REM Sleep and Emotional Regulation: What Sleep Disruption Does to Anxiety and Fear

What is REM sleep and why does it matter for the brain?

Sleep consists of several phases that alternate in cycles of 90–110 minutes. The rapid eye movement phase — REM — accounts for an average of 20–25% of nightly sleep in a healthy adult. It is during this phase that the brain is most active: electroencephalographically, REM resembles wakefulness, but skeletal muscles are atonic.

REM sleep is characterized by high acetylcholine and low serotonin and noradrenaline levels. This combination creates a neurochemical environment in which emotionally significant memories can be replayed and reinterpreted without a hormonal stress backdrop. It is precisely this property that underlies the "sleep to forget" hypothesis — or more precisely, "to preserve the content of a memory while reducing its affective load."

What happens to emotions when REM sleep is disrupted?

Viselli and colleagues (Sleep, 2026) conducted a randomized crossover experiment in 17 healthy young adults. On one night subjects slept without intervention; on another they received vibrotactile stimulation synchronized with the REM phase: the fragmentation index rose from 23.12 to 75.52 episodes (p<0.001), while REM duration was reduced by approximately 23 minutes.

The next morning and 48 hours later, participants were shown a series of emotional images while cardiac deceleration (CD) — a physiological marker of the orienting response to a stimulus — was recorded. Normally, CD diminishes with repeated presentations of the same stimulus: the brain "habituates" and reactivity fades. After the fragmented night this process was blocked: habituation was not observed either immediately after sleep or two days later (all p>0.529 in the fragmentation group; t=3.36, p=0.007 in the control group).

Parieto-occipital alpha power after fragmentation positively correlated with impaired habituation: r=0.710 (p=0.003) at 24 hours and r=0.618 (p=0.014) at 48 hours. An important caveat: the sample is small — 17 people; replications are needed.

Why might anxious people have more REM sleep?

Mediano, Fernández-Abascal, and Fondevila Estévez (Journal of Sleep Research, 2026) observed 128 non-clinical adults (75 women, 53 men; age 18–82, mean 35.5 years). Sleep was measured with a Fitbit Charge 5 tracker over several nights; anxiety was assessed using the STAI-T scale.

Result: trait anxiety positively correlated with REM sleep proportion (r=0.34, p<0.005). Similar correlations were found for depressive symptoms (r=0.31) and negative affect (r=0.32). Slow-wave sleep (SWS) was not associated with anxiety (p=0.49). Sex differences were also identified: women showed higher average anxiety and REM sleep proportion than men.

The authors interpret this association as an adaptive response: the brains of people with chronically high anxiety increase REM activity in an attempt to process the emotional load. This is consistent with neural models in which REM sleep functions as a "therapeutic window" for processing aversive experience — but the cross-sectional study design does not permit causal conclusions.

How does REM sleep participate in fear extinction?

Alongside clinical observations, neuroscientists are deciphering the cellular mechanism. Hong and colleagues (Current Biology, 2024) worked with a conditioned fear model in mice. Optogenetic suppression of pyramidal neurons of the infralimbic cortex specifically during REM sleep that followed fear conditioning disrupted the consolidation of fear extinction memory. The same intervention after an extinction session had no effect — meaning the window is critical: immediately after fear formation, not after its extinction.

The proposed neural substrate is theta oscillations at 4–7 Hz in prefrontal-limbic circuits during the REM phase. These appear to underpin the reactivation and reorganization of emotional memory traces without anxious affect, owing to reduced aminergic tone.

What does this mean for people with sleep disorders?

Sleep disturbances in post-traumatic stress disorder (PTSD) are characterized precisely by REM phase fragmentation and nightmares. If REM sleep is a window for fear extinction, its chronic disruption may create a vicious cycle: fear is not extinguished, anxiety does not diminish, and the next night is disturbed again.

Furthermore, certain sleep medications (benzodiazepines, some Z-hypnotics) suppress REM sleep — this is potentially significant when evaluating their long-term effects on emotional health. This area requires further randomized research.