Creatine and Sleep in Young Women: What Research Shows

Why Young Women Are the Focus

Most of what we know about creatine comes from studies of men, and most of it concerns muscle. Dr. Ann Brown, an exercise physiologist at the University of Idaho, is part of a small group of researchers trying to change both of those facts. Presenting at the Creatine Conference 2025, she laid out a case for studying creatine in one specific and underserved group: adolescent girls and young women, and for studying an outcome that has almost nothing to do with lifting weights. That outcome is sleep.

Brown’s route into this question was unusual. She trained as a dancer before she trained as a scientist, and her early research focused on nutrition in female dancers, a population she describes as riddled with dietary problems. Protein interventions worked well in that group, but dancers are reluctant to add calories, which pushed her toward creatine as a low calorie option and into a collaboration with creatine researcher Dr. Darren Candow. That collaboration eventually produced the study at the center of this article: a randomized, double blind, placebo controlled trial testing whether creatine changes sleep in naturally menstruating young women.

The answer, published in the journal Nutrients in 2024, was a qualified yes. Women taking creatine slept longer on the nights after they lifted weights, but not on other nights, and not by their own subjective account. It is a small study with a small effect, and Brown herself frames it as a starting point rather than a conclusion. To understand why she thinks it matters, though, you have to follow the longer argument she makes about what happens to girls between puberty and their early twenties.

Puberty, Body Image, and Comparison

Brown defines adolescence as a rapid period of biological, cognitive, and psychological change toward becoming more adult, and she splits it into three stages: early (roughly 12 to 14), middle (15 to 16), and late (17 to 19 and beyond). She points out that college students, who are often treated as adults in research, sit squarely in that late stage.

The early to middle window is where girls and boys diverge most sharply. Girls typically reach peak height first, mature earlier, and begin accumulating body fat at a time when many boys have not yet started to change. Brown notes that fat can be surprisingly good at hiding: in her own work with collegiate dancers, a substantial share of active, normal weight young women met the definition of normal weight obesity, meaning a normal body mass index alongside a high body fat percentage. She cited a figure of around 40 percent in her talk. Her 2021 paper on the prevalence of normal weight obesity in female collegiate dancers, published in the Journal of Strength and Conditioning Research, documents this pattern in that population.

The timing of these changes matters because puberty does not arrive on a schedule. Girls who develop before their peers become targets of comparison, and Brown argues that this comparison is one of the earliest drivers of body dissatisfaction. The research supports her on how early it can start. A study of 135 children aged five to eight, published in the British Journal of Health Psychology, found that six, seven, and eight year old girls already rated their ideal figure as significantly thinner than their current one, while boys of the same age did not (Lowes and Tiggemann, 2003). A followup study by the same group found that by age six, 42 percent of girls expressed a desire to be thinner, and that girls’ beliefs about their peers’ body dissatisfaction predicted their own (Dohnt and Tiggemann, 2006).

Brown also raises the role of aesthetic sports. Gymnastics and dance can begin as early as age three, and while sport participation is broadly positive, sports that judge appearance add a layer of pressure. Social media adds another. Brown’s reading of the evidence is that passive use, meaning scrolling and comparing without posting or interacting, is worse for body image than active use. The largest study on this question, drawing on 190,892 adolescents aged 11, 13, and 15 across 42 countries in a large international school survey, found that intensive and problematic patterns of social media use were associated with perceiving oneself as too fat or too thin, and that this association was significant for girls in both categories but only for problematic users among boys (International Journal of Public Health, 2024). That study did not isolate passive versus active use exactly as Brown described it, but its findings are consistent with the broader point that how girls use social media, not just whether they use it, shapes how they see their bodies.

The Cycle of Dissatisfaction, Mood, and Sleep

Brown’s central concern is not body image on its own but what it sets in motion. She describes a cycle: body dissatisfaction feeds anxiety and depression, which in turn drive behaviors like restrictive eating and reduced physical activity, which disrupt sleep, which can alter the menstrual cycle and feed back into mood. Each piece worsens the others.

The first link in that chain is well documented. A prospective analysis of 2,078 girls and 1,675 boys in the Avon Longitudinal Study of Parents and Children found that body dissatisfaction at age 14 predicted mild, moderate, and severe depressive episodes at age 18 in girls, after controlling for depression at baseline (Journal of Epidemiology and Community Health, 2021). Among girls, each step up in dissatisfaction at 14 was linked to a 63 to 84 percent higher chance of a depressive episode at 18, depending on severity.

Brown states that young females face roughly twice the risk of depression and anxiety, as well as sleep disturbances, compared with males. The depression figure holds up well. A meta analysis of nationally representative samples covering more than 1.7 million people in over 90 countries found that women were nearly twice as likely as men to be diagnosed with major depression overall, with an odds ratio of 1.95 (Psychological Bulletin, 2017). Crucially for Brown’s argument, the gap was already present at age 12 and peaked between 13 and 15, when girls were about three times as likely as boys to meet criteria. The sex difference in depression is, in other words, largely an adolescent phenomenon, which is exactly the window Brown wants to target.

She cites one figure in particular that she finds striking: in a sample of roughly 650 girls with an average age of 13 and a half, body dissatisfaction was highest not in those with a high BMI but in those at a weight to height ratio considered healthy or even underweight. Boys showed a more linear pattern, with dissatisfaction rising as BMI rose. Brown did not name the study in her talk, and we were unable to identify it independently, so the specific numbers should be treated as her report rather than an independently verified finding. Her broader claim, that girls can be deeply dissatisfied with bodies that are objectively healthy, is consistent with the childhood body image literature above.

In that same sample, Brown reports that 17 percent of girls said they often tried to restrict their eating. She calls that number too many, and connects it directly to creatine: girls who restrict food are likely to have low energy intake overall and, because creatine comes almost entirely from meat and fish, low creatine intake specifically.

From there the cycle turns to sleep. Brown notes that reduced physical activity is associated with worse perceived sleep quality, and that inadequate sleep and inactivity together are strongly associated with mental health risk. She also draws a distinction that becomes important later: sleep quantity and sleep quality are different things, and studies sometimes find that total sleep time changes while participants report feeling no different. Any serious sleep study, she argues, needs to measure both.

The Menstrual Cycle and Sleep

If you are going to study sleep in women, Brown says, you have to account for the menstrual cycle. She points to the work of sleep researcher Fiona Baker, whose group has spent two decades describing how the circadian rhythm is superimposed on the hormonal rhythm of the cycle.

The pattern Brown describes runs roughly as follows. In a textbook ovulatory cycle, core body temperature rises after ovulation and stays elevated through the luteal phase, the second half of the cycle. REM sleep, expressed as a share of total sleep, tends to be lowest during this same luteal window. Subjective sleep quality tends to be worst around menstruation itself, which Brown attributes to cramping, fatigue, and general discomfort rather than to hormones directly.

Baker’s reviews support this picture with some caveats. A 2018 review in Sleep Medicine Clinics summarizes that self reported sleep disturbance increases in the premenstrual and menstrual phases, particularly in women with premenstrual symptoms or painful periods, and that the luteal phase brings a blunted temperature rhythm, more sleep spindle activity, and reduced REM sleep (Baker and Lee, 2018). The same review is careful to note that not every study finds a menstrual cycle effect on sleep quality, and that women vary considerably: some show no relationship, some struggle mid cycle, and some struggle premenstrually. A separate study using the Oura ring in healthy women, cited in Brown’s own paper, found that sleep did not vary significantly by cycle phase. The consensus is that cycle effects on sleep are real but modest and uneven, which is a reason to document cycle phase rather than a reason to expect dramatic differences.

Brown raises two further complications. Hormonal contraceptives change the picture entirely: users have elevated body temperature relative to naturally cycling women and a different REM profile, so any sleep study needs at minimum to record whether participants are on them. And disrupted circadian rhythms, whether from jet lag or shift work, can feed back into both the menstrual cycle and mood. Her practical advice to researchers is blunt: even if your question has nothing to do with the menstrual cycle, collect cycle phase and contraceptive status anyway, because you will want to answer questions about them once your data come in. She adds, with some frustration, that when she called a local gynecology practice looking for a tool to screen out perimenopausal participants, the best they could offer was asking whether a woman had missed periods for six months.

How Creatine Might Affect Sleep

The mechanistic case for creatine and sleep starts with brain energy. Brown explains that the first hours of sleep bring a burst of brain activity and a high demand for ATP and phosphocreatine, the high energy phosphate molecules that creatine helps regenerate. She describes this surge as positively associated with REM sleep: the larger the surge, the more of the night spent in REM.

The surge itself is well established, at least in animals. A 2010 study in the Journal of Neuroscience reported that ATP levels in wake active regions of the rat brain rise sharply in the initial hours of spontaneous sleep, that the rise is abolished by keeping the animals awake, and that it tracks the depth of sleep (Dworak and colleagues, 2010). The detail that differs from Brown’s account is which kind of sleep the surge tracks: in that study, the ATP surge correlated with slow wave activity during non REM sleep, not with REM. Whether an analogous link to REM exists in humans is not established. The broader point, that early sleep is metabolically demanding and that creatine sits at the center of that metabolism, stands.

The second strand of evidence comes from sleep deprivation. Brown notes that creatine metabolites are altered during sleep deprivation in both humans and animals, and that a single dose of creatine appears able to partially reverse the cognitive decline that comes with staying awake. This refers to a 2024 study in Scientific Reports in which 15 healthy adults stayed awake for 21 hours after taking either placebo or a single very large dose of creatine monohydrate, 0.35 grams per kilogram of body weight, roughly 25 grams for a 70 kilogram person. Creatine improved cognitive performance and processing speed, changed brain phosphocreatine and ATP measures on phosphorus spectroscopy, and prevented a drop in brain pH. The effects appeared around three hours after the dose and lasted several hours. The authors themselves caution that a dose this large is not advisable outside a research setting. Brown’s framing is accurate: this is evidence that creatine helps the sleep deprived brain function, not evidence that it improves mental health broadly.

She adds a personal note that captures the gap in the literature: she would rather have better sleep without being deprived of it first. Almost all creatine and sleep research has been done in the context of deprivation. Whether creatine changes ordinary, everyday sleep in people who are not being kept awake is a different and much less studied question.

Brown then lists what population data suggest. First, most people do not get much creatine from food. Analyses of the U.S. National Health and Nutrition Examination Survey (NHANES) put average dietary creatine intake at roughly 0.7 to 0.9 grams per day, with intake among children and adolescents declining over the two decades to 2018 (Journal of the International Society of Sports Nutrition, 2021). Among older adults, about 70 percent consume less than one gram per day. Brown’s figure of almost 70 percent falling below one gram is consistent with these analyses, though the exact proportion depends on the age group and survey year.

Second, creatine may help with minor sleep complaints rather than clinical disorders. A cross sectional NHANES analysis of 5,988 people aged 16 and over found that those consuming less than one gram of dietary creatine per day were more likely to report trouble sleeping than those consuming more (23.7 percent versus 19.3 percent, an odds ratio of 1.30), but that rates of more serious sleep disorders did not differ between the groups (Nutrition and Health, 2024). Brown is explicit that nobody expects creatine to resolve insomnia. Her interest is in people like herself: no diagnosed sleep problem, but room to sleep better.

Third, Brown mentions, half jokingly, that higher creatine intake has been linked to more snoring, and that this gives her pause about her own supplement habit. We could not locate a published source reporting this association, and the NHANES sleep analysis above does not mention it in its published abstract. We treat it as unverified.

Fourth, she suggests creatine may be involved in melatonin production, which governs sleep initiation. This is the weakest link in the mechanistic chain. The one human study that directly measured melatonin during creatine supplementation, a 2007 trial in Physiology and Behavior in which participants took 20 grams of creatine daily for a week before 36 hours of sleep deprivation, found no difference in salivary melatonin between the creatine and placebo groups (McMorris and colleagues, 2007). Creatine did improve performance on a demanding executive function task at the 36 hour mark, but the melatonin hypothesis, as far as the published evidence goes, is unsupported.

Inside the Idaho Study

Brown’s team set out to answer a specific question: in healthy young women who menstruate naturally, does creatine combined with resistance training improve sleep more than resistance training alone?

They enrolled 27 women aged 18 to 24 who were not regular resistance trainers and who were not using hormonal contraception. Vegetarians and vegans were excluded, as were current creatine users. Participants were matched on body weight, physical activity, and habitual animal protein intake before being randomized, so that one group would not end up with all the low creatine eaters. Menstrual cycle phase was recorded at the start, and ovulation was tracked daily with urine test strips, which Brown describes as a logistical challenge that any researcher studying women should be prepared for. Six women withdrew, leaving 21 who completed the study: 12 on creatine and 9 on placebo.

The creatine group took 5 grams of creatine monohydrate plus 5 grams of maltodextrin each day; the placebo group took 10 grams of maltodextrin. Both were white powders mixed into flavored water. Brown’s lab took an unusually strict approach to compliance: participants came in every weekday to take their dose under supervision and sent video evidence on weekends. Everyone trained twice a week for six weeks on a Tonal, a wall mounted electromagnetic resistance machine, following the same prescribed full body program, so that exercise was held constant across both groups.

Sleep was measured every night with a second generation Oura ring, which estimates total sleep time and time in light, deep, and REM sleep. Brown notes that her team did not use the ring’s temperature data. Before and after the six weeks, participants also completed the Pittsburgh Sleep Quality Index, a validated questionnaire on perceived sleep quality; a food frequency questionnaire covering the previous 30 days; strength testing; and body composition scans.

What the Results Mean and What They Do Not

The headline finding, as Brown tells it, is that on nights following a resistance training session, the creatine group slept roughly an added hour compared with placebo, while on non training nights there was no difference at all. The published numbers are a little smaller than an hour but tell the same story. Using a statistical model that accounted for repeated nightly measurements, the creatine group averaged about 445 minutes of total sleep on training nights versus about 397 minutes for placebo, a difference of roughly 48 minutes that was statistically significant. On non training nights the two groups were within ten minutes of each other.

Brown is candid that she did not anticipate this. When the analysis came back, she recalls emailing Candow to say that either they had made a mistake or found something interesting. She also notes that the effect was specific to total sleep: there were no differences in deep, light, or REM sleep, and the questionnaire results showed no change in how participants rated their own sleep quality, either over time or between groups. In fact, most participants in both groups scored in the poor sleep range on the questionnaire at both the start and the end. That mismatch between objective duration and subjective quality is exactly the split Brown warned about earlier, and she speculates that the extra sleep might have shown up as better cognitive function or mood the next day, but her team did not measure those things because they did not expect to need to.

The secondary outcomes were unremarkable. Both groups got stronger over six weeks, with no advantage for creatine. Body composition did not differ between groups. Interestingly, both groups ate less over the course of the study, with drops in total calories, protein, carbohydrate, and fat that the authors could not explain, and estimated habitual creatine intake from food was below half a gram per day at both time points.

It is worth being clear about what this study can and cannot support. It is a single trial of 21 people, run for six weeks, with a modest difference in one outcome on a subset of nights. The authors themselves flag that six weeks may be too short to fully saturate creatine stores, that the training program may not have been intense enough, and that the small sample limits the ability to detect other effects. The result has not been replicated. Brown’s own conclusion is measured: creatine may increase acute total sleep when combined with resistance training in naturally menstruating young women, and the additive effect of the two deserves further study. That is an honest reading of the data, and it is why this article rates the evidence as early rather than promising or strong.

There is also a wrinkle in the animal literature that Brown’s paper acknowledges. A 2017 rat study found that four weeks of creatine supplementation reduced total sleep duration and sleep pressure, the opposite direction from the human result. Whether that reflects a species difference, a dose difference, or something about the interaction with exercise is unknown.

What the finding does offer is a plausible and testable idea: that creatine’s effects on the brain may be most visible under stress, whether that stress is sleep deprivation or a hard training session, and that a night of recovery after training is where a difference is most likely to show up. Independent evidence that resistance training itself improves sleep is fairly good. A 2018 systematic review of randomized trials in Sleep Medicine Reviews concluded that chronic resistance exercise improves most aspects of sleep, with the largest effect on sleep quality, though it noted that results in younger adults were less consistent than in older ones and that acute effects of a single session remain poorly studied (Kovacevic and colleagues, 2018). Brown’s contribution is the suggestion that creatine may add something on top of training, on the specific nights when the body has the most to recover from.

Diet, Habitual Creatine Intake, and Education

One of the most practical threads in Brown’s talk concerns how little creatine her participants were eating. Estimated habitual intake was below 0.5 grams per day before and after the study, even though every participant was an omnivore. An ordinary meat eating diet is often expected to supply two to four grams per day, and even the more conservative NHANES estimates put the population average near one gram. Brown calls this far too low, and she suspects it partly reflects total energy intake sitting below 2,000 calories per day in both groups.

She frames this as an opportunity for education, and one that extends well beyond physicians and the public to research participants themselves. When a young woman enrolls in a study, she is a captive audience for learning what the study is about and why it matters. When the participant is a minor, the parent consents too, which doubles the audience. Brown’s caution is about timing: educating participants mid study risks changing their behavior and contaminating the data, so the conversation belongs at the end.

Her position on food versus supplements is deliberately balanced. She believes people can reach adequate creatine intake through whole foods, and she promotes a food first approach, but she also says plainly that supplementation is always an option and that low dose supplementation could reasonably fill the gap left by a habitual diet that falls short. For the purposes of sleep, she keeps returning to the same qualifier: the benefit her study found came with resistance training, not from creatine alone.

There is a sex difference underlying all of this. Reviews of creatine in women’s health frequently state that women have substantially lower endogenous creatine stores than men and consume less dietary creatine, largely because they eat less meat, and that estrogen and progesterone influence creatine synthesis, transport, and creatine kinase activity across the cycle (Nutrients, 2021; Journal of the International Society of Sports Nutrition, 2025). These are the reasons researchers like Brown argue that creatine may matter more for women than the male dominated literature suggests, though the direct evidence in women is still catching up.

Creatine and Mood in Young Women

Brown is careful not to overreach on mental health. Her study measured sleep, not mood, and she says outright that connecting the two is a leap her data cannot make. She did not administer depression or anxiety scales, and she notes that mental health does not change in a day; any real test would need to be longitudinal.

That said, the reason she cares about sleep is its position in the cycle she described earlier. If body dissatisfaction raises the risk of anxiety and depression in young women, and those in turn disrupt sleep, then anything that adds sleep might, in principle, ease other links in the chain: mood, behavior, cognitive function the next day. That is a hypothesis, and Brown presents it as one.

There is some independent, preliminary evidence that creatine and adolescent female mood are connected. A small open label study at the University of Utah gave 4 grams of creatine daily for eight weeks to five adolescent girls whose depression had not responded to fluoxetine; their depression scores fell by more than half and brain phosphocreatine rose on spectroscopy, but there was no placebo group (Journal of Affective Disorders, 2011). A subsequent placebo controlled dose ranging study in the same population found that 2, 4, and 10 grams daily raised frontal lobe phosphocreatine in a dose dependent way, with the largest change at 10 grams, and no difference in adverse events (Amino Acids, 2016). These are small trials in a clinical population, and they measured brain chemistry more convincingly than they measured symptom change, so they support the biological rationale without establishing a treatment effect. A cross sectional NHANES analysis has also reported that adults with higher dietary creatine intake had lower rates of depression, though that design cannot show cause and effect.

Brown’s next project is designed to start filling this gap. She describes a new collaboration with Dr. CJ Brush, an exercise psychologist at Idaho, examining the relationship between habitual dietary creatine, cognitive function, and symptoms of depression and anxiety in girls aged 10 to 17. The study uses EEG during a response inhibition task, a food frequency questionnaire to estimate creatine intake, and standard pediatric screening tools for depression and anxiety. Brown believes it may be among the first studies to pair EEG with habitual creatine in a young female population, and the age range is chosen deliberately to span early, middle, and late puberty, so that if the sample is large enough the three stages can be compared.

This work sits within a larger NIH Center of Biomedical Research Excellence grant on nutrition and women’s health at the University of Idaho, on which Brown is one of the research project leads and Dr. Abbie Smith Ryan serves as her mentor. Brown referred to it as a 12 million dollar award; the university’s announcement puts the five year figure at just over 11 million dollars.

Open Questions and Future Research

Brown closes with a list of what she would like to see studied next, and it doubles as a fair summary of how much remains unknown.

Hormonal contraceptives come first. There are many formulations, users have elevated body temperature relative to naturally cycling women, and their REM profile differs. Whether creatine behaves any differently in this group, and whether it could moderate any of the sleep or mood symptoms some users report, is entirely open. Her study excluded contraceptive users on purpose, which was the right call for a first trial but leaves the larger population of young women unstudied.

Dosing is second. Five grams per day is a standard adult dose borrowed largely from male performance research. Whether a younger or smaller female needs less, whether needs change across puberty, and whether a loading phase would have produced a bigger sleep effect are all untested. Brown’s paper itself suggests that future work should try loading protocols and higher doses.

Third is timing across the lifespan. Brown has focused on adolescence and young adulthood, but she notes that body dissatisfaction and sleep disturbance also peak in perimenopause, where the evidence is equally thin. She frames early intervention as prevention: if girls can be given tools to manage sleep and mood before the late stages of puberty, they may carry those tools into later transitions.

Finally, she raises a mechanistic question outside her own expertise: whether creatine kinase activity rises and falls with estrogen across the menstrual cycle, as some reviews suggest, and whether it follows an individual’s actual hormonal pattern rather than the idealized textbook one. If it does, then the same dose of creatine might do different things at different points in the month, which would complicate both research design and practical advice.

For readers deciding what to do with all this, the honest summary is short. Creatine is well studied and generally safe at ordinary doses, and there is a reasonable biological case that it matters for the female brain. One small, carefully run trial found that young women taking 5 grams a day slept about 45 to 60 minutes longer on nights after strength training, without feeling any different, and not on other nights. That is an interesting result, not a proven benefit. Resistance training on its own has better evidence for improving sleep than creatine does, and most young women are eating very little creatine to begin with. If you are already lifting and already considering creatine, sleep is one more reason to be curious. It is not yet a reason to expect a change.