Nutrition 02: Creatine

Short answer. Creatine monohydrate alongside resistance training increases strength: roughly four and a half kilograms on an upper-body max in pooled trials of adults under 50, almost all of them men. Take it daily and do not fuss over timing. In older adults, dose and loading made no difference to lean mass, although the strength gains depended on a loading phase or a higher dose.

---

What it is

Creatine occupies an unusual position. Almost everything else in the supplement aisle has a weaker evidence base than its marketing implies. Creatine has a stronger one.

That inverts what is worth writing about. The question is not whether it works. It is which of the many arguments about how to take it are load-bearing, and the answer is fewer than you would guess from the amount of argument.

The proposed mechanism

Energy buffering. Phosphocreatine regenerates ATP, the cell's immediate energy currency, during short, high-intensity efforts. More stored phosphocreatine means more work available before that system is depleted, which allows more repetitions or more sets at a given load. On this account creatine does not build muscle. It lets you do more of the training that builds muscle.

Cell swelling. Creatine draws water into muscle cells by osmosis, and cell swelling has been proposed to activate protein synthesis directly.

These two accounts have very different implications for how you should read your own results. The second one means some of the size you gain in the first weeks is water inside the muscle rather than new contractile tissue, which is the same measurement problem that appears at the start of any training program. (→ Inflammation 02: Delayed-onset muscle soreness)

What has actually been tested

It works, and the effect on strength is modest in absolute terms. A meta-analysis of 23 studies in adults under 50 found that creatine combined with resistance training significantly increased upper-body strength against placebo, with a weighted mean difference (the average gap between groups, weighted by study size) of 4.43 kg (Wang et al., 2024).

The participant base is heavily male. In that analysis, 20 of the 23 studies involved males, contributing 447 male participants against 40 female participants across two studies and one mixed study. This is not a minor caveat about generalizability. It is most of the evidence.

The dosing arguments matter less than they sound, but not for everything. A meta-analysis in older adults compared lower doses at or below 5 g per day against higher doses above 5 g, with and without a loading phase of 20 g or more for five to seven days, and also examined supplementing only on training days. Creatine augmented lean tissue mass whichever strategy was used. Strength did not behave the same way: in the subanalyses, strength gains depended on a loading phase or a higher dose (Forbes et al., 2021).

So in that population the loading phase and the exact daily dose did not determine the lean mass outcome, and did matter for strength.

Timing appears not to matter either, or has not been shown to. A trial comparing creatine taken before against after training across eight weeks in collegiate athletes was designed to test exactly this (Dinan et al., 2022), and reviews of this question describe the evidence as inconclusive.

The effect is age and sex dependent. Reviews consistently report the clearest benefits in younger males, with effects diminishing in older adults and in women, which sits awkwardly beside the fact that the older-adult literature is where the dosing questions were answered.

Training status has been tested, without a clear answer. A dose-response meta-analysis examined whether prior training experience changes the effect of creatine on body composition. Experienced lifters came out about 0.6 kg ahead of novices, a difference that was not statistically significant (Ashtary-Larky et al., 2025).

Our reading

This is the supplement in this publication with the clearest yes.

The effect is real, it has been replicated across many trials and several independent research groups, and monohydrate is the form that has been tested. Nothing in this article should be read as hedging on that.

What we would push back on is the elaborate protocol culture around it. The dosing comparisons found lean mass insensitive to strategy, although strength gains in older adults depended on loading or a higher dose. Loading gets you to saturation faster, and for lean mass it does not appear to change where you end up. Timing has not been shown to matter. The forms marketed as superior to monohydrate are, at best, less studied.

The effect size deserves stating plainly. Roughly four and a half kilograms on an upper-body one-rep max, pooled across studies, over training blocks of a few months. That is worth having and it is not transformative, and the gap between that number and how creatine is discussed is the usual gap this publication exists to point at.

And some of the early gain is water. If your bodyweight jumps two kilograms in the first two weeks, that is largely intracellular fluid rather than contractile tissue. It is not fake, because cell swelling may itself contribute to the signal, but it is not what the scale appears to be telling you.

So: take monohydrate, take it daily, and stop optimizing the rest of it. A loading phase gets you to saturation sooner, and in older adults a loading phase or a higher dose went with larger strength gains. (→ Nutrition 01: Protein intake; Nutrition sub-hub)

No dose is stated here. The trials cluster in a narrow range, and the comparison of doses in older adults found lean mass insensitive to it, which is a better argument for not fixating on the number than for printing one.

Where the goals diverge

Young men training for size or strength. The population the evidence was built on, and where the effect is clearest. What it buys is capacity to do more training, not a substitute for it. (→ Loading fundamentals 07: Weekly volume)

Women. Forty participants across the whole strength meta-analysis is not enough to make claims about. The direction of the evidence is the same; the confidence should not be.

Older trainees. The effect diminishes, and this is also where the dosing-strategy questions were studied. Both facts are worth knowing and they pull in different directions.

Endurance athletes. The mechanism serves short high-intensity efforts, which is the same quality covered elsewhere here. (→ Loading fundamentals 09: Movement velocity and intent) Weight gain from fluid is a cost rather than a neutral in a sport where you carry your own mass.

Anyone worried about the scale during a cut. The water retention is real and will show up as weight that is not fat.

The case against

The evidence base is male. We have said this twice because it is the largest limitation on everything above and it is routinely omitted.

The dosing-strategy findings come from older adults, the population in which creatine works least well. Extending "dose does not matter for lean mass" to young trained men is an extrapolation, even though it is a plausible one, and for strength the same analysis found that loading or dose did matter.

Absolute effect sizes vary considerably between analyses, partly because the outcome measures differ and partly because training status and duration differ across the pooled trials.

Supplement research has a funding problem, and the trials pooled here have not been screened for industry relationships.

The non-muscle claims are outside this piece. Cognitive, bone and clinical outcomes have their own literatures and their own standards of evidence.

What is not known

Whether the effect differs in trained lifters. The most recent dose-response analysis was built to answer this and found no significant difference, which is not the same as showing there is none.

Whether it differs meaningfully in women, at the level of evidence rather than at the level of assumption.

Whether the cell-swelling account contributes to adaptation or is only a measurement artifact. Both are stated in the literature and neither has been isolated.

Whether there is a dose below which it stops working, given that the dosing analysis compared strategies rather than mapping a dose-response curve.

Safety

Creatine monohydrate has a long safety record in healthy people. If you have kidney disease or take medication affecting kidney function, speak to a doctor first.