after SD is increased.