Answers are grounded in this paper and cite their sources.
Creatine reliably raises serum creatinine, and prior meta-analyses report this without renal harm, relying on creatinine-derived filtration estimates. We asked how much of this reassurance rests on filtration measured independently of creatinine and, as an exploratory question, whether the creatinine rise differs by dosing protocol. We searched PubMed, Scopus, the Cochrane Library and OpenAlex unitl 6 August 2026 for randomised trials of oral creatine monohydrate versus placebo/control in healthy, physically active adults aged 18–45. Serum creatinine was pooled as a mean difference (µmol/L) and filtration/urea as standardised mean differences by random-effects meta-analysis (REML; DerSimonian–Laird, Paule–Mandel and Hartung–Knapp adjustment as sensitivity analyses). Dosing protocol was a pre-specified subgroup; filtration was additionally stratified by measurement method (post hoc) (PROSPERO CRD420261470453). Twelve trials (356 randomised, 325 analysed) were included; eight (11 comparisons) contributed to the creatinine analysis. Creatine raised serum creatinine by 12.10 µmol/L (95% CI 8.42–15.79; ≈ 0.14 mg/dL), with no detectable difference between dosing protocols ( P = 0.85; exploratory, low power). No change in urea was detected. For filtration, four comparisons used creatinine-derived estimates (SMD − 0.68, 95% CI − 1.50 to 0.13; I 2 = 73%) and one used 51 Cr-EDTA clearance (SMD 0.13, 95% CI − 0.61 to 0.88; 43% attrition); the overall pooled SMD (− 0.50, 95% CI − 1.19 to 0.19) mixes these methods and is not interpreted. Certainty was low to very low across outcomes. Of 46 reports sought, 32 could not be retrieved. Among the trials we could retrieve, only one small, high-attrition trial assessed filtration independently of creatinine, finding no change. The evidence is compatible with no effect of creatine on kidney function in healthy adults but, being largely indirect and of low certainty, does not establish renal safety. Trials with reference-standard filtration are a priority.
No takes yet. Share an insight, caveat, or question.
Raihan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: