Does creatine help memory? The cognition trials, by age

A little for memory, and almost nothing yet in healthy older adults. Across 16 trials, creatine improved memory scores by a small margin and left overall thinking unchanged.¹ The people who gained were the ill, the under-60s and women.¹ In adults aged 55 and over, one 2026 review found only two blinded trials.⁴

The memory result, and how big it is

Creatine helps cells rebuild the molecule they spend for energy. Most of the body's store sits in muscle, and a much smaller amount in the brain. Supplements fill the muscle store readily. Whether they fill the brain store is the question the cognition trials never quite reach.

A 2024 review gathered every trial that gave adults creatine and tested their thinking. It found 16 trials with 492 participants, ranging in age from their twenties to their seventies.¹ Healthy people and patients were mixed together, and every trial used creatine monohydrate, the cheap powder sold in tubs.¹

One domain moved with any conviction. Memory improved by 0.31 on a standardized scale, with a plausible range of 0.18 to 0.44.¹ That scale measures the gap between groups in units of the spread of the test scores themselves, so this is about a third of one standard deviation, which sits at the small end.

Attention and processing speed also improved, with the time taken falling by 0.31 (range 0.03 to 0.58) and 0.51 (range 0.01 to 1.01).¹ Both ranges run almost to no effect, and the reviewers had little confidence in either.¹ Overall cognitive function did not change. Executive function did not change.¹ Course length made no difference either: trials under 4 weeks and trials of 4 weeks or more told the same story.¹

Creatine and thinking: standardized change, with range0 = no effect

Memory0.31 (0.18 to 0.44)

Attention, time taken-0.31 (-0.58 to -0.03)

Processing speed, time taken-0.51 (-1.01 to -0.01)

Who it worked in

The 2024 review found larger effects in people with a disease, in adults aged 18 to 60, and in women.¹

An earlier review looked only at healthy people and found six trials with 281 participants.² Its reading was cautious. Short-term memory and reasoning may improve, while findings for attention, reaction time and mental fatigue pointed in different directions from one trial to the next.² In young participants, nothing moved.²

Creatine in food comes from meat, so vegetarians start with less of it. Nine studies of vegetarians found that supplements raised creatine in muscle, plasma and red blood cells, often past the levels omnivores carry, and raised memory and intelligence scores too.³ The striking part is what stayed put. Phosphocreatine, the brain's stored form of the molecule, did not rise.³

So the mental changes in vegetarians track no measured change in brain chemistry. Nobody can name the dose that matters for the brain, because the one measure of brain stores showed no rise. Those studies also carried a moderate to high risk of bias.³

Older adults: two trials and four surveys

This is where the marketing runs ahead of the evidence. A 2026 review looked for studies of creatine and thinking in adults aged 55 and over, and found six, covering 1,542 people.⁴ Only two were blinded trials in which participants actually swallowed creatine.⁴ The other four asked people what they ate, estimated their creatine intake from the answers, and looked for a link with test scores.⁴

Five of the six reported something positive, usually for memory or attention.⁴ That reads like a winning streak until the quality ratings are set beside it: one study rated good, two fair, three poor.⁴ The reviewers concluded that the evidence is limited and that better trials are still needed.⁴

A separate review of 17 trials that added creatine to exercise programs in older adults found limited data and no added effect on mental health beyond what the exercise did alone.⁷

What older adults do get is physical

Take the brain claims away and creatine still has a case in later life. A review of 33 trials in 1,076 participants used the sit-to-stand test as its main outcome, which is the everyday business of getting out of a chair unaided.⁵ Across the six trials reporting it, creatine improved that by 0.51 standardized, a moderate effect, though the range ran from almost nothing up to large.⁵

The other physical results were smaller. Upper-body strength improved by 0.25 and grip strength by 0.23, and lean tissue rose by 1.08 kg.⁵ Behind every one of those numbers the quality of evidence was rated low or very low, because the underlying trials were at high risk of bias.⁵

The exercise review sets the conditions. Any added gain in strength, function or lean mass in older people required creatine daily at a low dose, alongside at least 12 weeks of resistance training.⁷ A possible effect on thigh bone density may need a year of it.⁷ The gains were more consistent in older women than in men, and the review judged creatine monohydrate safe to use at this age.⁷

The dose, and where the strength result came from

Tub instructions descend from athlete studies run decades ago. A review of dosing in older adults sorted trials by whether the daily amount was above or below 5 g, roughly a level teaspoon of powder, and by whether people front-loaded first with 20 g a day or more for five to seven days.⁶ Front-loading means four of those teaspoons daily for the better part of a week.

Creatine added to the gains from resistance training under every strategy.⁶ Loading followed by a low daily dose beat placebo on chest press strength.⁶ But take the loading trials out of the analysis, and the advantage over placebo vanished for chest press and leg press alike.⁶ The strength signal rested on the loading phase, not on the daily teaspoon that follows it.

What remains unknown

Nobody has run a large, long, blinded trial of creatine for thinking in healthy older adults. The age-specific evidence is two trials and four dietary surveys.⁴ The brain dose is unknown, because the one review that measured brain phosphocreatine saw no change while muscle stores filled.³

Whether the gains in particular groups are real or noise from small trials is open; everything except memory was rated low certainty, and the reviewers asked for larger trials.¹ And the safety record in older adults covers weeks to about a year of use, not decades.⁷