Chondroitin won't relieve the pain of osteoarthritisBMJ 2007; 334 doi: https://doi.org/10.1136/bmj.334.7599.873-a (Published 26 April 2007) Cite this as: BMJ 2007;334:873
A meta-analysis has shown that chondroitin is no better than placebo for reducing the pain of osteoarthritis. A linked editorial (pp 611-2) points out that many patients have a powerful belief in the supplement, and that continuing to take it probably won't do them any harm⇑.
In an unusual step, the researchers chose to focus their analysis on the three biggest and best trials after an initial meta-analysis showed that early trials were small, heterogeneous, poorly done, and potentially biased. As usual, poor quality trials were more likely than better quality trials to report positive results.
The three best trials included 40% of all patients randomised. Most had osteoarthritis of the knee. Together, they suggest that chondroitin takes about half a millimetre off patients' scores on a 10 cm visual analogue scale compared with placebo. The difference was insignificant statistically and barely noticeable clinically. The effect of chondroitin on joint space narrowing was hard to evaluate from the existing data, but the researchers say benefits look unlikely.
This is another example of an intervention that looks useful in early trials but proves not so useful once bigger and better trials are complete, says the editorial.