SHREDDEDAF
Reference
10 NOV / 26
Reference room/Supplements
Supplements / Creatine

Creatine: what the evidence actually supports

Creatine monohydrate has an unusually deep evidence base for a supplement. That does not make every product equivalent, every claimed benefit proven or every population equally studied.

Creatine
without hype
05

The evidence does not support treating creatine as a miracle powder. It supports a narrower, more useful conclusion: creatine monohydrate can improve capacity for repeated high-intensity work and can add to some resistance-training adaptations.

Bottom line

For generally healthy adults who resistance-train, creatine monohydrate is one of the best-supported ergogenic supplements. Average effects are modest, individual response varies, and an increase in lean mass is not automatically an equal increase in new muscle tissue.

Body composition: a real signal with a measurement caveat

A 2024 systematic review and meta-analysis of 12 studies in adults younger than 50 found that adding creatine to resistance training increased measured lean body mass by 1.14 kg versus resistance training alone. It also reported small reductions in body-fat percentage and fat mass. Because creatine changes intracellular water, measured lean mass cannot be read as pure contractile-muscle gain.

Older adults: strength is clearer than function

A 2026 three-level meta-analysis of 11 randomized trials in adults aged 60 and above found a small additional strength benefit when creatine was combined with resistance training (g=0.31; moderate certainty). Pooled effects on muscle mass and physical function were not statistically significant and were graded low or very-low certainty. The authors warned against extrapolating directly to frail, sarcopenic or functionally limited older adults.

Safety: do not collapse “studied” into “safe for everyone”

A 2026 systematic review of randomized trials found no difference in urea or estimated glomerular filtration rate between creatine and control groups, while calling for trials extending beyond one year. A separate 2025 analysis covering 685 creatine trials found no overall difference in participant-reported side-effect frequency versus placebo.

Those are reassuring signals for studied populations, mostly using creatine monohydrate. They do not establish safety for contaminated or multi-ingredient products, undisclosed doses, pregnancy, children, or people with kidney disease or other clinical constraints. The 2025 safety paper also disclosed multiple supplement-industry relationships; its scale is useful, but the conflicts belong in the interpretation.

Practical audit translation

Creatine remains its own product class. The audit questions are form, disclosed quantity, independent quality testing, unnecessary ingredients, claims and cost—scored separately. “Advanced creatine matrix” is not an evidence grade. No brand or seller has paid for this review, and there are no affiliate links on this page.

References

  1. Desai I et al. Creatine plus resistance training and body composition in adults under 50: systematic review and meta-analysis of 12 studies. PubMed 39074168.
  2. Yao L et al. Creatine plus resistance training in adults aged 60+: systematic review and three-level meta-analysis of 11 randomized trials. PubMed 42712402.
  3. Kalaitzidis RG et al. Creatine and kidney function: systematic review and meta-analysis of randomized trials. PubMed 42035842.
  4. Kreider RB et al. Side effects in 685 creatine trials and adverse-event reports; industry conflicts declared. PubMed 40198156.
  5. U.S. NIH Office of Dietary Supplements. Exercise and athletic-performance supplements fact sheet. Issuing authority.
Evidence grade — A− for the core performance use case: multiple randomized trials, meta-analyses and an authoritative evidence review support creatine monohydrate for high-intensity exercise and some training adaptations. The grade is not an A because outcomes and response vary, lean-mass measurement is imperfect, and long-term or higher-risk population data are thinner. Last checked 9 November 2026.
← Back to the reference room