What is being claimed, and by whom
Creatine occupies an unusual position. It is one of the few supplements with a health claim authorised for use on labels, and the authorised wording is narrow: it concerns increased physical performance in successive bursts of short term, high intensity exercise, under stated conditions of use. That claim is about athletic performance in adults doing intense exercise. It says nothing about ageing.
The claim under review here is the broader one now made in longevity contexts: that taking creatine helps older adults hold on to muscle, strength, bone and cognitive function, and therefore extends the years lived in good health. That is a different proposition, it is being made well ahead of the evidence, and it is graded C.
It is worth saying plainly that C is not a dismissal. Creatine is among the better characterised compounds in this field, its mechanism is understood, it is cheap, and its safety record in healthy adults is better than most things graded on this site. The grade reflects the outcomes measured, not the plausibility of the idea.
The mechanism, which is not in doubt
Creatine is a compound synthesised in the body from amino acids and obtained in the diet mainly from meat and fish. Most of it is stored in skeletal muscle as phosphocreatine, which acts as a rapid buffer for regenerating adenosine triphosphate during brief, intense effort. Supplementation raises muscle creatine stores, and vegetarians and vegans, whose dietary intake is lower, typically start from a lower baseline and show a larger increase.
This is settled physiology, and it explains why the performance effect is confined to the kind of work that draws on that energy system. Creatine does not improve endurance performance and does not claim to.
Two secondary observations drive the ageing hypothesis. Ageing muscle loses mass and, disproportionately, power. And the brain also uses phosphocreatine, which is the basis for the cognitive claims. Both are reasonable starting points for research and neither is a result.
What the trials in older adults report
The literature has a recognisable shape. Randomised, placebo controlled trials, typically running from several weeks to several months, usually combining supplementation with a supervised resistance training programme, with lean mass measured by body composition scanning and strength measured with standard tests.
The general pattern reported is that creatine plus resistance training produces somewhat greater gains in lean mass and in some strength measures than training with placebo, and that creatine without training does considerably less. Effects are described as modest, results are not uniform across trials, and the trials differ in loading protocol, maintenance dose, sex distribution, baseline diet and training programme, which makes pooling awkward.
An immediate interpretive problem attends the lean mass finding. Creatine draws water into muscle cells. Some of the measured increase in lean mass, particularly early, is intracellular water rather than contractile protein. Body composition scanning does not distinguish the two. This does not mean the finding is spurious, since cell swelling has been proposed as a signal for protein synthesis, but it does mean that a lean mass number in a creatine trial is a weaker outcome than the same number in a training-only trial.
On bone, the position is thinner. Trials have measured bone mineral density and geometry alongside training, and results have not established a reliable effect. No trial has reported fractures as a primary endpoint.
On cognition, the evidence is early. Trials report scores on cognitive test batteries, effects appear more often in conditions of stress such as sleep deprivation or in people with low dietary intake, and findings in healthy older adults are inconsistent. Cognitive test scores are a surrogate for the outcome anyone cares about, which is whether someone develops dementia or loses independence, and no trial has approached that.
The limitations that constrain what this can tell you
| Limitation | What it affects |
|---|---|
| No clinical endpoints | Nothing has been reported on fractures, disability, dementia diagnosis or death, which is where a healthspan claim would have to be settled. |
| Trial duration | Weeks to months against a process measured in decades. Durability of any gain after supplementation stops is largely unstudied. |
| Water in the lean mass signal | Intracellular water inflates measured lean mass, and scanning cannot separate it from contractile tissue. |
| Heterogeneity | Loading protocols, maintenance doses, sexes, baseline meat intake and training programmes all differ between trials. |
| Effect requires training | Most of the reported benefit appears alongside resistance training, so the supplement is at best an amplifier of something that has its own evidence base. |
| Product quality | Creatine is sold as a food supplement, so composition and contamination are the manufacturer's responsibility and are not verified before sale. |
Safety, and the one laboratory result that causes confusion
In healthy adults at commonly used doses, creatine monohydrate has an unusually good tolerability record for a supplement, with gastrointestinal upset and weight gain from water retention the usual complaints. The persistent claims of kidney damage in healthy people have not been borne out.
The practical point that matters clinically is different. Creatine supplementation raises serum creatinine, the blood marker used to estimate kidney function, because creatinine is its breakdown product. A raised creatinine in someone taking creatine can be misread as impaired kidney function, prompting investigation that was not needed, or alternatively can obscure genuine impairment. Anyone taking it should tell whoever is ordering the blood test. That is a concrete, checkable statement and it is more useful than most of what is written about this compound.
People with existing kidney disease, and anyone taking medicines that affect renal function, are a different case and that is a conversation with a doctor rather than a supplement decision.
What would change the grade
A move to B requires a randomised trial in older adults, adequately powered, running for years rather than weeks, with a pre-registered primary endpoint that is clinical or robustly functional: fractures, falls resulting in injury, incident disability, or loss of independent living. Trials in fracture and fall prevention are the most plausible vehicle, because the infrastructure for them already exists.
Nothing about the current evidence would move the grade down. It would simply stay at C indefinitely if the field continues to run short trials with intermediate outcomes, which is the most likely outcome given who funds them. Our explainer on surrogate endpoints covers why that ceiling exists.