Why sleep is important for memory, mood and health
Sleep is important because it supports memory, emotional functioning and the physiological systems involved in cardiovascular and metabolic health (Ramar et al., 2021). Experiments also show that restricting sleep can impair new memory formation (Crowley et al., 2024). Neither finding means that every proposed biological benefit has been demonstrated in humans.
Sleep involves changing brain states, accumulated pressure to rest and an internal daily timing signal. Controlled studies measure what happens when sleep is shortened; reviews compare results across laboratories, while public health guidance addresses sleep duration and quality. Evidence about brain waste removal also comes from mice and requires separate interpretation.
Why is sleep important for the brain and body
The brain remains biologically active during sleep: Rasch and Born's 2013 review of human and animal research describes sleep-dependent consolidation, in which previously encoded memory representations are reactivated and reorganised for longer-term storage. Memory processing continues during sleep. Sleep therefore has a measurable relationship with learning even when a person is not consciously rehearsing information.
Physical health is part of the same account: the American Academy of Sleep Medicine position statement, written by Ramar and colleagues in 2021, identifies cognitive functioning, mood, mental health, cardiovascular health and metabolic health as functions relevant to healthy sleep. The statement also connects adequate sleep with fewer accidents linked to sleepiness and fatigue. An association cannot establish causation.
The role of sleep in brain clearance is less settled: in experiments involving live mice, Xie and colleagues (2013) used imaging and tracer measurements to compare brain fluid movement across sleep and wake states. They measured a 60% increase in interstitial space during natural sleep or anaesthesia and an accompanying increase in cerebrospinal fluid exchange. The mouse experiment did not measure human dementia incidence.
McKinley and colleagues (2013) discussed those mouse experiments in a commentary on neurodegeneration, and they noted that the invasive methods could not be reproduced in humans in their existing form. An improved movement of experimental tracers in mouse brain tissue is therefore separate from a claim that ordinary sleep prevents human Alzheimer's disease. Whether ordinary sleep reduces Alzheimer's disease risk in humans cannot be determined from these mouse experiments.
Why humans need sleep and what controls sleep timing
Sleep regulation involves the homeostatic process S and circadian process C interacting across the day in Borbély and colleagues' model (2016). These signals account for changes in sleep tendency across wakefulness and the daily cycle. The model predicts sleep timing.
Rasch and Born's 2013 review of human and animal research describes slow-wave sleep and rapid eye movement sleep as recurring stages in mammalian sleep, and their relative prominence changes through the night, with slow-wave sleep occurring predominantly earlier and REM sleep extending later. Consequently, the timing and continuity of a sleep period can alter which stages are represented, even before any cognitive result is measured.
Homeostatic sleep pressure
Homeostatic pressure accumulates during wakefulness and interacts with circadian timing in Borbély and colleagues' model (2016). The model successfully simulates sleep timing and intensity across different experimental conditions. A person can therefore be sleepy because sleep pressure has accumulated without the clock necessarily being at the same phase as on a different day.
Circadian timing
The other component, Process C, is controlled by a circadian pacemaker (Borbély et al., 2016), and it helps account for the timing of sleep episodes alongside the pressure accumulated by remaining awake. The model also describes continuing interactions between these regulatory processes rather than assigning sleep to a fixed clock hour alone.
Sleep quality introduces another dimension. Ramar and colleagues (2021) identify appropriate timing, regularity, good quality and the absence of sleep disorders alongside sufficient duration. Two people with similar reported sleep hours may therefore have different sleep continuity or timing. Duration misses continuity and regularity.
How sleep stages support memory and learning
Memory formation involves encoding and consolidation, with replay of recently encoded representations during slow-wave sleep described in Rasch and Born's 2013 review of human and animal physiology. Those representations may then be integrated into longer-term memory, with subsequent REM sleep potentially stabilising them. The proposed sequence describes an active mechanism, although different kinds of learning may depend on different conditions.
A systematic review by Crowley and colleagues (2024) examined restricted sleep and later memory formation in human experiments, and it analysed 125 effect sizes from 39 reports involving 1,234 participants. Across the comparison of restricted sleep, 3–6.5 hours, with normal sleep, 7–11 hours, the estimated effect on memory formation was Hedges' g = 0.29 (95% confidence interval 0.13–0.44). The estimate applies to the experiments in the review, not to an exact number of memories lost by an individual.
Crowley's team found evidence of impaired memory encoding and consolidation following restriction (2024): encoding concerns information as it is first learned; consolidation concerns the subsequent maintenance of that information. The distinction matters when a person sleeps poorly before studying compared with sleeping poorly after studying. Both conditions can influence performance, so one low test score cannot locate the failure in a single stage of memory processing.
Non-REM sleep and memory consolidation
Non-REM sleep includes the slow-wave stage discussed in Rasch and Born's 2013 review of human and animal studies, and their account links slow-wave activity with reactivation of recently encoded memories. In that account, reactivation is one route by which representations acquired while awake can be reorganised during sleep. The mechanism remains different from a claim that each moment of slow-wave sleep produces a fixed improvement in memory.
Rasch and colleagues (2007) tested a specific memory cue using an odour presented during learning and then presented again during sleep; re-exposure during slow-wave sleep improved retention of hippocampus-dependent declarative memories, but not the procedural memories tested. The cue did not have the same effect during REM sleep or wakefulness, or when it had been absent during earlier learning. Brain imaging also showed significant hippocampal activation when the odour was presented during slow-wave sleep.
REM sleep and emotional memory
REM sleep also recurs in the mammalian sleep cycle, as described in Rasch and Born's 2013 review of human and animal studies, and in the memory-consolidation account, REM sleep may help stabilise changes formed during slow-wave sleep. REM episodes also become more prominent towards the end of a typical human night. None of those observations assigns all emotionally significant learning to REM alone.
Schäfer and colleagues (2020) examined whether sleep preferentially preserves emotional material relative to neutral material in a meta-analysis; in the subset of 22 comparisons containing both sleep and wake groups, the emotional-over-neutral retention advantage was not statistically significant between those groups. The comparison assessed selective retention.
Schäfer and colleagues (2020) concluded that their aggregated evidence did not support a dependable special advantage of sleep for emotional material. This sits alongside physiological accounts describing complementary sleep stages. A proposed mechanism for memory processing and a meta-analysed comparison of emotional selectivity answer different questions, so one cannot substitute for the other.
How sleep affects mood and emotional regulation
Sleep loss changes measurable emotional responses: Palmer and colleagues (2024) combined 1,338 effect sizes from 154 experimental studies involving 5,717 participants aged 7–79 years. This synthesis examined changes following extended wakefulness, shortened sleep or night-time awakenings. Because the studies manipulated sleep, their findings give a more direct test of short-term emotional consequences than a survey asking how many hours someone usually sleeps.
Across these experiments, Palmer et al. (2024) found reduced positive affect, with standardised mean differences ranging from −0.27 to −1.14, and increased anxiety symptoms, with standardised mean differences ranging from 0.57 to 0.63. Positive affect refers to pleasant emotional experience rather than an absence of psychiatric illness. The pooled estimates describe measured reactions to experimental sleep loss; they do not count new diagnoses of anxiety disorders.
Emotional reactivity after sleep loss
Emotional effects are not confined to how people rate their mood: Palmer and colleagues (2024) also found blunted arousal in response to emotional stimuli after sleep loss, with standardised mean differences ranging from −0.20 to −0.53. Mood and arousal can diverge. Self-reported emotional arousal reflects participants' subjective experiences rather than physiological responses to emotional stimuli.
The experiments included several forms of disturbed sleep: Palmer et al. (2024) found that extended wakefulness, shorter nights and night-time awakenings adversely affected emotional functioning in the combined literature. Sleep was manipulated experimentally. Whether a given laboratory effect continues unchanged during long-term sleep problems requires a different type of study.
Memory research provides an instructive boundary. The selective advantage of sleep over wakefulness for emotionally valenced memories was not statistically significant in 22 comparisons (Schäfer et al., 2020). Emotional reactivity on waking and the preferential retention of emotional information are different outcomes. These comparisons do not exclude stage-specific effects on emotional memory.
Depression and mental health associations
Findings for depressive symptoms were mixed in Palmer and colleagues' 2024 meta-analysis and varied according to the type of sleep loss, and experimental declines in positive mood are evidence about short-term emotional function. They cannot independently establish the development of clinical depression, which involves a different outcome and observation period. The degree of certainty changes with the question being asked.
Ramar and colleagues (2021) identify mental health among the domains linked to healthy sleep, although that broad position statement does not assign the same strength of causal evidence to every diagnosis. Research on physical exercise, self-concept and well-being addresses different influences on how people feel, and its outcomes should not be attributed to sleep without a direct comparison. Mood is measured in overlapping ways across research fields, but the interventions differ.
Why sleep matters for the heart, metabolism and immunity
Inadequate sleep is associated with several long-term health outcomes: Ramar and colleagues (2021) link chronic insufficient sleep with increased risk of mortality and with cardiovascular disease, diabetes and obesity in the wider population. Observational associations can involve other characteristics of the people being studied. They establish a public health concern without proving that each condition is produced by sleep loss alone.
There is also controlled metabolic evidence. In a randomised crossover study reported by Zuraikat and colleagues (2024), women aged 20–75 years without cardiometabolic disease completed two six-week periods, one of adequate sleep and one of restricted sleep. Sleep was restricted experimentally. Its selected population limits direct application to people with existing disease or to men.
Cardiovascular and metabolic regulation
Zuraikat's trial included 38 women, including 11 postmenopausal participants (2024). The researchers reduced sleep duration during the restricted condition and reported impaired insulin sensitivity after six weeks, independent of changes in adiposity. Insulin sensitivity concerns the body's metabolic response to insulin; it is a physiological measure rather than an observed count of new diabetes cases. The trial did not measure new cases of diabetes.
In contrast, the cardiovascular evidence described by Ramar et al. (2021) includes associations between chronic insufficient sleep and disease risk. One study measuring insulin sensitivity in a controlled experiment does not directly measure heart attacks, cancer or mortality. Research into the Seven Countries Study and cardiovascular risk examines another route through which population observations can be interpreted, but it does not establish a sleep effect. The two studies measured different cardiovascular and metabolic endpoints.
Immune responses and infection
Immune outcomes have also been assessed directly. Spiegel and colleagues (2023) conducted a meta-analysis of sleep duration and antibody responses to vaccination. Among participants whose sleep was measured objectively, shorter sleep was associated with a decrease in antibody response; the analysis included 304 people aged 18–60 years, with an effect size of 0.79 (95% confidence interval 0.40–1.18). The endpoint was an antibody response following vaccination, rather than infection incidence.
The same review produced a less decisive result for questionnaires. With self-reported short sleep below 6 hours per night, the association with antibody response was not statistically significant among 504 participants aged 18–85 years; the effect size was 0.29 (95% confidence interval −0.04–0.63) (Spiegel et al., 2023). The questionnaire evidence was less conclusive.
Spiegel and colleagues (2023) called for larger controlled studies to clarify the period around vaccination when sleep might matter and to investigate differences between groups. Meanwhile, Ramar et al. (2021) include healthy sleep within general health and disease prevention discussions. An antibody measurement supports a narrower conclusion than saying that every additional hour of sleep necessarily prevents an infection.
What happens after insufficient or disrupted sleep
The effects of short sleep can accumulate even when subjective sleepiness changes little. Van Dongen and colleagues (2003) studied 48 healthy adults aged 21–38 years using controlled sleep-restriction and sleep-deprivation experiments. Participants in the restriction experiment were assigned opportunities of 4, 6 or 8 hours in bed on each of 14 consecutive days. Performance was then assessed across repeated testing rather than from a single morning's impression.
The laboratory comparison showed cumulative, dose-dependent deficits in cognitive performance in the 4-hour and 6-hour groups over the 14-day restriction period (Van Dongen et al., 2003). That is a finding about measured performance under a defined protocol. It does not imply that every person's impairment is immediately apparent to them or that a single night can be assigned the same effect as repeated restriction.
Attention, reaction time and safety
Van Dongen's participants reported an initial rise in sleepiness, but ratings changed relatively little on later restricted nights (2003). The researchers concluded that participants were largely unaware of their growing cognitive deficits. Subjective alertness can therefore fail to reflect performance measured on laboratory tasks during repeated sleep restriction. The difference is relevant to interpreting self-ratings of readiness for tasks demanding sustained attention.
The American Academy of Sleep Medicine statement by Ramar et al. (2021) links insufficient sleep with risk of accidents and injuries related to sleepiness and fatigue, including driving and workplace incidents. The statement concerns population safety as well as individual health. Laboratory tasks and real-world accidents are not interchangeable endpoints, although the former can identify a deficit in cognition before a serious event occurs.
Sleep duration, continuity and timing
Hours slept are one dimension of sleep health. Ramar and colleagues (2021) describe healthy sleep as adequate duration together with good quality, appropriate timing, regularity and the absence of sleep disorders. The CDC (2024) describes quality sleep as uninterrupted and refreshing, and lists repeated night waking and persistent tiredness as signs of poor quality. A duration figure alone cannot record those characteristics.
The distinction is relevant to sleep studies that use self-reported data. Van Dongen et al. (2003) found that sleepiness ratings did not keep pace with objective cognitive decline under continued restriction. The article on the accuracy of self-reported participation rates examines a separate measurement problem involving reports rather than sleep. Neither self-report in general nor a particular sleep questionnaire can be treated as equivalent to an instrumented measurement without validation for that outcome.
The CDC (2024) describes sleep disorders as health conditions that prevent some people from getting enough quality sleep. This makes persistent disruption distinct from a deliberately shortened sleep opportunity in a laboratory. Research comparing sleep stages, questionnaires and cognitive tests must specify which property was actually measured, rather than assuming the same duration implies the same biological exposure.
Sleep disruption can also arise in specific physiological contexts. A separate article on perimenopause symptoms and hormonal changes examines how night sweats and hormonal fluctuations can coincide with disturbed sleep.
What is proven about sleep benefits and what is not
Sleep supports memory and emotional functioning, and restricting it produces measurable changes under controlled conditions. Crowley and colleagues (2024) found impaired memory formation across human experiments, while Palmer and colleagues (2024) found reduced positive affect and heightened anxiety symptoms after manipulated sleep loss. These outcomes are supported more directly than broad claims that sleep prevents every condition associated with short nights.
Experimental sleep loss changes memory and mood, while mouse evidence on brain clearance cannot establish disease prevention in people (Crowley et al., 2024; Palmer et al., 2024; Xie et al., 2013).
A comparison of the experiments shows why the distinction matters:
Study | Evidence and participants | Measured outcome | Finding |
Van Dongen et al., 2003 | 48 adults across two experiments; 14-day sleep-restriction protocol | Repeated cognitive performance tests | Performance deficits accumulated in the restricted-sleep conditions. |
Crowley et al., 2024 | 39 reports; 1,234 participants | Memory formation | Restriction produced a small adverse effect (Hedges' g = 0.29; 95% confidence interval 0.13–0.44). |
Palmer et al., 2024 | 154 studies; 5,717 participants | Emotional responses | Positive affect fell and anxiety symptoms increased after sleep loss. |
Zuraikat et al., 2024 | 38 women; 6-week restricted-sleep condition | Insulin sensitivity | Sensitivity was impaired, independent of adiposity. |
Spiegel et al., 2023 | 304 participants with objectively assessed sleep | Antibody response to vaccination | Short sleep was associated with a lower response (effect size 0.79; 95% confidence interval 0.40–1.18). |
The table contains interventions, pooled experiments and associations that measure different outcomes. Van Dongen et al. (2003) directly manipulated sleep opportunity and tested cognitive performance. Spiegel et al. (2023) pooled associations between measured sleep and antibody responses. Their designs do not justify interpreting a lower antibody response as proof that sleep duration alone caused a particular infection.
The stage-specific memory account also has an unresolved boundary. Rasch and Born's 2013 review of human and animal studies describes reactivation during slow-wave sleep and possible stabilisation during subsequent REM sleep. Schäfer et al. (2020), however, did not find a reliable selective advantage for emotionally charged memories over neutral ones in their sleep-versus-wake comparisons. A plausible process does not determine the size of an emotional-memory effect in every experiment.
Xie et al. (2013) measured greater fluid exchange and faster amyloid beta clearance during sleep in mice. McKinley et al. (2013) cautioned that the invasive procedures could not be transferred directly to humans. A further conference supplement report by Vasciaveo and colleagues (2024) described changes in cognitive testing after sleep fragmentation in wild-type mice and plans to examine glymphatic mechanisms. That preliminary mouse evidence does not establish Alzheimer's prevention in people.
Better separation of these questions requires studies with clear sleep measurements and outcomes matched to the claim. Spiegel and colleagues (2023) specifically identified uncertainties around vaccination timing and the amount of sleep associated with antibody response. Ramar and colleagues (2021) also call for more sleep and circadian research into public health and health disparities. Neither statement replaces the stronger experimental findings already available for memory, mood and short-term performance.
Frequently asked questions
How much sleep do adults need?
For adults aged 18–60 years, the CDC (2024) recommends 7 or more hours each night. That is a population guideline, not an individual guarantee of feeling rested. Ramar et al. (2021) also identify timing, regularity, quality and sleep disorders as relevant. Adequate duration cannot compensate automatically for every form of disrupted sleep.
Why do children and teenagers need more sleep?
Sleep needs change with age, according to the CDC (2024), which lists 9–12 hours for school-age children aged 6–12 and 8–10 hours for teenagers aged 13–17. The NINDS (2024) also identifies growth and brain development as possible functions of the longer sleep observed in babies. Needs still vary among individuals.
Why do people dream?
The exact purpose of dreaming remains unknown. NINDS (2024) describes emotional processing as a possible function rather than an established explanation, and notes that dreams occur in several sleep stages, often most vividly during REM sleep. Evidence that sleep supports memory does not prove that consciously experienced dreams perform that function.
Can lost sleep be made up later?
Extra sleep may replace some lost sleep, but recovery depends on how substantial the deprivation was. NINDS (2024) notes that longer weekend sleep may be insufficient to replace missed sleep. Crowley et al. (2024) also caution that catch-up sleep after restriction may not eliminate cognitive costs. The recovery question remains outcome-specific.
Does sleeping longer always mean better sleep?
No. Ramar and colleagues (2021) describe healthy sleep as involving quality, timing, regularity and the absence of sleep disorders alongside duration. CDC (2024) identifies repeated awakenings and tiredness despite sufficient hours as signs of poor sleep quality. A longer reported night does not necessarily establish uninterrupted or refreshing sleep.
Sources
- Ramar K, Malhotra RK, Carden KA, et al. Sleep is essential to health: an American Academy of Sleep Medicine position statement. Journal of Clinical Sleep Medicine. 2021;17(10):2115–2119. DOI 10.5664/jcsm.9476.
- Borbély AA, Daan S, Wirz-Justice A, Deboer T. The two-process model of sleep regulation: a reappraisal. Journal of Sleep Research. 2016;25(2):131–143. DOI 10.1111/jsr.12371.
- Rasch B, Büchel C, Gais S, Born J. Odor cues during slow-wave sleep prompt declarative memory consolidation. Science. 2007;315(5817):1426–1429. DOI 10.1126/science.1138581.
- Crowley R, Alderman E, Javadi AH, Tamminen J. A systematic and meta-analytic review of the impact of sleep restriction on memory formation. Neuroscience & Biobehavioral Reviews. 2024;167:105929. DOI 10.1016/j.neubiorev.2024.105929.
- Van Dongen HPA, Maislin G, Mullington JM, Dinges DF. The cumulative cost of additional wakefulness: dose-response effects on neurobehavioral functions and sleep physiology from chronic sleep restriction and total sleep deprivation. Sleep. 2003;26(2):117–126. DOI 10.1093/sleep/26.2.117.
- Spiegel K, et al. A meta-analysis of the associations between insufficient sleep duration and antibody response to vaccination. Current Biology. 2023;33(5):998–1005.e2. DOI 10.1016/j.cub.2023.02.017.
- Xie L, Kang H, Xu Q, et al. Sleep drives metabolite clearance from the adult brain. Science. 2013;342(6156):373–377. DOI 10.1126/science.1241224.
- Centers for Disease Control and Prevention. About Sleep. 15 May 2024. https://www.cdc.gov/sleep/about/.
- McKinley J, McCarthy A, Lynch T. Don’t Lose Sleep Over Neurodegeneration-It Helps Clear Amyloid Beta. Frontiers in Neurology. 2013;4:206. DOI 10.3389/fneur.2013.00206.
- Vasciaveo V, Casile A, Ferro A, et al. Memory consolidation circuits and the glymphatic system in sleep. Alzheimer’s & Dementia. 2024;20(Suppl 1):e089231. DOI 10.1002/alz.089231.
- Rasch B, Born J. About sleep’s role in memory. Physiological Reviews. 2013;93(2):681–766. DOI 10.1152/physrev.00032.2012.
- Palmer CA, Bower JL, Cho KW, et al. Sleep loss and emotion: a systematic review and meta-analysis of over 50 years of experimental research. Psychological Bulletin. 2024;150(4):440–463. DOI 10.1037/bul0000410.
- Zuraikat FM, Laferrère B, Cheng B, et al. Chronic insufficient sleep in women impairs insulin sensitivity independent of adiposity changes: results of a randomized trial. Diabetes Care. 2024;47(1):117–125. DOI 10.2337/dc23-1156.
- Schäfer SK, Wirth BE, Staginnus M, et al. Sleep’s impact on emotional recognition memory: a meta-analysis of whole-night, nap, and REM sleep effects. Sleep Medicine Reviews. 2020;51:101280. DOI 10.1016/j.smrv.2020.101280.
- National Institute of Neurological Disorders and Stroke. Brain Basics: Understanding Sleep. May 2024. https://www.ninds.nih.gov/sites/default/files/2025-05/understanding-sleep.pdf.
Related
How exercise affects self-concept and well-being in midlife
The Seven Countries Study: Diet and Heart Deaths in 1980
Perimenopause symptoms and the hormone changes behind them
How Accurate Are Self-Reported Recreation Participation Rates?
Oil pulling for teeth and what the studies show about oral bacteria
How much water you should drink each day and why needs vary
Impostor syndrome and why some people doubt their success