The quoted benefit disappeared in the trials that hid who got what
Cochrane pooled twenty trials and put glucosamine's pain benefit at 0.61, still the quoted number. The eight that properly concealed who got what showed none.
What the study found
Cochrane's glucosamine benefit vanished in the well-concealed trials
Pooling twenty trials, Cochrane put glucosamine's pain benefit at a standardised difference of 0.61, still the quoted number. The same review reports that the eight trials which properly concealed who got what showed none, and neither did trials using another manufacturer's preparation.
The caveat · Standard arthritis scores: no change
Which preparation a trial used and how rigorously it was designed cannot be separated either: the older, weaker trials are also that manufacturer's. On the standard osteoarthritis questionnaire, pain, function and stiffness reached significance for neither preparation. Its search closed in 2005.