Efficacies of different preparations of glucosamine for the treatment of osteoarthritis: a meta-analysis of randomised, double-blind, placebo-controlled trials

Abstract

Objective: To determine the efficacies of different preparations of glucosamine for the treatment of osteoarthritis (OA).

Methods: Systematic searches of the bibliographic databases Medline, Embase, the Cochrane Central Register of Controlled Trials and the Cochrane Database of Systematic Reviews for randomised, double-blind, placebo-controlled trials (RCTs) concerning glucosamine treatment of OA. Effect size (ES) was estimated using Cohen's standardised mean difference. Consistency was evaluated via the I(2) index.

Results: Nineteen trials (3159 patients) contributed to the meta-analysis, revealing a large degree of inconsistency among the trials in terms of pain-reduction outcome: the combined ES in glucosamine sulphate (GS) trials was -0.22 [95% confidence intervals (CI) -0.48, 0.04], I(2) was 82.3%. The combined ES in glucosamine hydrochloride (GH) trials was -0.03 (95% CI -0.14, 0.08), with an absence of heterogeneity. No treatment ES was observed [-0.38 (95% CI -0.99, 0.23)] favouring GS in trials of less than 24 weeks duration and the I(2) remained high (I(2) = 88.5%). No significant treatment ES -0.09 (95% CI -0.21, 0.03) was observed in trials of more than 24 weeks duration compared with placebo, with a heterogeneity of zero. In terms of function-modifying outcomes, GS showed no significant effect on Lequesne Index reduction vs. placebo in trials of less than 24 weeks duration (ES -0.55 (95% CI -1.22, 0.11)) with a high degree of heterogeneity (I(2) = 92.9%). Pooling data from studies with durations of more than 24 weeks presented a significant combined ES of -0.36 (95% CI: -0.56, -0.17) with an absence of heterogeneity. No risk of publication bias could be detected using Egger test.

Conclusions: GH is ineffective for pain reduction in patients with knee OA. GS may have function-modifying effects in patients with knee OA when administered for more than 6 months. However, it showed no pain-reduction benefits after 6 months of therapy.