The Surprise Hidden in a Knee Osteoarthritis Trial
Creatine has a reputation problem: it's the supplement for 22-year-old lifters chasing PRs. But a randomized trial in Medicine & Science in Sports & Exercise tested it in a very different group, postmenopausal women with knee osteoarthritis (OA). The creatine group got better at getting out of a chair. Their strength gains were no bigger than the placebo group's, and neither were their pain reductions.
Here's what the study actually found, and what it doesn't prove.
The Study at a Glance
- Design: Randomized, double-blind, placebo-controlled trial (Neves et al., 2011, University of São Paulo)
- Participants: Postmenopausal women with knee OA
- Sample size: The published abstract doesn't state the number of participants. Check the full text before leaning hard on the results, and remember trials like this tend to be small.
- Intervention: Creatine (20 g/day for 1 week, then 5 g/day) or placebo
- Training: Everyone followed a lower-limb resistance training program
- Duration: 12 weeks, with assessments at baseline and at the end
- Primary outcome: Physical function, measured by the timed-stands test
Because both groups trained, the question wasn't "does creatine beat doing nothing?" It was "does creatine add anything on top of strength training?"
The Headline Number: Standing Up Faster
The timed-stands score in the creatine group rose from 15.7 ± 1.4 to 18.1 ± 1.8, a gain of 2.4 points, or roughly 15%. The placebo group went from 15.0 ± 1.8 to 15.2 ± 1.2, essentially flat. The within-group improvement for creatine was significant (P = 0.006), and so was the between-group difference (P = 0.004).
The placebo group did the same training and saw no meaningful change on this test. That's what makes the result interesting.
More Wins Only Creatine Showed
The benefits weren't limited to the primary test. Only the creatine group showed significant improvements in:
- WOMAC physical function: P = 0.005
- WOMAC stiffness: P = 0.024
- Lower-limb lean mass: P = 0.04
- Quality of life: P = 0.01
The placebo group showed no significant changes in the WOMAC function and stiffness subscales (P > 0.05). The WOMAC is a standard questionnaire where patients rate their own pain, stiffness and difficulty with daily tasks, so these are changes the women themselves noticed.
Where Creatine Did Not Win
This is where the headline could easily be oversold, so the caveats matter.
Pain: Both groups reported significant pain reductions (P < 0.05). Creatine didn't beat placebo here. The most likely driver is the shared element, the supervised resistance training, though the study can't isolate that. Creatine is not shown to be a painkiller for arthritic knees.
Maximum strength: Leg-press one-repetition max rose over time in both groups (P = 0.005), with no difference between them (P = 0.81). Creatine did not make these women stronger on the leg press than placebo did.
So the story isn't "creatine builds bigger strength." Something else seems to be going on: better functional performance and more lean tissue without a measurable edge in 1RM. The authors' conclusion is limited to physical function, lower-limb lean mass and quality of life.
Why Might Function Improve Without a Strength Edge?
The paper's abstract doesn't settle this, so treat the following as informed speculation. Chair-stand tests reward repeated efforts, not one maximal lift. Creatine's best-known role is supporting rapid energy turnover in muscle during repeated short bouts of effort, which fits a test built on repetition. The extra lean mass in the legs may also have helped. A 1RM test and a repeated-stand test measure different things, so it's plausible they'd respond differently.
What This Means for You
Dose: The trial used a 1-week loading phase of 20 g/day, then 5 g/day. The standard 5 g/day is the maintenance dose and the one to focus on. Loading was part of this protocol, but the study didn't test whether skipping it would change results.
Timing: The abstract doesn't specify when during the day participants took it, so no timing claims can be drawn from it. Consistency matters more than clock time.
Training: Creatine was paired with lower-limb resistance training for 12 weeks. The result applies to creatine plus training, not creatine alone. If you're not strength training, don't expect the same payoff.
Who it may be relevant for: Postmenopausal women with knee OA who are already doing, or about to start, supervised leg strengthening, and who want an edge in everyday function such as standing, rising and stiffness.
Who should skip it or talk to a doctor first:
- Anyone with kidney disease or reduced kidney function, or taking medications that affect the kidneys
- Anyone expecting pain relief from creatine itself, since the trial showed no advantage over placebo
- Men, younger adults and people with other types of arthritis, who weren't studied here
- Anyone with significant medical conditions or on multiple medications, who should clear any supplement with their physician
The Fine Print
This is a single 12-week trial in one population, and the abstract doesn't report sample size. It deserves replication before it changes clinical guidelines. The results are encouraging but not definitive, and the secondary outcomes in particular should be read as supportive, not proof.
Still, the core takeaway is practical. For women with knee OA who are strength training, adding 5 g/day of creatine produced meaningful functional gains that training plus placebo did not. Pain and max strength came out the same either way. Creatine may be useful well outside the young-lifter demographic, but it works alongside the squat rack or leg-press machine, not in place of it.
