Creatine benefits, ranked by how strong the evidence is

Strength and power: the best-supported claim

A 2025 meta-analysis of 69 randomized trials found that creatine combined with resistance training improved bench press strength, squat strength and jump and sprint power compared with placebo. The effects were small but statistically significant. Handgrip and leg press strength did not change significantly overall.

The gains were consistent in younger adults and in males, and smaller in older adults and females on several measures. The authors called for more trials in women and older adults to clarify how those groups respond.

Evidence strength: moderate to strong for the direction of the effect, given the number of trials. The size is modest, and it depends on training. Nothing here suggests creatine builds strength on its own. For background on the compound itself, see our creatine topic page.

Lean mass, water weight and fat loss

A 2022 meta-analysis of 35 trials found creatine raised lean body mass only when combined with resistance training, not with other exercise or no exercise. Men gained about 1.46 kg of lean mass. The 0.29 kg increase in women was not statistically significant.

Early weight gain is mostly water. In a one-week trial of 27 adults using a loading dose of 0.3 g/kg, total body water rose about 2% and fat-free mass readings rose 1.1 to 1.9 kg across three methods. That is too short a window for new muscle, and it is not fat.

Fat loss is the weakest version of the claim. In adults under 50, two meta-analyses of 12 trials each found creatine plus resistance training was associated with a body fat percentage difference of 0.88 and 1.19 points. Only one found a significant difference in absolute fat mass (0.73 kg less); the other found 0.18 kg, indistinguishable from chance.

In adults 50 and older, a 19-study meta-analysis found a 0.55-point difference in body fat percentage, while the roughly half-kilogram fat mass difference was not significant. A 143-study dose-response analysis found 0.86 kg more total body mass and only a 0.28-point body fat difference. Evidence strength: consistent for lean mass with training, weak and marginal for fat.

Older adults, women and bone

Older adults fare better than the subgroup results above might suggest. A meta-analysis of 22 trials (mean ages 57 to 70, lasting 7 to 52 weeks) found creatine plus resistance training increased lean tissue mass and both chest and leg press strength more than training alone.

Older women are a murkier case. A meta-analysis of 10 trials found upper-body strength improved, but lower-body strength and muscle mass did not, unless the training program lasted at least 24 weeks. A 2026 meta-analysis of 7 trials in postmenopausal women found small gains in lean mass (0.37 kg) and leg press strength (7.5 kg), mainly at 5 g/day or more alongside training.

Bone is where the claim weakens. A 12-month trial in postmenopausal women found creatine plus training slowed bone loss at the femoral neck (-1.2% versus -3.9% with placebo), but not at other sites. A 2-year trial of 3 g/day and a 1-year trial of 1 g/day, both without training, found no effect on bone density, lean mass or muscle function. A meta-analysis of 5 trials found no significant bone density difference at any site.

Perimenopausal women are barely represented, and no trial has tested creatine against hot flashes or mood in menopause. Evidence strength: moderate for muscle with training, weak to negative for bone.

Memory, focus and sleep deprivation

A 2024 meta-analysis of 16 trials found creatine improved memory and processing speed, with moderate-to-low certainty. It found no significant improvement in overall cognitive function or executive function.

The clearest effects come from sleep deprivation. A 2006 trial of 19 people found a 7-day loading protocol reduced the decline in mood and several cognitive and psychomotor tasks after 24 hours awake. A 2024 trial found a single 0.35 g/kg dose (roughly 24 g for a 150-lb adult) improved processing speed over 21 hours awake. A 2026 follow-up using 0.2 g/kg (about 14 g) reduced cognitive decline by up to 12%, with women benefiting more than men.

Those doses are far above a daily serving, and none of these trials measured sleep itself. In well-rested young adults, 6 weeks of about 2.1 g/day produced no improvement on any of seven cognitive tests. Evidence strength: low to moderate, and specific to stress conditions.

Depression, bipolar disorder, anxiety and ADHD

Depression has the most data and the least convincing summary. A 2025 meta-analysis of 11 trials (1,093 people) found an average effect of about 2.2 points on the Hamilton scale, below the 3.0-point threshold usually considered clinically meaningful. Certainty was very low, and the authors found signs of bias inflating the result.

The best individual trial gave 52 women on an SSRI either 5 g/day of creatine or placebo. Scores fell faster with creatine from week 2 through week 8. A smaller 2013 pilot in 18 treatment-resistant patients, using 5 or 10 g/day, found no overall advantage. A 2022 psychiatric taskforce rated creatine monotherapy for unipolar depression as not currently recommended.

The bipolar trial had 35 participants. Its cognitive sub-study found better verbal fluency with creatine and no difference on other tests, and a systematic review called the bipolar findings variable and mostly non-significant. The anxiety claim traces to one sentence of "promise" in a 2023 narrative review that cited no anxiety trial, and several of its authors disclosed ties to creatine makers.

ADHD-like behavior does appear in rare inherited creatine deficiency disorders, which are treated under a metabolic specialist with prescription-level dosing. That is a different situation from typical ADHD; our piece on creatine deficiency as a diagnosis explains the distinction.

Type 2 diabetes and blood sugar

A 12-week trial in 25 people with type 2 diabetes gave 5 g/day of creatine or placebo, with everyone in the same exercise program. HbA1c fell from 7.4% to 6.4% with creatine and stayed at about 7.5% to 7.6% with placebo, a 1.1-point difference.

That is a striking number from a small trial. A systematic review found only 3 trials worldwide (87 people), all at unclear-to-high risk of bias, and rated the evidence low-certainty. Other reviews call the findings preliminary and still unconfirmed. A 2025 review found no consistent kidney or liver marker changes in these studies, but said larger, longer trials are still needed.

Evidence strength: low. No trial tested creatine without exercise, and none supports treating it as a substitute for prescribed diabetes medication.

Dementia, cancer, heart and liver

Dementia. A 2026 scoping review found no controlled human trial of creatine in dementia. The one human study was an uncontrolled pilot in 20 people with Alzheimer's disease: 20 g/day for 8 weeks raised brain creatine 11% and some test scores rose, but with no placebo group, practice effects cannot be ruled out. Animal results conflict, including one rodent study in which creatine worsened memory in males while helping females.

Cancer. No human trial has measured cancer risk or outcomes. Older mouse studies with skin-implanted tumors found suppressed growth, while newer studies with tumors in their natural organs found more spread of some cancers. A 2026 study linked creatine to platelet activation and metastasis in mice, and found healthier-looking human platelets became more reactive, though no cancer outcome was measured. One mouse study found less muscle wasting with creatine. The picture is unsettled, and one review urges caution for people in cancer treatment.

Heart. Two small trials in heart failure patients (10 days and 6 weeks, 20 g/day) found improved muscle energy stores or strength without changes in ejection fraction, and one found no change in exercise capacity or quality of life. A review called the heart failure evidence small, mixed and inconclusive. A single case report ties an irregular heartbeat to creatine use, which cannot establish causation. Blood pressure was not a main outcome in either small older-adult trial that touched on it, so there is no answer there.

Liver. A small 8-week trial in football players found no significant change in liver or kidney markers. Mouse data split: creatine was linked to less liver fat on a high-fat diet and more damage with alcohol. A review found human research in chronic liver disease essentially absent.

Safety claims: kidneys, cramps and stomach

A review of 19 studies covering 5 to 30 g/day for 5 days to 5 years found no known harm to kidney function in healthy people. It advised against use in people who already have kidney disease. The 2017 International Society of Sports Nutrition position stand likewise concluded creatine is safe and well tolerated across the doses and durations studied.

The cramping myth is not supported. Observational data from college football players found less cramping and heat illness among users, though players chose for themselves, so that does not prove protection. A randomized crossover trial found a week of high-dose creatine did not worsen hydration or heat tolerance. A 2025 analysis of 685 trials found no overall increase in side effects, and a small cramping difference vanished when weighted by sample size. Our piece on creatine and muscle cramps covers this in more detail.

Stomach upset depends on how it is taken. In a 28-day soccer trial, a single 10 g dose caused diarrhea about twice as often as the same total split into two 5 g doses, or placebo. Safety in children and adolescents remains an open question in sports nutrition, and nobody has trialed creatine in pregnancy.