Does creatine really boost your brain?

Disclosure first, because it matters here: I formulate and sell supplements for a living. My company makes functional-mushroom products, not creatine, so I don't sell the thing this report is about. But I'm in the supplement business, so I hold the whole category (my own shelf included) to the standard I'd want a skeptic to use on me. Creatine has earned an unusual amount of goodwill, and that goodwill is exactly what a new claim borrows against. So let's read the paper the new claim rests on.

First, the part that's genuinely solid, because being fair to creatine is the whole point of being hard on it. For building muscle and strength alongside resistance training, creatine monohydrate is one of the best-supported supplements that exists. Decades of trials, a clear mechanism, real effect sizes. If you lift and you take it, you're on firm ground. That is not what's in question. What's in question is the newer second act: creatine sold as a nootropic, a supplement for memory, focus, and brain fog. That pitch almost always points at the same piece of evidence.

The paper everyone cites

In 2024, Frontiers in Nutrition published a systematic review and meta-analysis by Xu and colleagues titled, plainly, "The effects of creatine supplementation on cognitive function in adults." A meta-analysis pools many trials into one combined estimate, and it sits near the top of the evidence pyramid, so a positive one makes a strong marketing anchor. And the coverage treated this one as a green light: creatine, science says, is good for your brain.

Now open the actual results. The headline outcome, the combined measure of overall cognitive function, was not statistically significant (a standardized mean difference of 0.34, p = 0.22). Read that again, because it's the sentence the "creatine for your brain" posts skip: the paper's own top-line result did not reach significance. What did move was a scatter of sub-domains. Memory came back positive. So did "attention time" and "processing-speed time," meaning people completed certain timed tasks a bit faster. But attention scores (p = 0.49), processing-speed scores (p = 0.97), and executive function (p = 0.12) all came back null, right alongside the null overall result. The authors' careful conclusion was that creatine "may confer beneficial effects on cognitive function in adults, particularly in the domains of memory, attention time, and information processing speed." That is a much narrower and more hedged statement than "boosts your brain," and it's already the optimistic reading.

Then a reviewer pulled the thread

Here's where 2026 comes in. In April, Frontiers in Nutrition published a formal Commentary by Tom Citherlet flagging a statistical problem in that meta-analysis, the kind that's invisible unless you know to look for it. It's called a unit-of-analysis error, and it's simpler than it sounds. Many of the pooled trials reported several cognitive subtests measured on the same people, and the meta-analysis counted each subtest as if it were an independent result. As Citherlet put it, "the number of observations in the pooled analysis exceeds the number of unique randomized participants." You're effectively counting the same volunteers more than once.

That "double-counting," in his words, "violates the assumption of independent observations" and artificially inflates the statistical precision. In plain terms: it makes the confidence intervals look tighter and the p-values look smaller than the actual number of human beings in the studies justifies. It can turn a wobbly "maybe" into a confident-looking "yes." The commentary points out that the European Food Safety Authority had already looked at this kind of pooled data and concluded "no conclusions could be drawn" from it. And when a separate group (Prokopidis and colleagues) re-ran the analysis with proper statistical handling, the effect of creatine on memory "was no longer significant, except in older adults." The one sub-domain that looked most convincing largely dissolved once the same people stopped being counted as different people. It's worth noting the original paper had already needed a published correction in 2025, a sign it was under real scrutiny before this commentary landed.

The honest read, both directions

This is not "creatine does nothing for your brain," and I won't overclaim in the other direction to make a cleaner story. There's a plausible mechanism: the brain runs on a lot of energy, creatine is part of how cells buffer energy, and it's reasonable that a demand-stressed brain (sleep-deprived, aging, under load) might benefit more than a rested young one. The surviving "except in older adults" signal fits that idea. What the evidence does not support is the flat, universal "take creatine, think better" claim aimed at healthy adults. That claim rests on a meta-analysis whose overall result was null and whose positive pieces were challenged, on the record, by a peer reviewer doing exactly what peer review is supposed to do.

The signal

A meta-analysis feels like the final word because it's built from many studies, but it inherits every weakness of the studies feeding it, and it can add fresh ones in how it pools them. So when a label or a thread waves a meta-analysis at you, two questions do most of the work. First: did the overall, pre-specified outcome actually move, or is the story being told from a handful of sub-results that happened to land? Second: were the data points independent, or is the same handful of people being counted over and over? Those are the same instincts as reading any "clinically proven" label: find the real number, and ask what it's actually a number of. Creatine is a great muscle supplement with a real, unfinished brain question attached. "Unfinished" is an honest word. "Proven" isn't, not yet.

Disclosure, again, plainly: I founded and run Shroombiosis (a company I run), which formulates and sells functional-mushroom supplements. It does not sell creatine, but I have a commercial stake in the supplement category generally, which is why I read even the popular, well-liked products this critically. Nothing here is sponsored and no link earns a commission; here's the full policy. And a recommendation with no stake at all: creatine is squarely a performance-nutrition product, and Die Tryin Co., a fellow combat-veteran-owned brand, is one I'm glad to point you to for that category; I don't own it and earn nothing from the link.

Not medical advice. This is educational analysis, not a recommendation, and a study is not a prescription. Talk to a qualified clinician before acting on anything you read here. Full disclaimer →