Creatine helps most when you train hard and often — for everyone else it is optional, not a must
Asked (summary): Is creatine good or bad, at what dose and timing? How does it play out for sedentary people, people doing CrossFit about twice a week, and more active people — crossed with carnivorous, flexitarian and vegetarian diets? Is it a must-take, and what are the downsides and risks, given that lobbies might be pushing for it?
This report draws on a curated library of 83 evidence and resource pages across 43 independent web hosts, selected from much larger broad searches — 57 peer-reviewed papers (systematic reviews, meta-analyses and randomized trials), one professional position stand, and 25 secondary or commercial summaries kept for context. Doses are grams of creatine monohydrate per day; strength and lean-mass effects are trial differences versus placebo. Inclusion in the library is not endorsement, and some rows summarize other sources.
Decision matrix: activity level × diet pattern
Colour = likely marginal benefit of 3–5 g/day creatine monohydrate for that person. Hover or tap a cell for the reasoning; cells marked "inferred" have no direct trials and are read from the nearest evidence.
likely worthwhile benefitpossible, modest benefituncertain or small"inferred" = no direct trials for this exact cell
Hover a cellDirect evidence for "CrossFit twice weekly" and for carnivorous or flexitarian strata is sparse; those cells are labelled as inferences from the nearest studied groups, not established facts.
With resistance training, creatine adds about 1.1–1.5 kg lean mass across ages, and roughly none without exercise — doi.org
In adults under 50, creatine plus resistance training added ~4.4 kg upper-body and ~11.4 kg lower-body strength over training alone — mdpi.com
Across 684 randomized trials (>12,800 participants), side effects showed no dose- or duration-dependent increase; placebo arms often reported similar or higher rates — holdsup.app
Endurance scoping review: creatine is not supported as a general endurance aid; gains concentrate in sprint and power-endurance tasks — mdpi.com
Who is most and least likely to benefit
Most: people doing regular resistance or repeated high-intensity training — the strongest, most consistent evidence base (tandfonline.com).
Also likely: vegetarians and vegans, whose lower dietary creatine means larger increases in tissue creatine and sometimes cognition or performance (mdpi.com); and older or functionally at-risk adults, especially combined with resistance exercise (aspenjournals.onlinelibrary.wiley.com).
Modest at best: training only about twice weekly — repeated-effort work still responds, but improving programming and adherence likely pays more (inference; this stratum is unstudied).
Least: sedentary younger adults expecting muscle gain — creatine without exercise adds essentially no lean mass. Cognitive benefits look more promising with low baseline creatine, older age, sleep deprivation, or dietary restriction, and are not a universal guarantee (tandfonline.com).
Bottom line: creatine is not a must-take and no substitute for training, sleep, protein and overall diet — an optional, comparatively well-studied aid.
Dose and timing
Form: plain creatine monohydrate. HCl, buffered, ethyl-ester, nitrate and other variants have not shown superiority (sciencedirect.com).
Simple approach: 3–5 g every day. Consistency matters far more than workout timing; saturation without loading takes several weeks.
Optional loading: ~0.3 g/kg/day (about 20 g/day in 4 doses) for 5–7 days, then 3–5 g/day (tandfonline.com). Loading is not required.
Comfort: split doses to ≤5 g each to limit gastrointestinal upset (frontiersin.org).
Quality: choose reputable third-party-tested products, especially for tested athletes.
Risks and reading your lab tests
Early scale-weight gain from intracellular water is common and is not fat gain.
Gastrointestinal upset is possible, mostly with large single doses.
Randomized-trial safety analyses find adverse-event rates similar to placebo in healthy users (holdsup.app).
Kidney meta-analysis: a small rise in serum creatinine without a significant fall in measured or estimated GFR — tell your clinician you take creatine, because creatinine can create a misleading kidney-test signal (link.springer.com).
Long-term data beyond roughly a year are less abundant; hair-loss claims remain unproven rather than definitively excluded.
Pregnancy, breastfeeding, minors, pre-existing kidney disease, or potentially nephrotoxic medication are not routine self-supplementation situations — get clinician review, and stop and seek care for unusual symptoms.
Diet patterns and evidence gaps
Vegetarians and vegans have lower dietary creatine and often show larger tissue-creatine increases, and sometimes cognitive or performance gains, after supplementation (mdpi.com). Switching omnivores to a vegetarian diet reduces the body creatine pool (cambridge.org).
Omnivores and high-meat eaters still respond, but the average marginal benefit is likely smaller because baseline stores are often higher.
Gaps: no trials separate carnivorous from ordinary omnivorous or flexitarian diets; "CrossFit about twice weekly" has no direct trials; female-specific and long-term (>1 year) evidence is thinner. The matrix presents those cells as likely gradients, not facts.
Evidence versus advocacy
Several forceful pro-creatine reviews and position pieces disclose ties to creatine manufacturers, supplement companies, trade groups, advisory boards, consulting, grants, donations, or paid article-processing charges — for example, a piece arguing everyone should take 2–3 g/day "for general health" whose authors include the chair of a creatine manufacturer's scientific advisory board (frontiersin.org). That universal-need claim is advocacy, not settled consensus. Yet the core benefits for resistance and high-intensity training, and the reassuring kidney findings, also appear in independent systematic reviews declaring no funding and no conflicts (link.springer.com). Industry involvement is a reason to grade claims carefully, not to dismiss the literature. The table below flags every row with declared conflicts so you can filter on them.
Full evidence library
83 rows, one per source URL. Peer-reviewed and guidance rows lead; blogs, commercial summaries and secondary write-ups are shown greyed. Declared conflicts are flagged in orange. Inclusion is not endorsement.
Title
Tier
Dose / regimen noted
Finding
Limitations / conflicts
Source
Method: curated evidence library of 83 web resources on creatine supplementation across 43 independent hosts, drawn from much larger broad searches; each row backed by its own URL (no URL backs more than one row). Rows record title, population, dose, finding, limitations and declared conflicts as stated by each source; tiers (peer-reviewed / guidance / secondary / animal) were assigned by host and content type. Numbers measure trial differences vs placebo in the units each source reports. Long fields are truncated for space; follow each link for the full source. Decision-matrix cells without direct trials are inferences from the nearest evidence and labelled as such.