Understand and Use Dreams to Learn and Forget | Huberman Lab Essentials

Huberman describes this as his own protocol: bed at 10:30–11 PM, wake at 3–4 AM, NSDR, back asleep until 7 AM. He notes he has measured his sleep stages and confirmed that this pattern reliably delivers slow wave sleep early and REM late. The NSDR acts as a reset valve — it keeps the nervous system from entering full wakefulness, which would break the architecture needed for the return to sleep.

Huberman makes a precise claim: resistance exercise, unlike aerobic exercise, increases the percentage of slow wave sleep without disrupting the overall sleep pattern. Slow wave sleep is the stage that processes fine motor skill learning and specific factual details — so this protocol is especially useful when you are actively learning a physical skill. The mechanism runs through growth hormone: resistance training stimulates the endocrine cascade that produces the GH surge that typically accompanies early-night slow wave sleep.

The mechanism is specific: serotonin is the dominant neuromodulator of slow wave sleep, and it is largely absent during REM. Supplements that elevate serotonin or GABA push the brain toward a serotonin-heavy state that crowds out the low-serotonin, low-norepinephrine, zero-epinephrine window that defines true REM sleep. Even if sleep quantity appears normal, the REM stage is either shortened, suppressed, or architecturally distorted. The result is a morning with poorer emotional regulation, less motor skill consolidation, and less spatial-memory processing than an equivalent duration of drug-free sleep would have produced.

Huberman describes the convergence: REM sleep (no epinephrine = memories replayed without fear), EMDR (amygdala suppressed = trauma recounted without full fear activation), and ketamine (NMDA blocked = emotion-memory plasticity prevented) are all working on the same fundamental node. EMDR was approved by the American Psychological Association as a trauma treatment, representing a relatively rare instance of a behavioral intervention clearing the APA bar.

Huberman explains the neuroscience: there is a specific neural connection — literally a set of neurons and a nerve circuit — that goes from the bladder to the brainstem and activates waking mechanisms when bladder pressure reaches threshold. This is not just discomfort waking you; it is a dedicated neural alarm. Even if you return to sleep quickly, the continuity break disrupts the ultradian cycling, potentially resetting the REM-access clock.

Huberman is explicit: slow wave sleep is not just about rest. It is where motor programs are transferred from temporary working memory into durable motor cortex representation. Research using selective sleep deprivation — waking subjects only when electrode recordings showed slow wave stage — demonstrated dramatically impaired motor skill retention compared to those allowed to complete their slow wave cycles. The implication for athletes and performers: the night you train is the night that matters most for sleep protection.

Huberman explains the implication clearly: REM sleep creates a window in which the brain can process everything that happened during the day — including highly distressing events — in the absence of the molecule that would make you feel the fear. This is why people often wake from dreams about frightening events without feeling traumatized by the dream itself. The brain is essentially running a rehearsal that allows it to re-file the emotional weight attached to experiences. When REM is cut short — by waking early, by alcohol, or by stress-induced insomnia — this uncoupling cannot complete, and the person wakes with unprocessed emotional weight that manifests as irritability, catastrophizing, and an inability to regulate small stressors.

Huberman describes his own experience: when he misses his late-morning REM window by catching a flight or staying up, the world feels harder, daunting, and the little things feel like the big things. He connects this to a broader principle: REM sleep is where the brain trims away the emotional over-connectivity between experiences. A person lacking REM starts seeing problems everywhere — not because the problems are there but because the brain has lost its pruning function. He explicitly links this to the sleep disturbances seen in menopause: Dr. Sarah McKay's research suggests that menopausal temperature dysregulation disrupts sleep architecture, and it is the resulting REM reduction — not the hormones directly — that accounts for much of the emotional volatility.

Huberman explains that when you walk, run, or ride a bicycle — self-generating forward movement — your eyes automatically make reflexive side-to-side movements tied to the motor system. These are the exact eye movements used in EMDR. The amygdala suppression they induce creates a window in which someone can recount a traumatic or stressful experience with reduced fear response. Over repeated sessions, the uncoupling of emotion from the memory occurs through the same mechanism as REM sleep: the memory is replayed in the absence of the normal epinephrine-mediated fear signal. EMDR is approved by the American Psychological Association for trauma treatment and works best for single discrete events rather than complex prolonged trauma.

Huberman describes the clinical scenario: someone brought to an ER after witnessing a loved one killed in a car accident. Ketamine infusion can still allow the emotion to occur, but it prevents the plasticity — the actual rewiring of synaptic connectivity — that would lock that emotion permanently to the memory trace. He notes the ethical complexity: some emotional coupling to experiences is necessary and adaptive. The clinical judgment is about whether the degree of anticipated emotional encoding is likely to be disabling. The common thread with REM sleep and EMDR: all three mechanisms target the same node — the moment at which emotion and memory become neurologically fused — from different directions.

Huberman notes that for most people, striving for a fixed 8–9 hours is harder to achieve consistently than protecting a reliable 6–6.5. Since the brain builds the REM-heavy late-night cycles across sequential 90-minute ultradian blocks, disrupting the pattern — not just shortening it — is what most damages learning and emotional processing. This connects to his personal protocol: going to bed around 10:30–11 PM, waking around 3–4 AM, using NSDR to return to sleep, and waking around 7 AM — a pattern he has measured to reliably deliver both early-night slow wave sleep and late-morning REM.

Huberman explains the principle through everyday experience: you visit a new building, find the bathroom, interact with new people. If the experience was sufficiently important or engaging, it consolidates over the next few days. If not, it fades. What decides which way it goes is partly whether REM sleep replays the session. Wilson's rodent data shows the hippocampal place-cell sequences from daytime navigation replaying during REM at something close to real-time speed. This is distinct from the emotional uncoupling function — the spatial replay is about meaning-making and integrating new locations into a navigational map of the world.