Creatine
Creatine is one of the best-studied and best-tolerated supplements, and taken alongside resistance training it consistently adds a modest amount of muscle and strength. Its other promises — protecting thinking during acute stress such as sleep loss, or helping blood sugar, bone or everyday thinking in older age — range from tentative to unproven.
For decades creatine was filed under "sports nutrition" — a powder bodybuilders used to lift heavier. The best evidence still sits there: combined with resistance training, creatine adds fat-free mass and strength. Some researchers now describe it as a conditionally essential nutrient, but the case for benefits beyond muscle is much weaker — a small, still-tentative signal for thinking under acute stress such as sleep loss, and unproven effects on blood sugar, bone and cognition in older age. Its safety record in trials is extensive (685 clinical trials with more than 12,800 people taking creatine, some lasting up to 14 years).
What creatine actually does
Cells run on adenosine triphosphate (ATP), and they burn through it fast — particularly muscle and brain tissue under load. Creatine works by buffering local energy supply through the phosphocreatine shuttle: it carries a high-energy phosphate group from the mitochondria (where ATP is generated) to the contractile machinery in muscle and the synaptic terminals in neurons (where ATP is consumed). Locally, that means a few extra seconds of full-power energy availability before the slower energy systems take over.[1]
Recent work has expanded the picture. Beyond the energy buffer, sustained creatine availability appears to support mitochondrial proteostasis (quality control of mitochondrial proteins), reduce oxidative stress by sparing antioxidant cofactors like glutathione, and amplify the systemic release of muscle-derived signalling molecules (myokines) like brain-derived neurotrophic factor (BDNF) and insulin-like growth factor 1 (IGF-1) during exercise — the same molecules that support neurogenesis and synaptic plasticity in the brain.[2]
The body makes roughly half of what it needs (about 1–2 g/day, synthesised in the kidneys and liver from arginine, glycine and methionine); the rest comes from food, almost entirely meat and fish, which contain roughly 2–5 g of creatine per kilogram, so about 450 g (a pound) of beef or salmon delivers 1–2 g.[3] In U.S. national survey data, about 70% of adults over 65 ate less than 1 g of creatine on the day recorded, and about 1 in 5 ate none.[4] There is no official requirement. One recent proposal for adults aged 19–65 estimates that the body makes most of what it needs (about three-quarters, more than the older "about half" figure) and puts adequate dietary intake at roughly 0.24 g/day for a 60-kg woman and 0.4 g/day for a 75-kg man, while flagging older adults and vegetarians as at risk of falling short.[5] So low intake in older adults means a smaller dietary contribution, not a proven deficiency. Vegans and vegetarians get virtually none from diet.
What the evidence says
Strong — for muscle and strength when paired with resistance training; for safety in healthy people at usual doses (see Safety for people with kidney disease, and Mood and depression for people with bipolar disorder).
Moderate — for cognitive resilience under acute metabolic stress (sleep loss, low oxygen, mental fatigue), where small trials from several research groups point the same way but there is no pooled estimate and the single-dose sleep-loss work comes from one group; for small extra gains in lean mass and strength in postmenopausal women taking at least 5 g/day alongside resistance training.
Weak / preliminary — for cognitive benefit in well-rested healthy young adults; for cognitive benefit in older adults (a signal from a few small trials) and in clinical populations such as Alzheimer's disease; for extra cognitive benefit in vegans and vegetarians; for blood-sugar control alongside exercise in type 2 diabetes; for bone in postmenopausal women, where a few measures of hip bone shape held up better but bone density did not; for depression as an adjunct; for emerging fatigue indications (long COVID, fibromyalgia, myalgic encephalomyelitis/chronic fatigue syndrome, or ME/CFS); for pregnancy and perimenopause.
Caution — large randomised controlled trials (RCTs) in motor neurodegenerative disease (Parkinson's, Huntington's, amyotrophic lateral sclerosis) were null or stopped for futility despite a strong mechanistic rationale; creatine is not a disease-modifying therapy for these conditions.
Muscle and strength
The skeletal-muscle case is the most thoroughly validated. A 2025 systematic review and dose-response meta-analysis (61 trials, lasting from a few days to a year) found that creatine combined with resistance training increased fat-free mass by about 1.4 kg and total body mass by about 0.9 kg, without any meaningful gain in fat mass. It found no dose-response for fat-free mass — more than 5 g a day added nothing extra — and the gain was smaller and not statistically significant in trials of women only.[6] A separate 2025 meta-analysis (69 trials, about 1,900 participants) put numbers on the strength side: creatine plus training added about 5.6 kg to the squat and about 1.4 kg to the bench or chest press, while gains in leg-press and handgrip strength were not statistically significant, and the squat gain reached significance in younger adults and in men but not in adults over 50 or in women.[7] The bench-press result showed signs of small-study bias (Egger's test p = 0.0003), although a standard correction found no missing studies and left the estimate unchanged, so treat that figure with some caution.
A useful 2025 methodological refinement: a University of New South Wales trial in 63 adults who had not lifted weights in the past year added a 7-day non-exercise "wash-in" period before 12 weeks of resistance training. The comparison group received no creatine and no placebo, and the trial was single-blind. After the wash-in, the creatine group (5 g/day) showed about 0.5 kg more lean mass on a body-composition X-ray scan (DXA) than controls, mostly in the trunk and statistically significant only in women; the authors think fluid is a plausible explanation, though they did not measure body water. Over the next 12 weeks of training both groups gained about 2 kg of lean mass, with no extra gain from creatine, and the authors suggest a daily dose above 5 g may be needed.[8] The takeaway: some of creatine's early effect on lean-mass scans may be water, so trials without a wash-in can overstate how much muscle it builds. The dose-response meta-analysis above does not show when any muscle advantage appears: experienced lifters gained somewhat more fat-free mass than novices (1.82 vs 1.23 kg), but the difference was not statistically significant and the two groups were studied in trials of very different lengths (about 6 vs 13 weeks).[9]
The practical implication for resistance training: creatine lets you do more work in a session. In a small 3-day crossover trial in 10 young men with a few months of lifting experience, a high short-term dose (0.3 g per kg of body weight per day) produced more reps at 60–80% of one-rep max, faster bar speed and less next-day soreness than placebo, though the one-week gap between conditions was short enough that some carryover can't be ruled out.[10] Extra work sustained over months and years is the plausible route from creatine to more muscle.
See Resistance training for the broader programming context.
Brain and cognitive performance
The most interesting development in creatine science over the past five years: the brain story.
The brain runs on roughly 20% of resting metabolic energy despite being only ~2% of body mass. Phosphocreatine is critical to maintaining that energy supply, especially under demand. But the brain's barrier (the blood-brain barrier) is much more selective than muscle's blood supply, and the brain appears to rely mainly on making its own creatine, so cerebral creatine stores are harder to load. Brain creatine rises much less than muscle creatine with supplementation (roughly 5–10% vs 20–40%), and trials show no clear dose-response: both short high-dose loading (about 20 g/day) and 4–5 g/day for several months have raised brain levels,[11] while one week at about 20 g/day did not raise brain phosphocreatine in another study.[12] The dose needed for brain effects is unknown.
When does cognitive benefit show up?
The honest answer requires a statistical caveat. A 2023 critique in Nutrition Reviews showed that an influential creatine-and-memory meta-analysis had double-counted multiple cognitive subtests from the same participants as if they were independent — a "unit of analysis" error that inflates the apparent effect size.[13] In well-rested, healthy young adults the effect is essentially absent: a dedicated dose-response trial found no cognitive benefit from either 10 or 20 g/day over six weeks in 30 young adults.[14] The meta-analysis authors accepted the critique and redid their analysis (below), and the broader read is that the cognition evidence base is weaker than popular coverage implies. Keep cognition claims tightly bounded: the more consistent signals come from people under acute metabolic stress such as sleep loss, and even those rest on small trials.
The signals that remain come from a few specific groups, and each is thinner than it first looks:
- Older adults: a 2023 meta-analysis of memory trials in healthy people first reported a small overall benefit,[15] but after the double-counting was pointed out the authors redid the analysis and found no significant overall effect on memory. Only the subgroup of adults aged 66–76 still showed a benefit, with a wide range of uncertainty, resting on a few small trials.[16] Assessing a health-claim application covering 23 human studies in 2024, the European Food Safety Authority concluded that a cause-and-effect link between creatine and better cognition has not been established.[17]
- Vegans and vegetarians: they get almost no creatine from food and have lower muscle stores, but a brain-scan study found their brain creatine was about the same as meat-eaters', which suggests the brain relies mainly on making its own.[18] An early trial reported better memory after five days of 20 g/day in young vegetarian women but not in meat-eaters,[19] but a larger preregistered trial (123 adults, about half vegetarian, 5 g/day for six weeks) found vegetarians did not benefit more than omnivores, and the overall benefit was small and fell short of statistical significance.[20]
- Clinical populations: in an uncontrolled 8-week pilot of 20 people with mild-to-moderate Alzheimer's disease, 20 g/day raised brain creatine by about 11% and scores on some thinking tests (notably working memory) improved, but with no placebo group the authors could not rule out practice or placebo effects.[21] In a cognitive sub-study of a small placebo-controlled trial in bipolar depression (18 of the trial's 35 patients, 6 g/day for six weeks), creatine improved verbal fluency but not working-memory or executive-function tests.[22]
Sleep deprivation: a promising single-dose signal
The most striking recent finding is a small 2024 trial of creatine during acute sleep loss. In a double-blind crossover study run by Forschungszentrum Jülich in Germany, 15 healthy young adults took a single high dose (0.35 g per kg of body weight, roughly 25–30 g for a 70–85 kg adult) or placebo in the evening and then stayed awake through the night, about 21 hours without sleep in all.[23]
Using phosphorus and proton magnetic resonance spectroscopy (scans that measure the brain's chemical energy stores) at three points during the night, the researchers tracked brain pH, the phosphocreatine-to-inorganic-phosphate ratio and ATP. Creatine largely prevented the small drop in brain pH (about 0.03 units under placebo) and the fall in the phosphocreatine-to-phosphate ratio that sleep loss caused. ATP did not hold steady, though: in the upper brain slice its signal fell by about 9% relative to placebo, which the authors attribute mainly to extra creatine shifting the creatine kinase reaction, the brain's energy buffer, toward less free ATP; overall they describe a partial reversal of the metabolic changes caused by sleep loss. On thinking tests, creatine improved word-memory scores by about 10% relative to placebo and made people faster on word-memory, language, logic and number tasks, by roughly 16–29%, and later in the night they rated themselves about 8% less fatigued. Digit span, working memory (the N-back test), logic accuracy and a reaction-time vigilance test were not reported as significantly better than placebo, and each effect was estimated by pooling three test sessions from the same 15 people, which makes the statistics look stronger than they are. Brain scans began about 3 hours after the dose and the first thinking tests followed at about 4 hours, so earlier onset was never tested; the authors put the largest cognitive effect at about 4 hours, which was also the first test, and benefits were still present when the last session ended about 9 hours after dosing.[24] A follow-up by the same group in 29 healthy adults using a lower dose (0.2 g/kg) found smaller benefits, including on reaction-time vigilance.[25]
That makes acute high-dose creatine a promising option to test for forced wakefulness, such as night shifts or emergency work, but the evidence is still thin: the single-dose findings come from two small crossover trials (15 and 29 healthy adults) from one research group, with no independent replication yet, and a 2026 systematic review judged sleep-deprivation research on creatine sparse.[26]
For chronic sleep regulation, the underlying problem is night-time sleep itself, not creatine — see Sleep. Creatine is at most an occasional stopgap, not a substitute.
Where creatine has failed: motor neurodegenerative disease
The bioenergetic rationale made creatine an obvious candidate for slowing neurodegeneration — and large, well-powered trials tested it directly and came up empty. The NET-PD LS-1 Parkinson's trial randomised 1,741 patients to 10 g/day of creatine or placebo, planned for at least five years of treatment, and was stopped for futility after a median of about four years, with no effect on clinical progression.[27] The CREST-E Huntington's trial (553 patients, up to 40 g/day) was likewise halted for futility, with the decline trend if anything slightly favouring placebo.[28] In amyotrophic lateral sclerosis (ALS), three trials with 386 patients taking 5–10 g/day found no effect on survival or disease progression.[29] The honest read: creatine has repeatedly failed in motor neurodegenerative disease despite strong mechanism. The genuine clinical wins are elsewhere — the inborn disorders that block the body's own creatine synthesis (AGAT and GAMT enzyme deficiencies) are treated with high-dose oral creatine, with the best outcomes when treatment starts early in life.[30] Creatine transporter deficiency responds much less, because creatine cannot easily get into brain cells; in a review of 28 reported patients, most treated with creatine plus the amino acids arginine and glycine, about a third improved, mostly milder cases that started treatment young.[31]
Creatine across the female lifespan
Most creatine research was done in men, and the female-specific picture has only recently been pulled together. Some reports suggest women synthesise about 20% less creatine than men, and on average they eat 30–40% less of it, which is the basis for the argument that they may be more responsive to supplementation.[32] Endogenous synthesis, transport, and creatine-kinase expression also shift with estrogen across the menstrual cycle, pregnancy, and menopause.
The strongest case is the menopause transition and beyond, where creatine taken alongside resistance training modestly adds to muscle and strength gains; it does not appear to raise or preserve bone density (see below). Benefits for sleep, fatigue, and mood in this group are untested or preliminary: the first trial to include perimenopausal women (36 women split across four groups, 8 weeks, a creatine hydrochloride product) found faster reaction times and higher brain creatine on the 1,500 mg/day dose, but no clear effect on mood swings.[33] Pregnancy and perimenopause remain under-researched: human pregnancy work is still gathering dosing and tolerability data ahead of any efficacy trial. The human evidence so far is observational: in US national survey data, women whose diets supplied at least 13 mg of creatine per kg of body weight a day reported fewer pregnancy complications,[34] while an Australian cohort of 282 low-risk pregnancies found maternal blood creatine stayed stable and was not linked to newborn growth.[35] (Note: the women's-health review cited here is a narrative review whose publication costs were paid by a creatine manufacturer, and two of its six authors advise the company; its framing leans favourable — for example, it cites the two-year bone trial for bone shape but not for its null bone-density result.)
Bone health
The early signal came from a small 12-month trial: postmenopausal women taking creatine (about 0.1 g per kg of body weight per day) with supervised resistance training three times a week lost less bone mineral density (BMD) at the femoral neck, the top of the thigh bone, than those on placebo (1.2% vs 3.9%), but only 33 of the 47 women randomised were analysed.[36] The same group then tested this properly in 237 postmenopausal women over two years (0.14 g/kg/day plus resistance training and walking) and found no effect on BMD at the femoral neck, whole hip or spine. Creatine did hold steady a few measures of hip bone shape linked to bending strength and made a fast 80 m walk slightly quicker, but did not improve strength; the authors judged the shape changes too small to prevent fractures on their own, and many women dropped out or took the supplement irregularly.[37]
The important nuance: the one small positive trial was not confirmed by a larger one. A 2018 meta-analysis of five small trials had already found no effect on bone density, and noticed that the only positive trials used weight-based dosing with frequent training;[38] the two-year trial above used exactly that approach and still found no effect. Later pooled analyses agree. Among trials in older adults that measured whole-body bone density, adding creatine to exercise made no difference,[39] and a 2026 review of trials in postmenopausal women found bone density unchanged even though lean mass and strength improved slightly.[40] No trial has been large enough to test fractures, so any fracture benefit is unproven. Do not rely on creatine to raise or protect bone density.
Mood and depression
There is a real but oversold signal here. A landmark 8-week RCT in 52 women with major depression found that adding 5 g/day creatine to escitalopram roughly doubled remission (52% vs 26%) and produced a large extra drop in depression scores.[41] But that is a single small, women-only trial. The most recent meta-analysis (11 trials, 1,093 participants) found the pooled effect was small — about 2.2 points on the 17-item Hamilton Depression Rating Scale (HAM-D), below the ~3-point threshold for clinical importance — with high heterogeneity, very low certainty under the GRADE evidence-grading framework, and evidence of publication bias favouring creatine; remission was more common with creatine in the three trials that reported it, but treatment response in two trials was no better than placebo, and the authors caution that the true effect may be trivial or none.[42] Treat creatine as a plausible adjunct to standard treatment, not a stand-alone one, until larger and longer trials confirm it. That includes psychotherapy, not only antidepressants: in an 8-week trial in 100 adults with major depression who were not taking medication, 5 g/day added to cognitive behavioural therapy (CBT) left scores about 5 points lower on the 27-point PHQ-9 depression questionnaire than CBT plus placebo. But 40% of each group dropped out, and the authors call it a hypothesis-generating pilot.[43] People with bipolar disorder should talk to their psychiatrist before using creatine for low mood: in the only randomised trial in bipolar depression (35 patients on their usual medication, 6 g/day for 6 weeks), 2 of the 17 taking creatine switched into hypomania or mania early in the trial.[44]
Emerging: fatigue and post-viral conditions
A newer line of work treats creatine as a bioenergetic support in chronic-fatigue states, and all of it is preliminary. In fibromyalgia, one 16-week placebo-controlled trial raised muscle phosphocreatine and strength but did not ease pain or improve sleep, thinking or quality of life,[45] and a 2026 review of small trials in rheumatic pain conditions likewise found strength gains in fibromyalgia but inconsistent effects on pain.[46] In a six-month trial in just 12 people with post-COVID-19 fatigue syndrome, 4 g/day raised creatine levels in muscle and parts of the brain and eased symptoms such as body aches, headache and poor concentration, but fatigue scores were not significantly better than placebo.[47] The right dose is unsettled. In a four-week single-blind pilot trial in 67 people with long COVID who also exercised three times a week, 6 g/day eased fatigue and improved grip strength compared with placebo, while 18 g/day did neither and came with more nausea (not statistically significant) and small rises in blood creatinine and sodium.[48] A small ME/CFS study without a placebo group used 16 g/day and reported less fatigue.[49]
Cardiometabolic health: the blood-creatine and diabetes link
Population studies have found that people with more creatine in their blood are more likely to develop type 2 diabetes. In a Dutch cohort of 4,735 adults followed for about 7 years, each step up in plasma creatine (one standard deviation) carried about 27% higher risk (hazard ratio 1.27); the link was clear in men and not statistically significant in women.[50] None of these people were taking creatine supplements, and the authors note the finding does not point to harm from supplementation.
A 2024 study from Anna Krook's group at the Karolinska Institutet suggests the link runs the other way.[51] Men with type 2 diabetes had higher plasma creatine and lower muscle phosphocreatine than men with normal glucose tolerance, along with lower levels of mitochondrial creatine kinase 2 (CKMT2), the enzyme that turns creatine into phosphocreatine inside muscle mitochondria. In insulin-resistant mice, neither giving creatine for 2 weeks nor depleting it with a creatine-blocking compound changed glucose tolerance. The authors read raised blood creatine as a marker of insulin resistance rather than a cause, though the human data are a one-time snapshot of 52 men and the cause-and-effect test was short and done in mice.
In the other direction, creatine does not appear to harm glucose control, but whether it improves it is unproven. In the best-known trial, adults with type 2 diabetes took 5 g/day of creatine or placebo during 12 weeks of aerobic and resistance training; among the 25 who were analysed, HbA1c (a marker of average blood sugar) fell from 7.4% to 6.4% with creatine and stayed near 7.5% with placebo, and blood sugar was lower before and in the first hour after a test meal. Insulin levels and insulin sensitivity did not differ; the authors linked the benefit to muscle moving more of its main glucose transporter, GLUT4 (glucose transporter type 4), to the cell surface.[52] A 2022 systematic review of 9 studies in healthy and diabetic people found, in its pooled analysis, no clear effect on fasting glucose or insulin resistance,[53] and reviewers describe the trials as small, short and exploratory, with inconsistent effects on muscle GLUT4 and glycogen.[54]
Translation: there's no sign that creatine causes or worsens type 2 diabetes. A blood-sugar benefit, seen so far mainly in people with diabetes who also exercise, is promising but unconfirmed. Creatine also does not appear to change blood lipids: a 2026 meta-analysis of eight randomised trials found no meaningful change in total cholesterol, LDL ("bad") or HDL ("good") cholesterol, or triglycerides, though the certainty of the evidence is low to very low.[55] See Metabolic flexibility for the broader picture on glucose handling.
Form, dose, and timing
Form: stick with monohydrate
Multiple alternative forms exist — hydrochloride (HCl), nitrate, magnesium chelate, ethyl ester, buffered ("Kre-Alkalyn"), and others. They tend to dissolve better in water and feel less gritty, and the marketing emphasises "better absorption". The actual evidence:
- Monohydrate is the standard: about 80% of randomised trials of oral creatine published from 1994 to 2023 used it (275 of 343).[56] In a small human study, no creatine turned up in the stool of four men taking 2 g doses three times a day, which suggests nearly all of it is absorbed.[57] A rat study by researchers holding patents on a rival form found less complete absorption (16–53%), falling further at higher doses.[58]
- The alternative forms have only a handful of head-to-head trials, often at unequal doses, and none has shown a reliable advantage over monohydrate. Some do worse: creatine ethyl ester partly breaks down into creatinine, and in one trial it failed to raise blood creatine and more than doubled blood creatinine,[59] while a buffered form taken at its 1.5 g/day label dose did not significantly raise muscle creatine.[60] Hydrochloride roughly matched monohydrate in one trial, and nitrate appears to deliver creatine in proportion to its creatine content, with no advantage.[61][62] One small 2026 trial in soccer players (10 per group) did report larger sprint and jump gains with ethyl ester and hydrochloride than with monohydrate, an outlier that has not been replicated.[63]
- Alternatives cost 2–5× more per gram.
There is no evidence-based reason to pay the premium. The International Society of Sports Nutrition position stand recommends monohydrate as the gold standard.[64] Creapure is one German manufacturer's high-purity ingredient brand — a product claim rather than independent verification; see Quality below for what third-party testing covers.
Dose, by purpose
| Goal | Dose | Notes |
|---|---|---|
| Muscle saturation, daily maintenance | 3–5 g/day | Reaches full muscle saturation in ~28 days. No loading needed. |
| Optional fast-loading | 20–25 g/day for 5–7 days, then 3–5 g/day | Saturates in ~1 week. Skip unless you have a specific deadline. |
| Brain creatine | No established dose; trials have used 2–20 g/day | Brain creatine rises far less than muscle creatine (roughly 5–10% vs 20–40%). Higher doses may help, but trials show no clear dose-response, and 4–5 g/day for several months has also raised brain levels.[65] |
| Acute cognitive support during sleep loss | 0.35 g/kg single dose (~25–30 g for a 70–85 kg adult) | One small trial (15 people): brain changes from the first scan about 3 h after the dose, thinking benefits from the first tests about 4 h after (the largest effect was at that first test), still present when testing stopped about 9 h after.[66] A lower 0.2 g/kg dose helped less.[67] For low oxygen, the one human trial used 7 days of creatine, not a single dose.[68] |
Timing
The body's creatine stores are saturated rather than acutely sensitive — timing of the daily dose doesn't appear to matter much. Take it whenever is convenient. Take it with food, carbohydrate, or protein if you want a marginal absorption boost (insulin nudges muscle uptake). There is a mild lean toward post-exercise dosing in some trials, but follow-ups in older adults and athletes found timing immaterial; consistent daily intake matters far more than time-of-day. Caffeine probably doesn't blunt creatine at everyday intakes. In small trials that paired daily caffeine with creatine loading, two found caffeine interfered with creatine's benefit, three found no clash and one found the two worked better together.[69] Expert reviewers think a clash is more likely at about 5 mg of caffeine per kilogram of body weight a day or more (roughly 350 mg for a 70 kg adult), say long-term effects are unknown, and note the pair may upset the stomach more.[70] A morning coffee with your creatine is fine; very high daily caffeine is the less certain case.
Why some people respond less
Response varies between individuals. People who already have high muscle creatine, such as habitual red-meat eaters, gain the least, while vegetarians and others starting low gain the most. In one small biopsy study of 11 young men, the strongest responders also had more fast-twitch (type II) muscle fibres and more muscle mass.[71] One review estimates that about a quarter of people show little response, and suggests that common small variants in the creatine transporter gene SLC6A8 add to this, though that remains a hypothesis.[72]
Quality
Plain monohydrate powder (or capsules) is the most reliable format. In a small, non-peer-reviewed 2025 test run by a supplement company, all five powders held what the label said, while four of six creatine gummies held little or no creatine, and even the gummies that passed contained a lot of creatinine, likely because creatine slowly turns into creatinine when kept moist.[73] For independent third-party testing, look for NSF Certified for Sport or USP Verified; Informed Sport mainly screens for substances banned in sport. Creapure is a manufacturer's ingredient brand, not a third-party certification.
Safety
This is where creatine's track record stands out even among well-tolerated supplements.
A 2025 analysis of 685 clinical trials, with 12,839 people taking creatine and 13,452 taking placebo, covered children and adolescents through adults in their 70s, athletes, military personnel and people with a range of medical conditions. Supplementation lasted about two months on average, often starting with a short high-dose loading phase, and the longest trials ran for 14 years. Overall, side effects were no more common with creatine than with placebo (4.6% vs 4.2% of participants, most of them from one long Parkinson's disease trial), and kidney markers did not differ. Slightly more creatine trials reported stomach upset or muscle cramps, but counted per person neither difference was statistically significant: stomach upset affected about 5.5% on creatine and 4% on placebo, cramps about 0.5% and 0.1%.[74] Adverse-event databases point the same way, with a caveat. Across 28.4 million reports in five national and international databases, only 203 involved a product that actually contained creatine (0.00072%), although most of those reports came from a European drug database with no creatine reports at all; in the US supplement database the share was 0.12%. Such reports cannot show that creatine caused anything. In the US database, almost half of the reports naming creatine were for products with no creatine in them, and most of the rest involved other forms of creatine or creatine taken with other supplements or drugs.[75] The analysis's lead author chairs a scientific advisory board for a creatine manufacturer. People with bipolar disorder should see the caution under Mood and depression above.
The kidney myth
The most persistent misconception: creatine damages kidneys. The mechanism people imagine: serum creatinine (a routine kidney lab) goes up after creatine supplementation. That's true — but it's because creatine metabolises non-enzymatically into creatinine, which is then filtered out by the kidneys. The bump is benign metabolic turnover, not damage.
The functional measure of kidney health is the glomerular filtration rate (GFR). When it is measured directly with a tracer, creatine has not been shown to lower it. But the GFR on a routine blood report is usually estimated from creatinine, and a 2026 meta-analysis of 26 randomised trials, in healthy people and people with kidney disease, found that estimate ran about 11 mL/min lower on creatine on average, most likely because of the extra creatinine rather than kidney injury.[76] If you're getting blood work, mention you're taking creatine so a high creatinine or a low estimated GFR isn't misread. Meta-analyses find no evidence of kidney damage in healthy people, though the authors of one call for trials lasting longer than a year.[77][78]
If you have chronic kidney disease, check with your nephrologist before taking creatine. It pushes up the creatinine-based labs used to track the disease, and its safety in people with kidney disease hasn't been well tested: even a review favourable to creatine concludes that more studies in this group are needed.[79]
What about the GI side effects?
About 5% of people in creatine trials report mild stomach upset, but so do about 4% of those taking placebo.[80] When a trial asks participants to write down anything they notice, the gap can look larger: in a six-week crossover trial of 5 g/day in 123 adults, 17% reported a negative side effect on creatine against 4% on placebo (digestive problems 6% vs 2%, weight gain 3% vs 0%), though no one dropped out because of one.[81] Splitting the dose across the day, taking it with food, or skipping the loading phase usually resolves it. Cramping is rare and not consistently linked to creatine in controlled trials.
The hair-loss myth
The fear traces to a single 2009 study in college rugby players, in which dihydrotestosterone (DHT) rose by 56% during a week of creatine loading and was still 40% above baseline after two more weeks on a lower dose.[82] The levels stayed within the normal clinical range, the finding has never been replicated, and hair was never measured.[83] A dedicated 12-week RCT has now tested the claim directly: 5 g/day in resistance-trained men, with dermatologist-assessed hair plus androgen panels. There was no effect on DHT, the DHT:testosterone ratio, or any hair parameter.[84] The trial was men-only and 12 weeks, but it is strong evidence against the hair-loss claim.
Dehydration, cramping, cycling, and kidney stones
- Dehydration and cramping. A systematic review of high-quality studies found no evidence that creatine impairs heat dissipation or fluid balance, and a dehydration trial found no extra cramping or thermoregulatory problems.[85] Because creatine raises intracellular water, it may if anything aid thermoregulation.
- No need to cycle. There is no physiological basis for "cycling" creatine. Stores stay saturated on 3–5 g/day, and years of continuous use show no transporter down-regulation requiring a washout.
- Kidney stones. Creatine is not an oxalate, calcium, or uric-acid precursor, and there is no direct evidence it causes stones. Concentrated urine is the actual risk factor — so the practical advice is simply to stay well hydrated.
The conditionally essential argument
The contemporary case for thinking of creatine as a near-essential nutrient (rather than an optional ergogenic aid) rests on three observations, each softer than it is often presented:
- Creatine turnover shrinks with age. Older adults break down less creatine each day, mainly because they carry less muscle.[86] Whether the liver and kidneys make enough to offset low dietary intake in old age is unknown.[87]
- Dietary intake is low, in older adults as in everyone else. In U.S. survey data, about 70% of people over 65 ate less than 1 g/day, about 1 in 5 ate none on the day recorded, and the average was about 0.76 g/day.[88] U.S. adults of all ages look much the same, with average intake below 1 g/day across two decades of surveys.[89] Even raw, meat and fish are modest sources, at roughly 2–5 g of creatine per kilogram (plants ~0),[90] and cooking converts part of it to creatinine — so matching a 3–5 g dose would take on the order of 1 kg or more of meat or fish a day, which is impractical and expensive. Older adults may also eat less meat and fish, for example because it is harder to digest.
- Muscle and cognitive reserves shrink with age. Sarcopenia, frailty and cognitive decline all involve cellular energy supply, which is the theoretical case for topping up creatine. Whether supplementation changes any of them outside resistance training is unproven (see below and the brain section above).
Adding creatine to resistance training gives older adults a modest extra gain in strength.[91] Whether that turns into better everyday function is unclear. Earlier meta-analyses found better chair-stand scores and about 1 kg more lean mass,[92][93] but the newest one, with only three trials measuring function, could not confirm an added benefit for rising from a chair, walking or the Timed Up and Go test (standing up, walking a short distance and sitting back down), rated that evidence very low certainty, could not confirm the extra lean mass either, and cautions against applying its results to frail or sarcopenic older adults.[94] A reduction in falls or fractures has not been shown. The critical caveat: whether creatine helps muscle or function in older adults without resistance training is uncertain — an 8-week pilot in adults over 65 at a low dose found no change in grip strength or fat-free mass, but few trials have tested it this way.[95]
Sustained 3–5 g/day supplementation, paired with resistance training, is a cheap, low-risk way for older adults to add a modest amount of strength. The claim that it also protects thinking in older age, through a muscle-brain link, is unproven. At ~$0.05/g, that's a few dollars a month.
For vegetarians and vegans, dietary intake is essentially zero and muscle creatine stores are lower, so supplementation raises muscle stores more than in meat-eaters.[96] Brain creatine, however, appears similar to meat-eaters',[97] and an extra cognitive benefit has not been shown.[98]
What to actually do
For most healthy adults: 3–5 g/day creatine monohydrate, taken whenever convenient, daily. No loading phase needed. Skip the alternative forms.
For people doing resistance training: Same protocol; this is where the muscle evidence is strongest.
For older adults (65+): Same protocol, always alongside resistance training. The extra strength it adds is small, and in one large meta-analysis the gain was statistically clear in adults under 50 but not in those over 50.[99] The case rests on low cost and a good safety record in trials rather than a large effect.
For vegetarians or vegans: Same protocol; you're starting from lower muscle stores, so muscle creatine rises more, though no trial has shown that you will notice more of a difference.
For people hoping for cognitive benefit: There is no established dose. Brain creatine has risen on 4–5 g/day taken for several months as well as on about 20 g/day for weeks, and in well-rested young adults neither 10 nor 20 g/day improved thinking. Taking more than the standard 3–5 g/day for this purpose is not supported by current evidence.
For people facing a night of forced wakefulness (night shifts, emergency work): A 0.35 g/kg bolus (~25–30 g) taken a few hours before the demanding period; in the one trial of this dose, brain changes were seen from the first scan about 3 hours after dosing and thinking benefits from the first tests about 4 hours after, and both were still present when testing stopped about 9 hours after. The evidence is still thin, and this is an occasional use, not a daily one.
If you have bipolar disorder or chronic kidney disease: Talk to your psychiatrist or nephrologist first (see Mood and depression, and Safety).
Before a routine kidney lab: Tell your clinician you take creatine so a raised serum creatinine or a low estimated GFR isn't misinterpreted.
What's overrated and what to know
Overrated:
- Alternative forms (HCl, nitrate, ethyl ester, "buffered"). No reliable advantage over monohydrate, some do worse, and label doses are often too low, at 2–5× the cost.
- The loading phase. It just gets you to the same steady state about 3 weeks faster, which few people need.
- Marketing claims about "rapid hypertrophy" in beginners. Some of the early gain on body scans may be water rather than new muscle.
- Creatine gummies. In one small industry test, four of six held little or no creatine (see Quality).
Worth knowing:
- Brain effects are not clearly dose-sensitive. No threshold dose has been established: brain creatine has risen on 4–5 g/day taken for months, a memory benefit in vegetarians came from 5 g/day,[100] and 10–20 g/day did nothing in well-rested young adults.[101]
- Any cognitive benefit is most likely under acute stress. Small trials hint at effects during sleep deprivation, low oxygen and mental fatigue, but a 2026 review by several of the field's leading creatine researchers still describes creatine's effects on the brain as largely unknown, and the findings in Alzheimer's disease, depression and mild traumatic brain injury (concussion) as preliminary and inconsistent.[102] Well-rested young adults generally notice little.
- Pair with training. Gains in muscle mass, the strongest part of the evidence, depend on structured resistance training, at least in young men.[103] The cognitive findings under stress, such as sleep loss, come from trials without training. Without training the picture is uncertain: one small 12-week trial in 45- to 65-year-olds reported gains in lean mass and strength without exercise, along with some changes in blood sugar and blood fats, but its senior author chairs a creatine manufacturer's advisory board and the authors call the cognitive and biomarker findings exploratory.[104]
Further reading
- International Society of Sports Nutrition position stand on creatine.[105]
- Antonio J et al. Common questions and misconceptions about creatine supplementation: what does the scientific evidence really show? J Int Soc Sports Nutr 2021.[106]
- Prokopidis K et al. Effects of creatine supplementation on memory in healthy individuals: a systematic review and meta-analysis of randomized controlled trials. Nutrition Reviews 2023.[107]
- Gordji-Nejad A et al. Single dose creatine improves cognitive performance and induces changes in cerebral high energy phosphates during sleep deprivation. Sci Rep 2024.[108]
- Rizo-Roca D et al. Decreased mitochondrial creatine kinase 2 impairs skeletal muscle mitochondrial function independently of insulin in type 2 diabetes. Sci Transl Med 2024.[109]
- Candow DG et al. Creatine monohydrate supplementation for older adults and clinical populations. J Int Soc Sports Nutr 2025.[110]
- Desai I et al. The Effect of Creatine Supplementation on Lean Body Mass with and Without Resistance Training. Nutrients 2025.[111]
- Kazeminasab F et al. The effects of creatine supplementation on upper- and lower-body strength and power: a systematic review and meta-analysis. Nutrients 2025.[112]
- Moriarty T et al. Dose-Response of Creatine Supplementation on Cognitive Function in Healthy Young Adults. Brain Sci 2023.[113]
- Kreider RB, Stout JR. Creatine in health and disease. Nutrients 2021.[114]
- Smith-Ryan AE et al. Creatine in women's health: bridging the gap from menstruation through pregnancy to menopause. J Int Soc Sports Nutr 2025.[115]
- Chilibeck PD et al. Effects of creatine and resistance training on bone health in postmenopausal women. Med Sci Sports Exerc 2015.[116]
- Chilibeck PD et al. A 2-yr randomized controlled trial on creatine supplementation during exercise for postmenopausal bone health. Med Sci Sports Exerc 2023.[117]
- Lyoo IK et al. A randomized, double-blind placebo-controlled trial of oral creatine monohydrate augmentation for enhanced response to a SSRI in women with major depressive disorder. Am J Psychiatry 2012.[118]
- Eckert I, Lima J, Dariva AA. Creatine supplementation for treating symptoms of depression: a systematic review and meta-analysis. Br J Nutr 2025.[119]
- Lak M et al. Does creatine cause hair loss? A 12-week randomized controlled trial. J Int Soc Sports Nutr 2025.[120]
- Kieburtz K et al. Effect of creatine monohydrate on clinical progression in patients with Parkinson disease: a randomized clinical trial. JAMA 2015.[121]
- Hersch SM et al. The CREST-E study of creatine for Huntington disease: a randomized controlled trial. Neurology 2017.[122]