Creatine muscle — water weight vs hypertrophy

Creatine's reputation for adding bulk has a well-earned asterisk: a chunk of that early weight gain is water, not muscle. A 2025 washout trial by Desai and colleagues pinned down exactly how much. Participants loading creatine gained 0.51 kg within seven days (p=0.03), and the researchers attributed that rise to fluid retention [1]. After 12 weeks of resistance training, creatine showed no lean-body-mass advantage over placebo (p=0.71) [1]. That's a rigorous finding from a controlled design, and it merits serious consideration.

But one well-powered trial isn't the whole picture.

What the meta-analysis adds

Chilibeck et al. (2017) pooled data across many creatine-plus-resistance-training studies in older adults and found a mean lean tissue gain of roughly 1.37 kg compared to resistance training alone, alongside strength improvements [2]. That's a small but real benefit at the population level. The acute scale jump is mostly water, while a modest genuine lean-mass effect accumulates across longer timelines and multiple trials. The Desai null result and the Chilibeck meta-analysis address different questions rather than contradicting each other.

One caveat applies here. Darren Candow, a researcher with a well-documented pro-creatine publication history, is a co-author on the Chilibeck meta-analysis [2]. That doesn't invalidate the pooled data, but it is worth keeping in mind when weighing the confidence interval around that 1.37 kg figure.

Fibre-level evidence

DEXA and lean-body-mass scales both have a weakness: they can't distinguish water inside a muscle cell from actual contractile protein. Biopsy data sidestep that problem. Volek 1999 used direct muscle biopsies and found greater cross-sectional area gains in Type I, IIA, and IIAB fibres in creatine-supplemented subjects compared to controls [3]. Fibre cross-sectional area is a structural measure. Fluid doesn't show up as enlarged myofibrils under a microscope, so this finding points toward genuine hypertrophy rather than cell swelling alone.

Verhoeven's analysis suggests the biopsy evidence, together with the meta-analytic lean-mass data, supports the conclusion that creatine produces real fibre growth on top of its cell-volume effect [4]. That framing aligns with the primary literature from Desai, Chilibeck, and Volek, though Verhoeven's synthesis should be read as interpretive rather than a primary finding.

How the pieces fit together

The three studies land in different places for a reason: they're measuring different things at different time points. Desai captured a single 12-week block with a careful washout to strip out water [1]. Chilibeck pooled results across many trials where cumulative training adaptations had more time to compound [2]. Volek looked directly at fibre architecture rather than whole-body composition [3]. Together, they support a coherent picture: creatine causes rapid fluid retention that inflates early scale readings, and separately promotes a modest but structural increase in muscle fibre size over time when combined with resistance training.

Creatine doesn't work by magic, and the initial weight gain isn't muscle. The muscle that does accumulate appears to be genuine, smaller and slower than the first week of the scale suggests.

Where the evidence diverges

The Desai 2025 RCT found no lean-body-mass advantage after 12 weeks of creatine-supplemented training (p=0.71), a clean null result from a study designed to separate water from tissue [1]. This is primary evidence that the short-term LBM bump is predominantly water.

The Chilibeck 2017 meta-analysis, by contrast, reports a meaningful positive lean-mass effect across the broader literature [2]. A single 12-week trial and a pooled estimate integrating variation across populations, durations, and training protocols will not always agree; the two findings are not fully harmonized at the individual-study level. Researchers who weight Desai's washout methodology heavily will read the evidence as more skeptical of creatine's hypertrophic effect than those who weight the meta-analytic average. The co-authorship conflict of interest on [2] adds a further reason for caution on that side of the ledger.

References

1. Desai et al. 2025, Nutrients 17(6):1081 — creatine washout LBM RCT · PubMed 40292479
2. Chilibeck et al. 2017 — creatine+RT lean mass meta (older) · PubMed 29138605
3. Volek et al. 1999, Med Sci Sports Exerc — creatine + RT, muscle-fibre hypertrophy (biopsy) · PubMed 10449017
4. Physionic — "Creatine Debunked: No Extra Muscle Gained" (video 02)