Creatine Monohydrate: Strength, Cognition, Safety and Funding

Creatine monohydrate can support strength gains during resistance training, with varied individual response. Cognition findings are less secure: a positive pooled review was explicitly criticized by EFSA for double-counting related test results. Kidney studies generally show no significant pooled filtration change, but are too small and short to guarantee universal safety. Alternative-form superiority is not established by the comparisons reviewed here. Studied loading and maintenance protocols differ in speed; select dosing and product quality with your clinical and training context. Known commercial relationships and unreported trial finances remain explicit. Wang 2024; Xu 2024; EFSA 2024; Kabiri Naeini 2025; ISSN 2007; NIH ODS.

Table of contents

• Evidence summary
• What creatine monohydrate is
• Selected creatine forms
• How creatine monohydrate works
• What works and what does not
• Benefits and evidence limits
• Risks and reported side effects
• Medicine combinations and precautions
• Who needs clinical advice before use
• Dosage and how to take creatine monohydrate
• Infographics with full text versions
• Related research
• Frequently asked questions
• Sources and funding notes

Evidence summary

What creatine monohydrate is

Creatine is an amino-acid-derived nitrogenous compound made through pathways involving the kidney, liver and other tissues, using glycine, arginine and a methyl group from methionine metabolism. Meat and fish provide dietary creatine. About 95% of the body pool is in skeletal muscle, with free creatine and phosphocreatine fractions. The cited review’s 5 g dose corresponds to roughly 1.1 kg uncooked beef; Food creatine content varies by type and preparation. Kreider, Jäger & Purpura 2022.

Creatine monohydrate contains creatine with one water molecule. Creatine is not an anabolic steroid or stimulant. USADA explicitly says the ingredient is not prohibited; this does not clear every supplement product or jurisdiction. Its creatine/phosphocreatine system helps regenerate ATP, with exercise performance depending on the task and response. Kreider, Jäger & Purpura 2022; USADA ingredient guidance; NIH ODS.

The two-thirds figure in the cited review refers to the approximate phosphocreatine fraction of the muscle pool, with the remainder free creatine. It does not describe the fraction obtained from diet and synthesis. Endogenous synthesis and dietary intake replenish daily turnover. Supplementation can raise muscle stores, with substantial variation in baseline content and response; saturation does not guarantee a particular strength or cognitive gain. Kreider, Jäger & Purpura 2022; ISSN 2007.

Vegetarian and vegan diets generally supply less preformed creatine, and the reviewed studies describe lower muscle stores. The narrative review does not establish lower brain stores or a guaranteed larger response in every clinical outcome. Diet is one relevant factor among several. Nutrients 2024 narrative review; McMorris 2024.

Selected creatine forms

The table describes selected forms discussed in the 2022 comparative review and the original buffered trial. Their known commercial relationships are disclosed. Solubility, blood concentration, muscle loading and clinical outcomes are different endpoints; a null comparison is not an equivalence trial. The sources do not provide a current census of every form, brand, dose or price. Jagim 2012; Kreider, Jäger & Purpura 2022.

How it works

Creatine kinase transfers phosphate between phosphocreatine and ADP to help regenerate ATP. Higher muscle stores can support repeated high-intensity work and training adaptations, but the mechanism does not predict the same response for every exercise, person or brain outcome. Kreider, Jäger & Purpura 2022; NIH ODS.

Other proposed mechanisms are background explanations rather than separate proven clinical benefits:

• Cell water: Creatine can alter intracellular fluid. An anabolic signaling explanation remains distinct from direct proof of additional tissue growth in an individual. Kreider, Jäger & Purpura 2022; Burke 2023; Moore 2023.
• Carbohydrate-related retention: The cited reviews describe increased creatine retention with carbohydrate or carbohydrate/protein. Greater retention does not necessarily produce additional performance benefit. Kreider, Jäger & Purpura 2022; ISSN 2007.
• Brain energetics: The brain uses creatine/phosphocreatine, but uptake and energy-demand hypotheses do not prove general cognitive benefit. EFSA judged the submitted mechanism weak; the acute sleep-deprivation trial concerns a specific high-dose laboratory setting. EFSA 2024; Gordji-Nejad 2024; McMorris 2024.
• Recovery mechanisms: Oxidative, inflammatory and hormonal explanations are discussed in reviews, but the selected women’s recovery trial did not show significant HRV or general sprint-performance benefit. Biomarkers and mechanisms are not interchangeable with recovery outcomes. Nutrients 2024 narrative review; Gordon 2023.

The comparative review describes nearly complete absorption of standard monohydrate, which differs from tissue retention and clinical response. Carbohydrate or carbohydrate/protein co-ingestion can increase retention in reviewed experiments; the historical ISSN statement notes that this need not improve performance beyond monohydrate alone. Kreider, Jäger & Purpura 2022; ISSN 2007.

What works and what does not

Muscle strength, power, and exercise performance

Wang’s 2024 pool included 23 trials, predominantly men under 50, generally 4–12 weeks of resistance training. Additional upper-body strength was 4.43 kg (95% CI 3.12–5.75) and lower-body 11.35 kg (95% CI 8.44–14.25). Direct sex comparisons were nonsignificant—upper p = 0.067, lower p = 0.446—so the small female subgroup does not establish a sex difference. Upper-body publication-bias testing was positive; adjusted estimates still favored creatine. The review had no external funding, but Candow disclosed industry research, donations, travel, advisory and consulting relationships. Wang 2024.

Burke’s 2023 Bayesian review of 10 studies/44 site-specific outcomes estimated a small hypertrophy effect, 0.11 (95% credible interval −0.02–0.25). The age contrast was imprecise and its interval included zero; it is not a confirmed younger-person advantage. Only two studies involved trained participants and one was female-only. No external review funding was reported, with Schoenfeld, Candow and Forbes commercial relationships disclosed; underlying trial finances were not all traced. Burke 2023.

ODS supports creatine’s potential usefulness for intermittent high-intensity activity and strength training, with varied responses. Pure endurance benefit is limited. Some studies describe single-sprint improvements, so isolated efforts cannot categorically be described as having no benefit. A precise universal 30-second threshold or ranking is not established here. NIH ODS.

Cognitive function, sleep deprivation, and mental fatigue

Xu’s 2024 review reports 16 RCTs/492 participants and a pooled memory signal, SMD 0.31(0.18–0.44). Attention time and processing time improved in the reported analyses, but their score outcomes, overall cognition and executive outcomes were nonsignificant. Memory was rated Moderate; processing Low and other domains Low to Very Low. EFSA explicitly found that pooling related tests from the same participants double-counted them and inflated sample sizes, and did not accept this meta-analysis as substantiation of the proposed claim. NSFC grant 72364006 and no commercial relationships were declared; original trial finances and analytic reproduction remain incomplete. Xu 2024; EFSA 2024.

Gordji-Nejad’s 2024 double-blind crossover included 15 healthy young adults under induced sleep deprivation. A single 0.35 g/kg monohydrate dose was compared with corn starch; selected word-memory and processing-time findings favored creatine after multiple-testing adjustment. This does not establish routine 3–5 g/day benefit in well-rested people or an optimal cognition dose. Jülich/RWTH Aachen financed expenses/personnel; Drzezga disclosed industry research, advisory, speaking, stocks and a patent. Alzchem product was named, but purchase/donation terms were unreported. Gordji-Nejad 2024.

McMorris’s 2024 systematic review found mixed cognitive results and questioned a simple brain-energy explanation, while suggesting further research might identify selected benefits. The complete published paper reports no specific grant; a separate author-COI statement was not identified. EFSA’s 2024 opinion assessed an Alzchem-submitted 3 g/day claim, with 10 interpretable healthy intervention studies and 3 patient studies considered separately. It did not establish cause/effect for improved cognition. This opinion is neither proof of zero benefit in every setting nor an operative authorisation decision. McMorris 2024; EFSA 2024.

Older-adult exercise findings and prevention limits

Sharifian’s 2025 review pooled 20 studies/1093 adults aged 55+ across varied healthy and clinical exercise settings. Pooled 1RM improved by 2.122 kg (p = 0.001). Fat percentage fell by 0.548 percentage points (p = 0.026), but this lost significance when one study was omitted; BMD was nonsignificant. Eleven trials were high risk of bias and five had some concerns. No review funding or competing interests were declared, without a complete underlying finance/procurement audit. Sharifian 2025.

These exercise studies support a possible adjunct strength benefit. They do not directly establish prevention of sarcopenia, falls or fractures, and the pooled comparison cannot prove that supplementation alone has no effect in sedentary adults. Sharifian 2025.

Creatine for women: dosing, menstrual cycle, and pregnancy

Women’s evidence differs by endpoint. Small loading studies are not a formal Moderate rating for every female, menstrual or pregnancy claim. Moore 2023; Gordon 2023.

Moore’s 2023 trial enrolled 43 women and analyzed 30 completers, with supplement groups maintained while menstrual phase crossed over. Loading 20 g/day for 5 days did not significantly change body mass; luteal fluid changes favored creatine (TBW p = 0.021, ECF p = 0.013, ICF p = 0.041). Phase verification and attrition limit inference. These measurements do not establish symptomatic bloating, improved thermoregulation or contraceptive safety. Alzchem donated creatine; Smith-Ryan was an Alzchem advisor. Moore 2023.

Gordon’s 2023 study enrolled 52 women and analyzed 39 after dropouts. It found no significant creatine effect on HRV, average power or peak power. A fatigue-index phase-by-supplement interaction (p = 0.048) was exploratory; decomposed phase changes were nonsignificant. Cycle verification and small contraception subgroups limit a dosing recommendation. Alzchem donated product and Smith-Ryan disclosed an advisory role. Gordon 2023.

The 2024 narrative review describes women’s endogenous creatine synthesis rate as roughly 70–80% of men’s, not women’s muscle creatine content. It also describes a higher intramuscular phosphocreatine proportion in women. These biological observations do not prove a universally larger supplementation response; Wang’s direct sex comparisons were nonsignificant. The narrative declares no external funding/no conflicts, without clearing all underlying trials. Nutrients 2024 narrative review; Wang 2024.

Pregnancy findings in the safety review are mainly animal experiments, observational nutrition data and an unusual synthesis-disorder case. They do not establish routine supplementation safety or clinical benefit for healthy pregnant women. Discuss use with the obstetric team rather than extrapolating reassuring general-adult findings. Longobardi 2025.

Loading dose vs. maintenance dose

Historical ISSN guidance describes loading about 0.3 g/kg/day in divided doses for 5–7 days, followed by 3–5 g/day, and gradual loading with 3 g/day for roughly 28 days. These are studied or review-described store-loading protocols; speed of raising stores is different from proof of identical efficacy, safety or optimal dosing for every person. ISSN 2007.

Loading raises stores faster. Larger single doses can increase gastrointestinal symptoms in selected studies; dividing higher daily amounts is a practical option. Skipping loading is also an option, but these sources do not guarantee lower symptoms or identical clinical outcomes for every user. ISSN 2007; Longobardi 2025.

"Non-responders" to creatine

Muscle uptake and performance responses vary. Baseline stores and muscle characteristics are possible contributors discussed in the reviews, but the evidence does not identify one proven primary cause of all nonresponse or guarantee a larger benefit in lower-muscle-mass or vegetarian users. Pooled estimates do not predict an individual result. Kreider, Jäger & Purpura 2022; Burke 2023; Nutrients 2024 narrative review.

Vegetarians and vegans

The narrative review describes lower muscle creatine in vegetarian/vegan groups, while reviewed studies did not show a clear brain-content difference. McMorris likewise found mixed cognition results and heterogeneous diets. More room to increase muscle stores is not proof of greater strength or cognitive benefit in every vegan or vegetarian. Review declarations do not certify complete underlying independence. Nutrients 2024 narrative review; McMorris 2024.

Selected form comparisons and evidence gaps

Jagim’s 2012 buffered-versus-monohydrate trial lasted 28 days, with 36 men completing and 25 in the muscle-content analysis. It found no significant additional training or loading benefit and no fewer reported side effects, without an unsupplemented control or an equivalence design. Alzchem funded the grant and capsules; investigator consulting relationships were disclosed. The 2022 review also has Alzchem APC support and author industry ties. Jagim 2012; Kreider, Jäger & Purpura 2022.

The 2022 review did not establish alternative-form superiority, but its comparisons and original financial chains vary. Ethyl ester had poorer loading in a described protocol; HCl, buffered, micronized and branded nano products cannot be treated as one adequately powered, independently replicated 2026 comparison set. No current all-form superiority or price census was completed. Kreider, Jäger & Purpura 2022.

Kidney findings, laboratory interpretation and clinical limits

Creatine can increase creatinine production, potentially complicating a blood test and creatinine-based eGFR. This does not mean every elevated result is benign or that kidney disease is excluded. Tell the clinician about the product and dose; they can interpret trends and use other measures when appropriate. Kabiri Naeini 2025; Longobardi 2025.

Kabiri Naeini’s 2025 review included 21 heterogeneous studies. It reported 440 participants in the creatinine pool, but the GFR pool comprised only five studies and 143 participants, combining estimated and measured filtration. Creatinine increased in some duration groups, including beyond 12 weeks; it was not uniformly transient. The GFR difference was nonsignificant, which does not prove universal normality or exclude uncommon/long-term harm. The paper’s small reported creatinine-unit estimate was not independently reconciled to original assays here. Review funding and conflicts were declared none; underlying financial chains remain incompletely traced. Kabiri Naeini 2025.

Longobardi’s 2025 narrative provides supporting kidney, safety and pregnancy context with original-study limits. FAPESP/CNPq support coexisted with Alzchem APC support. Gualano disclosed Alzchem grants, products, travel, speaking and advisory work; Roschel reported Alzchem and other company grants/products. Their paper and the Kabiri review declaring no funding are distinct evidence sources, not proof that either complete underlying trial chain is independent. Longobardi 2025; Kabiri Naeini 2025.

Kidney-injury case reports warrant clinical attention but cannot measure incidence or establish causality alone. The narrative describes pre-existing disease or other exposures in several cases; the original 1998 letter was not recovered in full here, so its detailed laboratory trajectory and financing remain unresolved. Existing kidney disease or medicines affecting kidney function require clinical assessment before starting creatine. Longobardi 2025; Original case citation.

Hair outcomes and remaining uncertainty

The 2025 paper describes an earlier 2009 rugby trial with a DHT rise after loading, without actual hair-loss endpoints. A hormone observation is not proof of alopecia. This is the newer paper’s account of the earlier experiment, rather than a separately recovered full 2009 original here. 2025 hair trial.

The 2025 double-blind trial tested 5 g/day for 12 weeks in healthy resistance-trained men 18–40. Thirty-eight completed, with 37 hair assessments; hormone and hair group-by-time findings were nonsignificant. It did not use the earlier loading protocol, measure family history or establish long-term equivalence. The paper declared no specific funding but acknowledged Tose’e Sabz/BPI support; Forbes, Antonio and Tinsley disclosed company advisory, donation, research or paid consulting relationships. These findings do not prove lifelong freedom from hair risk. 2025 hair trial.

If hair changes occur, seek assessment for common causes and discuss the exact supplement with a clinician. The short trial does not establish an individual risk estimate for people with strong family history or prior hair treatment. 2025 hair trial.

GI tolerability and water retention

Gastrointestinal symptoms can occur, particularly with large single doses. The safety review describes one 59-player trial with more diarrhea at 10 g once daily than 5 g twice daily, alongside inconsistent findings in other studies. Splitting larger amounts may help, but the reviewed sources do not prove universal loading-versus-maintenance symptom rates or that food and hydration always prevent symptoms. Longobardi 2025.

Water-related weight changes are different from fat gain or symptoms of bloating. In Moore’s 30 completers, a short loading protocol increased selected luteal fluid measures without significant body-mass change. That result cannot predict every user’s weight, symptoms or phase-specific response. Moore 2023.

The caffeine interaction debate

The safety review describes mixed older caffeine-combination findings and possible gastrointestinal sensitivity. It does not provide a complete financial audit or settle all doses, formulations and performance outcomes. A pre-workout blend adds its other ingredients; neither routine co-use nor an absence of established harm certifies that a combination is safe or effective. Longobardi 2025; NIH ODS.

Benefits by claim

Risks and reported side effects

The reviewed clinical literature contains generally reassuring findings within selected settings. Populations, doses, follow-up, event reporting and finances vary. Small trials and absent reported events cannot exclude rare or long-term harms. Kabiri Naeini 2025; Longobardi 2025; NIH ODS.

These sources do not establish hair loss or dehydration as common effects at studied doses, but they also do not guarantee no kidney, liver or hair harm in every person. The hair study was short and male-only; kidney pools were small. Product purity, clinical conditions and other exposures require separate assessment. Kabiri Naeini 2025; 2025 hair trial; Longobardi 2025; NIH ODS.

Medicine combinations and precautions

The reviewed sources do not provide dedicated interaction evidence for every medicine combination below. These are topics for product-specific clinical review, not verified mechanisms, incidence estimates or a list of combinations proven safe. NIH ODS.

An unestablished interaction is a gap, not proof of safety. Review the exact formulation and medicine list with a clinician or pharmacist, especially when kidney function, pregnancy or other clinical conditions affect suitability. Do not change prescribed treatment based on supplement theories. NIH ODS; Longobardi 2025.

Who needs clinical advice before use

• Existing kidney disease or reduced filtration: obtain treating-clinician advice before use. Selected studies cannot establish suitability for every stage or patient. Kabiri Naeini 2025; Longobardi 2025.
• Medicines and other supplements: review the exact combination with a clinician or pharmacist, particularly when illness or treatment affects kidney function. A specific additive creatine interaction is not established by the reviewed guidance. NIH ODS; Longobardi 2025.
• Pregnancy or breastfeeding: the general-adult evidence should not be treated as safety clearance. Discuss use with the obstetric or pediatric care team; the pregnancy review’s mainly animal/observational findings do not establish routine human supplementation benefit. Longobardi 2025; NIH ODS.
• Children or adolescents: involve an appropriate pediatric or sports-medicine professional. The historical ISSN statement gives supervised adolescent conditions, while ODS cautions that performance-supplement evidence often comes from healthy adults. This is not blanket pediatric safety or efficacy. ISSN 2007; NIH ODS.
• Bipolar disorder or psychiatric treatment: discuss supplementation with the treating psychiatrist. The reviewed cognition studies in patient groups are not psychiatric-treatment recommendations, and do not establish a creatine-specific mania incidence or causal interaction. EFSA 2024; NIH ODS.
• Hair concerns or prior alopecia treatment: the short male trial does not supply an individual risk estimate or long-term clearance. Discuss observed changes with a clinician rather than treating the study as a universal avoid/use rule. 2025 hair trial.

Dosage and how to take creatine monohydrate

These practical notes draw on historical guidance, the comparative review and official ODS context. They are not a complete independently financed trial census or an individualized prescription. ISSN 2007; Kreider, Jäger & Purpura 2022; NIH ODS.

• Regular intake is used in reviewed store-loading protocols. These sources do not establish a universal best time of day or prove every pre/post-workout schedule equivalent. ISSN 2007; NIH ODS.
• Carbohydrate or carbohydrate/protein can enhance retention in reviewed experiments. That does not guarantee extra clinical benefit compared with creatine alone. Kreider, Jäger & Purpura 2022; ISSN 2007.
• Historical maintenance examples commonly use 3–5 g/day after loading. Body size, clinical context and study protocol differ; this is not a verified optimum or universal adequate dose for the vast majority of adults. ISSN 2007; NIH ODS.
• Muscle stores can decline over weeks after stopping. The historical statement discusses cycling protocols but does not establish a universal performance or safety advantage of cycling or continuous use for every setting. ISSN 2007.
• Powder, capsules and blends need appropriate delivered dose, identity and quality checks. The comparative review does not establish current superiority of every exotic delivery format or certify every plain powder. Athlete certification can reduce, not eliminate contamination risk. Kreider, Jäger & Purpura 2022; NIH ODS; USADA ingredient guidance.

Infographics with full text versions

Infographic 1: Selected form comparisons and limits

Infographic 2: Creatinine, filtration and clinical interpretation

Infographic 3: Loading and low-dose store timelines

Related research

For broader background on sleep, mental health, blood pressure, glucose management and cardiovascular health, see these related guides: sleep prevention guide; stress, anxiety, and depression prevention guide; blood pressure prevention guide; diabetes prevention guide; heart disease prevention guide. These links do not establish that creatine prevents those conditions or that every source in those guides is financially independent.

Frequently asked questions

Sources and funding notes

Role of animal and laboratory evidence

This article distinguishes human outcomes, official guidance, reviewed mechanisms and laboratory formulation context. Known commercial support is disclosed and unreported finances remain unknown. The cited evidence does not establish every benefit, safety, medicine or product claim; those limits are retained rather than treating all sources as independent human trials. Kreider, Jäger & Purpura 2022; Longobardi 2025; NIH ODS.