Creatine memory — meta-analysis double-counting

The most widely shared statistic about creatine and the brain, that it improves memory with an effect size of SMD 0.29, comes from a 2023 meta-analysis by Prokopidis et al. [1]. That number is now formally disputed, and the dispute matters if you are deciding whether to take creatine for cognitive reasons.

What the meta-analysis found

Prokopidis et al. pooled randomized controlled trials and reported a statistically significant memory benefit (SMD 0.29, 95% CI 0.04-0.53, p=0.02), with the largest effects appearing in adults aged 66-76 years [1]. That result got wide coverage, and it's the foundation for most current enthusiasm about creatine as a brain supplement.

The double-counting problem

A 2023 letter to the editor by Eckert and Pascher (PMID 36644917) argues the pooled effect is a false positive [2]. Their critique: the same participants and studies were counted multiple times across cognitive measures, inflating the apparent signal beyond what the underlying data support. The meta-analysis authors published a reply (PMID 36644912), but as of the current evidence review the dispute remains unresolved [2].

Verhoeven's analysis independently reaches the same conclusion, naming specific over-counted studies in the pool and pointing out that other cognitive outcomes, processing speed, for instance, showed no effect [3]. Two independent sources converging on the same methodological flaw is meaningful, even though Verhoeven's critique arrives as an analyst synthesis rather than peer-reviewed literature.

What does survive scrutiny

Stripped of the contested meta-analytic claim, the remaining evidence is narrower but more solid. Creatine improves cognition specifically under sleep deprivation [4] and acute stress [5]. The biological rationale fits: the brain's creatine-phosphate system buffers ATP, and that buffering matters most when energy demand is highest, during prolonged wakefulness or physiological stress. In rested, healthy adults, that mechanism appears to offer little additional benefit [6].

A 2024 systematic review found the proposed mechanism does not translate into a consistent cognitive benefit in the general healthy population [6], which aligns with the narrowed picture from the stress-context trials.

Where enthusiasm outruns the evidence

Some researchers and communicators have argued that 10 g daily is the appropriate brain-targeted dose, and pilot work has explored doses as high as 20 g in Alzheimer's disease populations [7, 8]. Those claims go beyond what the current evidence base supports. The solid signal, cognition under acute stress or sleep deprivation, comes from specific controlled conditions, not from general supplementation in healthy adults. The Alzheimer's work is at the pilot stage, and inferring a general neuroprotective effect from it is premature.

Where the evidence is contested

The central dispute is whether creatine produces a genuine general memory benefit. Prokopidis et al. say yes, reporting a moderate effect size driven largely by older-adult studies [1]. Eckert and Pascher counter that the statistical pooling double-counts data, making the headline result unreliable [2]. Verhoeven's analysis supports this methodological critique. It adds that correcting for the over-counting may still leave a residual effect, though that corrected re-analysis has not been independently verified [3].

Neither side questions the stress-context evidence. The disagreement is specifically about whether a benefit exists in unstressed, healthy populations. The double-counting question awaits a properly conducted re-analysis of the individual RCTs. Until then, the defensible position is this: creatine aids cognition under sleep deprivation and acute stress; a general memory benefit in healthy people is disputed and should not be assumed.

References

1. Prokopidis et al. 2023, Nutrition Reviews — creatine & memory meta-analysis (disputed)
2. Eckert & Pascher 2023 — Letter: double-counting in Prokopidis meta-analysis · PubMed 36644917
3. Physionic — "Creatine Expert uncovers why 5 grams isn't Enough!" (video 06)
4. Gordji-Nejad 2024, Sci Rep — creatine & sleep-dep cognition · PubMed 38418540
5. McMorris et al. 2006 — creatine, sleep deprivation & cognition · PubMed 17046034
6. 2024 systematic review — creatine cognition (skeptical) · PubMed 38582412
7. FoundMyFitness — Darren Candow: optimal creatine protocol
8. FoundMyFitness — Rhonda Patrick: creatine cognitive benefits