Beyond the gym. Into a more difficult question.

A plain tub of white powder has acquired an unexpectedly ambitious job description. Once associated chiefly with the weights room, creatine now appears in conversations about clear thinking, exhausted brains and Alzheimer’s disease. Its familiarity is part of the appeal. Here is an ingredient that has already spent decades being studied, usually costs less than elaborate supplement blends, and has a biological explanation that sounds satisfyingly direct: cells need energy.

That explanation opens a worthwhile scientific question. It does not settle it. The most interesting creatine story today is the distance between two bodies of evidence: a substantial exercise literature and a much smaller, less consistent cognitive literature. A reader can take the second seriously without borrowing certainty from the first. Indeed, that separation is what makes the emerging research easier to understand.

The energy reserve behind the reputation

Creatine participates in the recycling of adenosine triphosphate, or ATP, the molecule cells use to power work. In muscle, stored phosphocreatine helps regenerate ATP during short, demanding efforts. The body makes creatine, and foods such as meat and fish supply some too. A supplement changes the available supply; it does not introduce an entirely foreign energy system. The US National Institutes of Health exercise fact sheet explains why that matters most for repeated, high-intensity activity.

For a gym-goer, the useful outcome is not simply a higher creatine measurement. It is whether a programme becomes more productive: another quality repetition, improved strength over time, or a better capacity to repeat brief efforts. Creatine is an addition to training, food and recovery. It cannot make an inconsistent programme consistent, and an impressive label cannot decide what physical task a person should train for.

The Australian Institute of Sport supplement framework includes creatine among ingredients with evidence for particular performance applications. That wording matters. Recognition within a sports framework is not approval for every claim attached to a product. A training aid can be useful within its tested setting while a memory, longevity or disease-treatment claim remains unproven.

Why strength matters beyond appearance

Research in older adults makes the muscle story more consequential than a before-and-after photograph. A controlled study of vegetarians and nonvegetarians undertaking eight weeks of weight training examined muscle creatine, body composition and performance, illustrating how starting conditions and an actual training stimulus belong in the experiment. It is not enough to ask whether an ingredient has an effect somewhere; the question is whether it adds value to the activity and population being considered.

A 2017 synthesis of resistance-training trials in older adults found greater gains in lean tissue and strength with creatine than with training alone. Lean tissue is not synonymous with newly built contractile muscle, because body-composition methods also reflect water. Strength measurements provide a different piece of the picture. Neither endpoint, by itself, proves fewer falls, less disability or longer life.

Our editorial interpretation is that the most persuasive practical case is therefore specific: a person undertaking appropriate resistance training may have a reason to discuss creatine for that purpose. The appeal does not require an additional promise about the brain. Giving an established use a realistic boundary strengthens it; piling on speculative benefits makes the whole proposition harder to assess.

A brain hypothesis meets a demanding test

The brain also relies on energy metabolism. That makes creatine biologically interesting, but a plausible mechanism is only the beginning of an intervention study. A supplement must reach the relevant tissue, alter something meaningful, and produce a benefit that can be distinguished from ordinary variation, practice and expectation. Brain chemistry and daily cognitive function are related questions, rather than interchangeable measurements.

One particularly useful corrective to sweeping claims is a 2023 randomised, double-blind crossover trial involving 123 participants. Participants received creatine and placebo during separate six-week periods. Roughly half were vegetarian. The principal tests examined working memory and abstract reasoning. The working-memory result did not reach the conventional significance threshold, and the reasoning result was not significant. Eight exploratory tasks showed no indication of improvement. Vegetarians did not benefit more than omnivores, and side effects were reported more frequently with creatine.

The investigators’ analyses left room for a small possible benefit, rather than supporting a dramatic enhancement. That is a scientifically useful outcome. It narrows the scale of what remains plausible in that setting. A product advertisement might prefer the simple word “improves”; a reader benefits more from knowing which task, which participants and how uncertain the result was.

What happens when the brain is tired?

A different experiment asks whether creatine matters more under metabolic strain. A 2024 crossover study during sleep deprivation measured cognitive performance alongside brain energy-related signals after a single large research dose. The small study reported improvements in some performance measures compared with placebo. It provides a reason to investigate the interaction between energy availability and fatigue, not a routine for replacing sleep.

The limitation is built into the question. A result obtained during a controlled night without sleep cannot automatically describe an ordinarily rested adult, a shift worker over several years, or someone with a neurological condition. The research dose should not become an internet challenge. An experiment can reveal a mechanism while leaving the everyday clinical decision unresolved.

A 2026 sleep-deprivation trial also reported less deterioration on selected cognitive tasks following creatine. It adds to that specific line of inquiry. The outcome remains performance under an experimentally imposed stressor, rather than proof that regular use prevents memory loss. Future replication needs to establish who benefits, the magnitude of that benefit, tolerability and whether laboratory advantages translate into meaningful functioning.

The Alzheimer’s pilot: a beginning, not a prescription

The most emotionally powerful headlines concern dementia. A 2025 pilot involving 20 people with Alzheimer’s disease investigated eight weeks of creatine supplementation. Researchers found that supplementation was feasible, brain creatine increased, and several cognitive measures improved. Those findings justify further investigation, particularly because measuring brain exposure helps connect an intervention with the biological hypothesis behind it.

However, the pilot had no placebo group. Without that comparison, repeated testing, attention from a research team, expectation and changes unrelated to treatment remain possible explanations for improvement. Eight weeks also cannot establish a change in the long-term course of Alzheimer’s disease. An encouraging signal and a demonstrated disease-modifying treatment occupy very different places in medical evidence.

For families, the distinction can feel frustrating. It is also protective. The next valuable study is one that can challenge the promising explanation fairly: random allocation, a suitable control, enough participants and follow-up that includes everyday function. Brain creatine measurements matter, but so do the outcomes people live with—communication, independence and the capacity to manage familiar tasks.

Read the label, then name the goal

Creatine monohydrate is the form most extensively studied for exercise. More expensive formats and proprietary blends do not inherit superiority from a persuasive product story. A useful comparison begins with the actual ingredient, the amount supplied, product testing and the purpose for taking it. It should also include what a product leaves unclear: a blend may advertise several ingredients while making each contribution difficult to evaluate.

Sport Integrity Australia advises attention to batch testing because supplement contamination can create risks, particularly for tested athletes. Batch testing reduces a specific risk; it cannot certify that a cognitive promise is true or that a product suits every medical situation. Quality control and efficacy are separate checks, and a premium-looking container answers neither.

A practical conversation with a qualified practitioner should establish whether the intended outcome is training performance, fatigue, a diagnosed condition or an unexplained symptom. These goals require different evidence. A person taking creatine for resistance training does not need to frame every ordinary lapse of concentration as proof of a second benefit. Equally, an unresolved memory concern deserves investigation even if training is going well.

The safety conversation has useful detail

The NIH fact sheet describes water-related weight gain and gastrointestinal complaints among relevant practical considerations. A changing scale reading is therefore not a clean measure of either fat gain or new muscle. Recording the reason for use, the product and any symptoms is more informative than chasing a daily weight response.

Kidney disease, abnormal kidney results, pregnancy, medicines and complex medical histories warrant an individual discussion rather than assumptions drawn from healthy exercise participants. Tell the clinician interpreting blood tests about supplementation: creatinine results require context. The point is not that every user needs an elaborate monitoring programme; it is that evidence from one population does not remove the need to understand another person’s circumstances.

There is no established universal creatine dose for cognition or dementia prevention. We have deliberately not turned high-dose research protocols into a recommendation. If a claim relies on a supervised experiment, the supervision and eligibility criteria are part of the evidence, not details that can be discarded when translating it into a home routine.

One useful question for an appointment is what would count as a worthwhile result. For training, that might be a documented change in a relevant performance measure over an agreed period. For cognitive symptoms, it first means understanding the cause. Decide what will be reviewed and when, rather than continually adding ingredients whenever a result is difficult to interpret. This turns an open-ended purchase into a more accountable decision.

What would make the next headline stronger?

We will be looking for larger controlled trials with predefined cognitive outcomes, independent replication and clear adverse-event reporting. Studies should distinguish rested adults from sleep-deprived participants, healthy ageing from diagnosed disease, and improved test performance from sustained functional benefit. They should also explain the clinical relevance of a difference, rather than relying on whether a statistical threshold was crossed.

For now, creatine is compelling partly because it does not need mythology. Its training applications have a substantial foundation. Its cognitive applications pose credible, unfinished questions. The responsible next chapter will emerge from better comparisons, not a louder claim that one inexpensive ingredient has solved the complicated business of keeping a mind well.