Newborns spend roughly half their sleep time in REM, and that proportion drops steadily across childhood, settles around 20 to 25 percent in young adults, then drifts downward by about half a percentage point per decade through middle and later life. Those numbers come from decades of sleep-lab recordings, and they hold up remarkably well across studies. But the raw percentages only tell part of the story, because the amount of REM you actually get on any given night depends on how long you sleep, what you ate or drank, which medications you take, and how stressed you are. Understanding the normal trajectory helps you figure out when a change is just aging and when something else might be going on.
Newborns and Infants
The youngest humans are the biggest REM sleepers. A full-term newborn spends about 50 percent of total sleep time in what researchers call “active sleep,” the infant equivalent of REM. Premature infants can spend even more. This is not wasted time. The prevailing explanation, originally proposed in the 1960s and still widely accepted, is that the enormous volume of REM sleep in early life provides endogenous stimulation to a brain that is still wiring itself together.
More recent work has filled in the details of that idea. REM sleep in early life appears to support neuronal differentiation, migration, myelination, and the formation and pruning of synapses, all processes that peak in the first year or two of life.
1PubMed Central. Rapid Eye Movement Sleep during Early Life: A Comprehensive Narrative Review The brain essentially uses REM to generate its own neural activity patterns during periods when there is not yet much external sensory input to work with.2PubMed Central. The Ontogenetic Development of Rapid Eye Movement Sleep in Early Life and Its Regulatory Mechanisms As the brain matures and begins relying more on waking experience, the proportion of REM drops quickly. By about age two, REM has typically fallen from 50 percent to somewhere around 30 percent.
Children and Adolescents
Through childhood, REM continues to decrease as a share of total sleep, though total sleep time is decreasing too, so the absolute minutes of REM shrink from both directions. By the time a child enters school, REM usually accounts for roughly 20 to 25 percent of their night, not far from the adult range. The bigger sleep-architecture story during childhood involves slow-wave sleep, which is at its lifetime peak in the prepubertal years.
Adolescence introduces a wrinkle. A longitudinal study tracking teens on both school-night and extended-sleep schedules found that REM sleep durations increased slightly but significantly across the adolescent years, even as non-REM sleep followed a different trajectory.3PubMed Central. The maturational trajectories of NREM and REM sleep durations differ across adolescence on both school-night and extended sleep When teens were given the chance to extend their time in bed, both non-REM and REM durations exceeded what they got on school nights, but the REM increase was the one that grew with age. In other words, the adolescent brain seems to have a growing appetite for REM that school schedules and early alarms routinely frustrate.
A separate study confirmed the mechanism: when adolescents were given more time in bed, total sleep time increased, and that gain was composed of increases in both REM and non-REM stages.4Pediatrics. Earlier Bedtime and Its Effect on Adolescent Sleep Duration The practical takeaway for parents is that a teenager who sleeps only six or seven hours is likely shortchanging REM more than any other stage, because REM episodes grow longer in the final cycles of the night, and those are the cycles that get cut when the alarm goes off early.
Adults Through Midlife
Young adults in their twenties typically spend about 20 to 25 percent of the night in REM. Over the next few decades, that proportion edges downward, but slowly. A large analysis of polysomnography data found that REM as a percentage of total sleep decreased in a roughly linear fashion, losing about 0.6 percent per decade from early adulthood onward.5Oxford Academic (Sleep). Changes in REM-sleep percentage over the adult lifespan That is a gentle slide. A 25-year-old averaging 22 percent REM would be expected to average around 20 percent at 55. Most people would never notice.
Interestingly, a study of healthy men found that from early adulthood to midlife, there were no statistically significant changes in REM sleep at all. The dramatic aging-related shift during that window was in slow-wave (deep) sleep, which plummeted from about 19 percent in young men to just 3 percent by middle age. That lost deep sleep was replaced almost entirely by lighter non-REM stages, not by wakefulness.6JAMA. Age-Related Changes in Slow Wave Sleep and REM Sleep and Relationship With Growth Hormone and Cortisol Levels in Healthy Men REM stayed essentially stable until midlife, and the meaningful REM losses only kicked in after about age 50.
Older Adults
After midlife, REM sleep starts declining more noticeably. The same JAMA study found that increases in wake time and decreases in REM became significant starting in the fifties and continued into the seventies.6JAMA. Age-Related Changes in Slow Wave Sleep and REM Sleep and Relationship With Growth Hormone and Cortisol Levels in Healthy Men Older adults also tend to wake up more often during the night, which fragments sleep cycles and makes it harder to accumulate full REM episodes, since those episodes normally lengthen as the night progresses.
There is a curious twist at the very end of the lifespan. The analysis of REM percentage across adulthood found that the small linear decline in REM stopped around the mid-seventies, after which REM percentage actually ticked upward slightly. This was not because older adults were suddenly getting more REM in absolute terms. Instead, total sleep time was shrinking faster than REM minutes, so REM’s share of a shorter night inched up.5Oxford Academic (Sleep). Changes in REM-sleep percentage over the adult lifespan
A broader review of sleep in aging noted that most age-related sleep changes, including advanced sleep timing, more awakenings, and decreased slow-wave sleep, seem to occur between young and middle adulthood. Among healthy older adults, sleep parameters remain largely unchanged.7PubMed Central. Sleep in Normal Aging The key word there is “healthy.” Many of the sleep problems blamed on aging are actually driven by medications, chronic pain, mood disorders, and sleep-disordered breathing, all of which become more common with age and all of which can suppress REM independently.
Sex Differences in REM Sleep
Men and women follow broadly similar REM trajectories across the lifespan, but there are meaningful differences in the fine structure. A 2025 study examining the microstructure of REM sleep found that women showed a steeper decline over the decades in phasic REM percentage and REM density compared to men.8PubMed. Age and sex-related changes in the human REM sleep microstructure Phasic REM is the portion of REM sleep containing the bursts of actual rapid eye movements, which are associated with vivid dreaming and heightened neural activity. A steeper decline in these measures could mean that women’s REM sleep becomes qualitatively less intense over time, even if the total percentage of the night spent in REM looks similar on paper.
These findings are relatively new, and it is not yet clear what they mean for health outcomes. But they are a good reminder that “percent of the night spent in REM” is a coarse measure. Two people with the same REM percentage can have very different internal activity patterns during that REM, and those patterns may matter for memory, emotional regulation, and dreaming experience.
Why REM Matters Beyond Development
In infancy, REM’s role in brain maturation is its headline function. But REM stays important throughout life for different reasons. One of the best-studied adult functions is emotional memory processing. REM sleep appears to help the brain process fear memories and other emotionally charged experiences, with rhythmic interactions in specific brain regions playing a role that researchers are still mapping out in detail.9PubMed Central. Emotional Memory Processing during REM Sleep with Implications for Post-Traumatic Stress Disorder
One study found that REM sleep may initially increase emotional reactivity to charged stimuli in the short term, but that this heightened reactivity seems to serve a purpose: over subsequent nights, it facilitates deeper processing and leads to fewer intrusive memories in the long run.10PubMed. Differential Effects of REM Sleep on Emotional Processing: Initial Evidence for Increased Short-term Emotional Responses and Reduced Long-term Intrusive Memories The idea that a bad night of REM can make tomorrow’s emotions harder while good REM smooths out emotional memories over time has obvious relevance for conditions like PTSD, where disturbed sleep and intrusive memories feed each other.
What Commonly Disrupts REM Sleep
Several everyday substances and widely prescribed medications can substantially alter how much REM you get, often without your awareness.
Alcohol
Alcohol is one of the most reliable REM suppressors. A 2024 systematic review and meta-analysis confirmed a dose-response relationship: even a low dose of alcohol, roughly two standard drinks, delayed the onset of REM sleep and reduced REM duration. Higher doses made the disruption progressively worse.11PubMed. The effect of alcohol on subsequent sleep in healthy adults: A systematic review and meta-analysis An earlier review found that the delay in the first REM period was the single most consistent effect of alcohol on sleep architecture, appearing at all dose levels tested.12PubMed. Alcohol and sleep I: effects on normal sleep
Chronic alcohol use creates a different problem. People with alcohol dependence tend to have more fragmented sleep overall, and paradoxically may show more REM sleep than normal, a pattern that persists well into abstinence and may contribute to relapse risk.13PubMed Central. Alcohol and the sleeping brain So the acute effect of alcohol (less REM) and the chronic adaptation (disrupted, REM-heavy sleep) point in different directions, both bad.
Antidepressants
Most antidepressants powerfully suppress REM sleep. In animal studies, common SSRIs like paroxetine and citalopram reduced REM sleep by about 84 percent compared to controls, while the tricyclic antidepressant imipramine reduced it by about 69 percent.14PubMed. REM sleep homeostasis in the absence of REM sleep: Effects of antidepressants These are enormous effects, and the REM deficits from SSRIs were not fully recovered even after the drugs cleared, whereas imipramine-induced deficits were. This has been a puzzle for sleep researchers: if REM is so important, why do millions of people tolerate medications that nearly eliminate it? The answer is not fully settled, but it suggests either that the brain compensates in ways we do not fully understand or that the functions of REM are partially redundant with processes occurring during other sleep stages.
Exercise
Physical activity modestly reshapes sleep architecture. A meta-analysis of exercise and sleep studies found that exercise increased total sleep time and slow-wave sleep while delaying REM onset by about 10 minutes and reducing total REM sleep by 2 to 5 minutes.15Sleep Medicine Reviews. Exercise and sleep These are small effects compared to alcohol or medications, and the net trade is generally considered favorable, since the gains in deep sleep and sleep efficiency outweigh a few lost minutes of REM.
What Happens When REM Is Chronically Cut Short
Experimentally depriving animals of REM sleep produces a cascade of problems. A systematic review of both human and animal studies found that REM deprivation in animals causes aggressive behavior, increased pain sensitivity, reduced sexual behavior, and impaired consolidation of fear memories. When REM was disrupted during critical developmental windows, the consequences on emotional behavior were long-lasting.16PubMed. The effect of REM-sleep disruption on affective processing: A systematic review of human and animal experimental studies The few human studies in the same review pointed toward increased pain sensitivity and stronger consolidation of emotional memories following REM deprivation, suggesting that some of the animal findings translate.
Cognitive effects are also documented. REM-deprived mice showed impaired fear memory on both contextual and cue-based tasks, with measurable changes in how synapses functioned in the amygdala.17PubMed Central. The beneficial effects of leptin on REM sleep deprivation-induced cognitive deficits in mice At the cellular level, REM deprivation in rats increased markers of oxidative stress in the brain, though these changes reversed after a period of recovery sleep.18PubMed Central. Effect of REM sleep deprivation on the antioxidant status in the brain of Wistar rats That reversibility is somewhat reassuring: a few bad nights of REM are not permanently damaging, at least in animal models. Chronic, sustained REM loss is another matter, and much harder to study in humans for obvious ethical reasons.
How Reliable Are Consumer Sleep Trackers for Measuring REM
If you wear a smartwatch or fitness tracker that reports your REM sleep, take the numbers with a generous grain of salt. A 2023 validation study compared 11 consumer sleep trackers, including popular wrist-worn and bedside devices, against polysomnography, the clinical gold standard. The performance varied enormously. The best device achieved a macro F1 score (a measure of classification accuracy) of 0.69, while the worst scored just 0.26. Different devices had different strengths: some were better at detecting wake periods and REM, while others were better at identifying deep sleep.19PubMed Central. Accuracy of 11 Wearable, Nearable, and Airable Consumer Sleep Trackers: Prospective Multicenter Validation Study
The practical implication is that your tracker’s nightly REM number is best used as a rough trend indicator, not a precise measurement. If your watch says you got 18 percent REM last night, that might actually be 14 percent or 23 percent. But if you notice that your REM percentage has dropped consistently over several weeks, that trend is more likely to be real than any single night’s readout. And if you are anxious about hitting a specific REM target each night, the tracker is probably creating more stress than useful information, which is ironic, since stress itself can alter sleep architecture.
REM Sleep Behavior Disorder
Normally, the body is effectively paralyzed during REM sleep, a feature called atonia that prevents you from physically acting out your dreams. When that paralysis fails, the result is REM sleep behavior disorder (RBD), a condition in which people kick, punch, shout, or thrash during REM episodes. This can be dangerous for both the person and any bed partner.
RBD is more than a nuisance. People with isolated RBD, meaning they have the disorder without an existing neurological diagnosis, carry a high lifetime risk of eventually developing a neurodegenerative disease, particularly synucleinopathies like Parkinson’s disease and Lewy body dementia.20PubMed Central. REM Sleep Behavior Disorder as a Pathway to Dementia: If, When, How, What, and Why Should Physicians Disclose the Diagnosis and Risk for Dementia A meta-analysis of longitudinal studies found that the onset of RBD frequently precedes the neurodegenerative disease by several years, sometimes more than a decade.21PubMed. The risk of neurodegeneration in REM sleep behavior disorder: A systematic review and meta-analysis of longitudinal studies This makes RBD one of the strongest early markers for these conditions, and it is being studied as a potential window for early intervention. If you or a partner notice repeated episodes of physically acting out dreams, especially in middle age or later, it is worth bringing up with a doctor, not just for safety but for what it could signal.
REM Sleep and Breathing
REM poses a particular challenge for anyone prone to obstructive sleep apnea. The same muscle paralysis that keeps you from acting out dreams also affects the muscles that hold your upper airway open. During REM, the genioglossus muscle, the main tongue muscle that helps keep the airway dilated, loses both its tonic (baseline) and phasic (burst) activity. For someone whose airway is already narrow or floppy, this tips the balance toward collapse.22European Respiratory Review. REM sleep obstructive sleep apnoea
Some people have sleep apnea that occurs primarily or exclusively during REM, a pattern sometimes called REM-predominant OSA. Because REM concentrates in the second half of the night, these individuals may have a completely normal first few hours of sleep and then experience repeated apneas and oxygen drops from the early morning hours onward. Standard overnight sleep studies capture this, but a shorter or split-night study might miss it. REM-predominant OSA is particularly common in women and in people who are only mildly overweight, populations where sleep apnea is often underdiagnosed in the first place. The muscle-tone suppression specific to REM makes this a mechanistically distinct problem from the garden-variety apnea that occurs across all sleep stages, and some clinicians argue it deserves different treatment thresholds.