Clinical Question: Is glucosamine or chondroitin (or both) effective for osteoarthritis of the knee?

Setting: Outpatient (any)

Study Design: Randomized controlled trial (double-blinded)

Allocation: Concealed

Synopsis: A previous meta-analysis of glucosamine and chondroitin for osteoarthritis found a consistent benefit from treatment and good safety (McAlindon TE, et al. Glucosamine and chondroitin for treatment of osteoarthritis: a systematic quality assessment and meta-analysis. JAMA 2000;283:1469–75). However, many of the studies were small, did not report allocation concealment, were sponsored or conducted by drug manufacturers, or had other limitations that could lead to bias.

In this study, 1,583 patients with clinical and radiographic evidence of osteoarthritis of the knee were randomized (allocation concealed) to one of five groups: placebo; glucosamine hydrochloride 500 mg three times daily; chondroitin 400 mg three times daily; glucosamine hydrochloride 500 mg plus chondroitin 400 mg three times daily; or celecoxib (Celebrex) 200 mg once daily.

Most previous studies used glucosamine sulfate, although the importance of this substitution is of uncertain clinical significance. The mean age of patients was 58 years, 63 percent were women, and most were American Rheumatology Association functional class II (i.e., able to perform usual self-care and vocational activities but limited in avocational activities). The use of up to 4,000 mg acetaminophen was allowed (other than on the day of evaluation), and patients were followed up for 24 weeks.

Results showed an 85 percent probability of identifying a clinically significant 15 percent improvement difference between active treatment and placebo groups. The drop-out rate was approximately 20 percent in each group. The primary outcome of a 20 percent decrease in the Western Ontario and McMaster universities (WOMAC) osteoarthritis score, a validated pain score, was achieved in 60 percent of patients taking placebo, 64 percent taking glucosamine, 65 percent taking chondroitin, 67 percent taking glucosamine plus chondroitin, and 70 percent taking celecoxib. Only the difference between celecoxib and placebo was significant.

A review of the 14 secondary outcomes revealed trends toward greater benefit with glucosamine plus chondroitin or celecoxib, but these benefits were statistically significant for only two of these outcomes. The authors saw an interaction between the degree of pain and response to the study drugs. Although there was essentially no trend toward benefit, significant or otherwise, for patients with mild pain, a significant benefit occurred for the 354 patients with moderate or severe pain (WOMAC scores of 301 to 400). In these patients, a 20 percent decrease in the WOMAC pain score occurred for 54.0 percent receiving placebo, 79.2 percent receiving glucosamine plus chondroitin, and 69.4 percent taking celecoxib (P = .002 for placebo versus glucosamine plus chondroitin; number needed to treat = 4). The benefit for glucosamine plus chondroitin in these patients with more severe disease was larger than that for celecoxib. Similarly large benefits were seen for eight of 14 secondary outcomes. There was no important difference between groups regarding adverse events.

Bottom Line: Glucosamine hydrochloride plus chondroitin provides modest, if any, symptomatic benefit for patients with mild osteoarthritis of the knee. This study was well designed and avoided many of the design flaws of earlier studies. However, it had a high drop-out rate and used a different glucosamine salt than most previous studies. In addition, post hoc analysis suggests a large benefit in patients with moderate to severe pain. There also were consistent trends toward benefit for many secondary outcomes. (Level of Evidence: 1b)