Scoping Review Highlights Preliminary Nature of Creatine for Dementia

Dementia and its various subtypes represent a growing global health crisis, yet disease-modifying therapies remain frustratingly limited. Researchers are increasingly investigating metabolic interventions targeting shared pathological features, such as impaired brain energy metabolism, neuroinflammation, and synaptic dysfunction.

Because of its critical role in the phosphocreatine system—essentially a battery backup for cellular energy—creatine has emerged as a candidate intervention. However, while public interest in creatine for brain health is soaring, the actual scientific evidence remains in its infancy. A recent scoping review¹ sought to map the landscape of research evaluating creatine supplementation for dementia-related cognitive outcomes.

Key Takeaways

• No randomized trials exist at all: The search screened 8,317 records and identified no relevant RCTs. The entire human evidence base is a single uncontrolled pilot in 20 people.
• Preclinical dominance: Six studies met the criteria — one in vitro, four in animal models, one human pilot — so the field rests almost entirely on cells and rodents.
• Potential sex differences: A rodent study observed cognitive benefits in females but a trend toward harm in males, a divergence paralleled by bioenergetic responses in the single human pilot.
• Mechanistic plausibility, unproven efficacy: While creatine may support brain energy and modulate neuroinflammation, we currently lack the controlled trials needed to prove it improves symptoms of dementia.

The Study Design

To assess the state of the literature, the authors conducted a scoping review following Joanna Briggs Institute (JBI) methodology and PRISMA-ScR guidelines. They systematically searched four major databases (PubMed, Embase, Scopus, and CENTRAL) for English-language reports without date restrictions.

To be included, studies had to evaluate creatine supplementation in older adults with dementia or in preclinical animal models of dementia, and they needed to report at least one cognitive outcome. Beyond counting the studies, the review set out to map what they actually did — dose, duration, formulation and route of administration, along with the characteristics of the preclinical models used.

The Findings

The most striking finding of the review is how sparse the literature actually is. From 8,317 screened records, only six studies — reported across seven publications — met the criteria: one in vitro study, four animal model studies, and a single human pilot trial. The authors state it plainly: no relevant randomized controlled trials were identified.

In preclinical models, creatine demonstrated predominantly neuroprotective effects by providing bioenergetic support and modulating neuroinflammation. However, results were not universally positive; one rodent study reported a worsening of spatial memory.

The lone human pilot trial involved 20 participants who took a high dose of creatine (20 grams per day for 8 weeks). It increased total brain creatine by 11% and was associated with improved fluid cognition.

The design matters as much as the number. That trial is the CreAD feasibility pilot in Alzheimer's disease² — open-label and single-arm, with no placebo group, no randomization, and every participant knowing exactly what they were taking. It was built to establish that such a trial can be run at all, not to show that creatine works, and an eight-week cognitive retest in an unblinded cohort picks up practice effects as readily as any treatment effect. Our calculator carries this protocol for that reason and labels it feasibility evidence rather than demonstrated benefit.

Interestingly, researchers noted a potential sex divergence in the data. One rodent study found cognitive benefits exclusively in female animals, with males showing a trend toward harm. The human trial mirrored this divergence, noting sex-specific differences in bioenergetic responses to the supplement.

Analyzing the Study: Strengths & Limitations

Strengths and Reputability

The primary strength of this paper is its methodological rigor in mapping the existing literature. By utilizing established scoping review guidelines and casting a wide net across multiple databases, the authors successfully capture the current state of the field. Critically, the authors do not overstate their findings; they explicitly note that the current data cannot support causal inferences regarding creatine and dementia.

Limitations

The overwhelming limitation is the underlying evidence base itself. The entire human clinical literature for this application consists of 20 people in one uncontrolled pilot — the problem is not only that the sample is small, but that there is nothing to compare it against. Preclinical animal models, while helpful for understanding mechanisms, notoriously fail to translate into effective human therapies for Alzheimer's and other dementias. What little exists is also heterogeneous: doses, durations, formulations and routes of administration vary from study to study, so six reports cannot be pooled into anything resembling a dosing recommendation.

The authors' own prescription is worth repeating precisely. They call for adequately powered, sex-stratified randomized controlled trials in adults with dementia. The stratification is not a detail: benefit appeared only in female animals, males trended toward harm, and the human pilot showed sex-divergent bioenergetic responses. A trial that ignores sex could average a real effect in one group against harm in the other and conclude that nothing happened.

Conflicts of Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

1. Whitford T, Donlon C, Pech A, et al. The cognitive effects of creatine supplementation in older adults and preclinical models of dementia: A scoping review. Ageing research reviews. DOI: 10.1016/j.arr.2026.103317
2. Smith AN, et al. Creatine monohydrate pilot in Alzheimer's: Feasibility, brain creatine, and cognition. Alzheimer's & Dementia: Translational Research & Clinical Interventions. 2025. DOI: 10.1002/trc2.70101

Glossary

1. Scoping review: A type of literature review that aims to map the existing evidence base on a particular topic to identify gaps, rather than evaluating clinical efficacy or effect sizes. ↩

2. Open-label, single-arm: A trial in which everyone knows who is receiving the treatment and there is no comparison group — every participant gets the intervention. Such a design can show that a trial is feasible and that a biomarker moves, but it cannot separate the treatment from expectation, practice or the natural course of the disease. ↩