We searched seven databases up to November 2013, including the Cochrane Central Register of Controlled Trials (CENTRAL), Ovid MEDLINE, CINAHL, EMBASE, Science Citation Index (Web of Science) and Current Controlled Trials. We searched the US Food and Drug Administration (FDA) and European Medicines Agency (EMEA) websites for adverse effects. Trial registers were not searched.
SELECTION CRITERIA: Two review authors independently performed all title assessments, data extractions, and risk of bias assessments.
MAIN RESULTS: Forty-three randomized controlled trials including 4,962 participants treated with
chondroitin and 4,148 participants given placebo or another control were included. The majority of trials were in knee OA, with few in hip and hand OA. Trial duration varied from 1 month to 3 years. Participants treated with
chondroitin achieved statistically significantly and clinically meaningful better
pain scores (0-100) in studies less than 6 months than those given placebo with an absolute risk difference of 10% lower (95% confidence interval (CI), 15% to 6% lower; number needed to treat (NNT) = 5 (95% CI, 3 to 8; n = 8 trials) (level of evidence, low; risk of bias, high); but there was high heterogeneity between the trials (T(2) = 0.07; I(2) = 70%, which was not easily explained by differences in risk of bias or study sample size). In studies longer than 6 months, the absolute risk difference for
pain was 9% lower (95% CI 18% lower to 0%); n = 6 trials; T(2) = 0.18; I(2) = 83% ), again with low level of evidence.For the Western Ontario and McMaster Universities
Osteoarthritis Index Minimal Clinically Important Improvement (WOMAC MCII
Pain subscale) outcome, a reduction in knee
pain by 20% was achieved by 53/100 in the
chondroitin group versus 47/100 in the placebo group, an absolute risk difference of 6% (95% CI 1% to 11%), (RR 1.12, 95% CI 1.01 to 1.24; T(2) = 0.00; I(2) = 0%) (n = 2 trials, 1253 participants; level of evidence, high; risk of bias, low).Differences in Lequesne's index (composite of
pain,function and disability) statistically significantly favoured
chondroitin as compared with placebo in studies under six months, with an absolute risk difference of 8% lower (95% CI 12% to 5% lower; T(2)= 0.78; n = 7 trials) (level of evidence, moderate; risk of bias, unclear), also clinically meaningful. Loss of minimum joint space width in the
chondroitin group was statistically significantly less than in the placebo group, with a relative risk difference of 4.7% less (95% CI 1.6% to 7.8% less; n = 2 trials) (level of evidence, high; risk of bias, low).
Chondroitin was associated with statistically significantly lower odds of serious adverse events compared with placebo with Peto odds ratio of 0.40 (95% CI 0.19 to 0.82; n = 6 trials) (level of evidence, moderate).
Chondroitin did not result in statistically significant numbers of adverse events or withdrawals due to adverse events compared with placebo or another
drug. Adverse events were reported in a limited fashion, with some studies providing data and others not.Comparisons of
chondroitin taken alone or in combination with
glucosamine or another supplement showed a statistically significant reduction in
pain (0-100) when compared with placebo or an active control, with an absolute risk difference of 10% lower (95% CI 14% to 5% lower); NNT = 4 (95% CI 3 to 6); T(2) = 0.33; I(2) = 91%; n = 17 trials) (level of evidence, low). For physical function,
chondroitin in combination with
glucosamine or another supplement showed no statistically significant difference from placebo or an active control, with an absolute risk difference of 1% lower (95% CI 6% lower to 3% higher with T(2) = 0.04; n = 5 trials) (level of evidence, moderate). Differences in Lequesne's index statistically significantly favoured
chondroitin as compared with placebo, with an absolute risk difference of 8% lower (95% CI, 12% to 4% lower; T(2) = 0.12; n = 10 trials) (level of evidence, moderate).
Chondroitin in combination with
glucosamine did not result in statistically significant differences in the numbers of adverse events, withdrawals due to adverse events, or in the numbers of serious adverse events compared with placebo or with an active control.The beneficial effects of
chondroitin in
pain and Lequesne's index persisted when evidence was limited to studies with adequate blinding or studies that used appropriate intention to treat (ITT) analyses. These beneficial effects were uncertain when we limited data to studies with appropriate allocation concealment or a large study sample (> 200) or to studies without
pharmaceutical funding.
AUTHORS' CONCLUSIONS: A review of randomized trials of mostly low quality reveals that
chondroitin (alone or in combination with
glucosamine) was better than placebo in improving
pain in participants with
osteoarthritis in short-term studies. The benefit was small to moderate with an 8 point greater improvement in
pain (range 0 to 100) and a 2 point greater improvement in Lequesne's index (range 0 to 24), both seeming clinically meaningful. These differences persisted in some sensitivity analyses and not others.
Chondroitin had a lower risk of serious adverse events compared with control. More high-quality studies are needed to explore the role of
chondroitin in the treatment of
osteoarthritis. The combination of some efficacy and low risk associated with
chondroitin may explain its popularity among patients as an over-the-counter supplement.