Creatine and Aging: Does the Internet's Favourite Powder Work After 50?

How a gym-bro powder ended up in the longevity aisle

For about thirty years creatine had one job and one demographic: help large men move heavy objects slightly more times. Then, somewhere around 2024, it started turning up in menopause podcasts, in "what I take at 62" videos, and in the kind of supplement stack that also contains three things with the word "cellular" on the label.

This is normally the point at which we roll our eyes and write a hype detox piece. Not this time. Creatine is the rare case where the supplement got popular after the evidence had already accumulated, rather than instead of it, and the trials in older adults have been quietly stacking up for two decades.

But "the evidence is good" and "the claims on your feed are good" are different statements. The gap between them is where this article lives, because the trial data contain one variable that decides whether creatine does anything for you at all — and it is not the brand.

What creatine actually does, in one paragraph

Creatine (C₄H₉N₃O₂; sold as the monohydrate, C₄H₉N₃O₂·H₂O) is not a hormone, a stimulant or a signalling molecule with a fashionable pathway attached. It is a battery. Your body makes about a gram a day from glycine, arginine and methionine, and you eat roughly another gram if you eat meat. Inside muscle and brain cells it gets phosphorylated into phosphocreatine (C₄H₁₀N₃O₅P), which exists to hand its phosphate group straight back to ADP to regenerate ATP in the first seconds of hard work — faster than glycolysis can get out of bed. Supplementing simply raises the size of that buffer.

That mechanism explains the pattern in the data almost perfectly. Creatine improves things that are limited by rapid ATP resupply: a few extra reps, a bit more work per set, and — in tissue with high energy demand and declining supply — possibly some cognitive tasks. It does not do anything to the senescent cells, the epigenetic clocks or the other hallmarks of aging that the longevity field usually shops for. Nobody has shown it extends lifespan in humans, and nobody has run that trial.

The muscle numbers: real, replicated, and smaller than your feed says

Start with the December 2025 systematic review in the European Review of Aging and Physical Activity. Eight randomised controlled trials, 482 older participants, all doing resistance training two to three times a week for anywhere from 8 to 104 weeks, half of them also swallowing creatine.^[1]

Creatine plus training beat placebo plus training for lower-limb strength (SMD 0.29) and lean tissue mass (SMD 0.27). Upper-body strength: no significant difference. And the leg-strength result carried a p-value of exactly 0.05, with a confidence interval whose lower bound was 0.00 — which is statistics for "we found it, but do not build a religion on it."

A standardised mean difference of 0.27 is conventionally a small effect. In plain terms: over a training block, the creatine group tends to end up a bit ahead of the placebo group, in a way you would detect with a DXA scanner and probably not in a mirror. The same review found the benefit was clearer in interventions of 32 weeks or less, which is a slightly awkward finding — it may mean the placebo group's own training gains eventually catch up, which would be a very sensible thing for them to do.

The honest counterweight: what happens when you measure properly

Lean body mass from a DXA scan is a generous measure for creatine, because creatine pulls water into muscle cells. That first 1–2 kg people gain in the first fortnight is real, useful intracellular water, and it is not contractile tissue.

So the more interesting analysis is the 2023 review that only accepted direct imaging of muscle — MRI, CT or ultrasound — across 44 outcomes from 10 trials. Pooled effect: 0.11, described by the authors as "a very small effect favouring creatine".^[5] Regional muscle thickness gains came out at 0.10–0.16 cm. And the moderator analysis pointed the wrong way for our purposes: the benefit was somewhat larger in younger adults than in older ones.

Both things are true at once. Creatine reliably adds a small amount on top of training. It is not a growth agent, and if you are 55 and not lifting, the ceiling on what it can do for your body composition is approximately the floor.

Where the case is strongest: postmenopausal women

The best paper of 2026 on this is a meta-analysis published in May in the Journal of the International Society of Sports Nutrition: seven randomised placebo-controlled trials, 608 women averaging about 62 years old, running from 12 to 104 weeks with a median of 38.^[2]

And then the sentence that should really be the headline: benefits were evident when creatine at 5 g/day or more was combined with resistance training, while trials using 3 g/day or less without training showed no measurable effect.

That is the whole article in one line. Creatine is not a supplement that does something to you. It is a supplement that lets training do slightly more to you. Take it while sedentary and you have bought a very affordable way to retain water. This is the same lesson as most of the free levers, arriving in powdered form.

The brain claims, where the hype outruns the data

Creatine's second act is cognitive, and it is the part of the story most likely to be oversold to you this year.

The pro case is decent. A 2023 meta-analysis in Nutrition Reviews pooled randomised trials of creatine on memory in healthy people and found a benefit of SMD 0.29 — with a striking age split. In adults aged 66 to 76 the effect was 0.88; in people aged 11 to 31 it was 0.03.^[3] Dose, duration, sex and country did not move the needle. Age did.

Now the cold water. That older-adult subgroup carried I² = 83%, meaning the trials disagreed with each other a great deal, and it rested on a small number of studies. And the larger follow-up — a 2024 review of 16 randomised trials in 492 adults — found the benefit for memory held up at SMD 0.31 with moderate GRADE certainty, but reported no significant improvement in overall cognitive function or executive function, and rated attention and processing-speed findings as low certainty.^[4] Benefits also skewed toward people with an underlying disease.

Read together, these say something quite specific: creatine appears to help a brain that is short of energy do one particular job — hold and retrieve information — and does not measurably improve thinking in general. That is an interesting, plausible, mechanistically coherent finding. It is not "creatine protects against dementia", which is a claim nobody has earned. If cognitive aging is your actual worry, the evidence for what SuperAgers have in common remains a better guide than any powder.

Dose, form, and the three things people get wrong

Form. Creatine monohydrate was the form used in every single trial in the cognition meta-analysis, and it dominates the muscle literature too.^[4] The HCl, buffered and "advanced" versions are not better-evidenced; they are better-marketed. The rule from reading the label applies: when a supplement renames itself, check whether the trials came along for the ride.

Dose. 5 g/day. The loading protocol (20 g/day for a week) saturates muscle faster but ends up in the same place by roughly week four or five, so it is optional and mainly buys you the gastrointestinal experience sooner.

Timing. Largely irrelevant. Saturation depends on cumulative intake, not on hitting a window. Take it with whichever meal you will not forget.

One genuine nuance: people with low baseline muscle creatine — typically vegetarians and vegans, who get almost none from food — tend to have the most room to move, and the position-stand literature notes that habitual low intake of around 3 g/day may itself carry health relevance across the lifespan.^[6] If you eat steak four times a week, you are already partway saturated and should expect less.

Safety, and the blood-test trap

The kidney fear attached to creatine has been remarkably durable and remains unsupported in healthy people. The International Society of Sports Nutrition position stand — the standard reference on this — concluded that supplementation at up to 30 g/day for as long as five years is safe and well tolerated across populations from infants to the elderly.^[6] The 2026 postmenopausal meta-analysis, which specifically looked, found renal indices unchanged and adverse events indistinguishable from placebo.^[2]

There is, however, a real trap, and it is a measurement one. Creatine breaks down to creatinine — which is exactly what labs measure to estimate kidney function. Supplementing can nudge serum creatinine up and therefore make your calculated eGFR look worse without anything having happened to your kidneys. Every year this sends people into an entirely avoidable panic. If you take creatine, say so before a blood draw; cystatin C is the usual way to check properly. And if you already have kidney disease, this is a conversation with a clinician, not with a supplement label.

So should you take it?

Here is our read, which is deliberately narrower than the internet's.

The cost argument deserves its own sentence. Five grams a day of plain monohydrate runs to roughly fifteen cents. Compare that with the shelf around it: creatine is arguably better evidenced for muscle in older adults than NMN is for anything, and it costs a small fraction of urolithin A. On the criterion we use across the short list — evidence per euro — creatine is close to the top of the sheet.

What it is not is a longevity drug. Nothing in this literature touches lifespan, biological age or the hallmarks of aging. What creatine plausibly does is help you keep slightly more strength and muscle into your seventies, which matters enormously for how long you stay independent — but it only does that on top of the thing that was already doing the work. Which is, as ever, and rather annoyingly, lifting something heavy twice a week. If you are also wondering how much protein belongs alongside it, that argument had its own strange year.