Chondroitin's best evidence: 53 in 100 helped, against 47 on placebo
Cochrane pooled 43 trials in more than 9,000 people and rated almost everything it found poor quality. On its one high-quality outcome the gap was six in a hundred.
What the study found
Chondroitin's best-quality result is six people in a hundred
Cochrane pooled 43 trials in more than 9,000 people, and rated almost everything it found poor quality. On the single outcome it rated high quality, 53 people in 100 taking chondroitin reached a meaningful drop in knee pain against 47 in 100 on placebo. Small is not the same as nothing.
The caveat · Fades in large, independent trials
The 10-percentage-point figure usually quoted from it is the one it rates low quality, with high bias risk and heavy disagreement. The benefit became uncertain whenever the analysis narrowed to trials that concealed allocation properly, to large trials, or to independently funded ones.
What we don't know
Chondroitin's knee-pain benefit is small, and whether it is worth having is unsettled. The largest trials put it below the smallest change a patient would notice, while Cochrane counts six people in a hundred helped.