21 October 2010
A network meta-analysis has concluded that glucosamine and chondroitin, given either separately or together, do not affect pain intensity in osteoarthritis (OA) to a clinically meaningful extent compared with placebo. This supports existing NICE guidance that the use of glucosamine or chondroitin products is not recommended for the treatment of OA.
Level of evidence:
Level 2 (limited quality patient-oriented evidence) according to the SORT criteria.
Health professionals should continue to follow NICE guidance and not usually prescribe glucosamine or chondroitin products for the treatment of OA. However, if patients want to trial over-the-counter glucosamine, they should be advised that the only potential benefits so far identified are purely related to a reduction of pain (in some people, and to only a mild or modest degree) with glucosamine sulphate 1500mg/day. Health professionals should also advise patients on how to perform their own trial of therapy, that is, to evaluate their pain before starting glucosamine and ensure they review the benefits of glucosamine after 3 months. In addition, they should note that the MHRA has warned against its use by people who have seafood allergies or those who are taking warfarin.
What is the background to this?
Glucosamine and chondroitin are so called ‘nutraceuticals’ – foods with [alleged] medical or health benefits. NICE OA guidance clearly states that the use of glucosamine and/or chondroitin is not recommended for the treatment of OA under the NHS, based on a review of the evidence. Nevertheless, between 2003 and 2008 global sales of glucosamine supplements increased by about 60% to almost £1.3bn, with a forecasted continued growth reaching about £1.5bn by 2013. In the quarter to March 2010, NHS primary care in England spent about £7.8m glucosamine and chondroitin products (personal communication, NHSBSA PPD). Glucosamine sulphate 1500mg tablets have been included in the October 2010 Drug Tariff at £36.28 for 30, and glucosamine sulphate 500mg /chondroitin 400mg tablets are included at £37.53 for 30. These are much more costly than generic ibuprofen or naproxen.
There have been numerous RCTs and meta-analyses (MAs) of glucosamine and chondroitin (see the ‘how does this relate to other studies’ section, below). The authors of this MA used a novel methodology (network meta-analysis) to incorporate direct and indirect comparisons between glucosamine or chondroitin or the combination and placebo. The primary outcome considered was the absolute difference in pain scores, translated to differences in a 10cm visual analogue scale (VAS) . Secondary outcomes were changes in the minimum radiographic joint space between baseline and the end of treatment; the number of individuals who withdrew because of an adverse event, and the number of patients experiencing any adverse event. The MA included 10 trials of 3,803 patients with hip or knee OA.
What does this study claim?
The authors specified in advance that to be clinically meaningful, the improvement in pain score would have to be greater than 0.9cm on a 10cm VAS. They chose this based on previous studies of what is likely to constitute a clinically meaningful improvement from treatments in OA: other studies have found that 1.5 to 2.0cm would be needed to be clinically meaningful.
Neither glucosamine nor chondroitin, separately or in combination, produced an improvement in pain score compared to placebo which was likely to be clinically meaningful (based on the 0.9cm criterion) used here at conventional levels of statistical significance (see the ‘study details’ section below). There were no statistically significant effects on radiographic joint space, withdrawals due to adverse events or rates of adverse events.
How does this relate to other studies?
For some readers, it may be sufficient to know that the MA’s conclusions are consistent with NICE guidance. For those who would like more background, this section briefly reviews other important MAs.
There have been numerous MAs of glucosamine and chondroitin. The 2009 Cochrane review of glucosamine found a statistically significant improvement in pain from baseline overall, but not when the analysis was restricted to studies with adequate concealment of allocation. Inadequate allocation concealment is an important potential source of bias, which is discussed more fully in part 4 of the Information mastery skills eLearning recording on NPC.
A systematic review which sought to investigate the heterogeneity in glucosamine trial results found that this was greater than would be expected by chance. It concluded that potential explanations include different glucosamine preparations, inadequate allocation concealment, and industry bias. Similarly, a large MA of chondroitin trials found that although analysis of all trials suggested a benefit from chondroitin, this was not seen when the analysis was restricted to methodologically robust trials of adequate sample size.
In the MA discussed here, a number of stratified analyses were conducted, for example comparing trials with adequate or unclear allocation concealment, adequate or unclear blinding of participants, etc. Analyses based on higher quality studies or studies independent of industry funding all found a smaller effect on pain than methodologically questionable or industry-funded studies. Analyses of trials which used preparations which were quality controlled, or trials where the OA was confined to the knee, found a greater effect than those without quality assurance of ingredients, or which included OA of the hip or knee. When the results were stratified by glucosamine salt, glucosamine sulphate was found to have a statistically, but not clinically significant effect (see the ‘study details’ section below).
This study has attracted media attention and a number of individuals and groups have made rapid responses to it. Some of these include personal anecdotal experiences of improvement associated with glucosamine or chondroitin. Health professionals may also have experience of patients who have apparently benefited from glucosamine or chondroitin. These can be difficult to interpret: in OA analgesics and NSAIDs have only limited effects at a population level, but individual patients may experience significant benefit. If a patient with OA tolerates glucosamine or chondroitin and appears to obtain a benefit, they may avoid or reduce their need for other therapies which may have a greater risk of causing undesirable effects. However, as the authors of this MA point out, this apparent individual benefit might be because of the natural course of OA, regression to the mean, or the placebo effect. This MA supports the decision by NICE that the best available evidence suggests that glucosamine and chondroitin do not bring sufficient clinical benefit at a population level to justify their use under the NHS.
Other criticisms of this MA include the inclusion of trials in patients with different degrees of pain at baseline, trials of different preparations, and trials of differing methodological rigour. However, none of the stratified analyses found that selecting only methodologically better studies, or studies independent of industry funding, or studies of glucosamine sulphate only, or looking at only knee OA, showed a clinically meaningful improvement. Perhaps an ongoing trial set to report in late 2011, which seems to address most of these concerns, will finally provide definitive information
Design and patients Network meta-analysis of randomised trials with an average of at least 100 patients with knee or hip osteoarthritis per arm. 12 reports of 10 trials (n=3,803) met inclusion criteria.
Intervention and comparison glucosamine, chondroitin or the combination versus placebo.
Outcomes and results
Primary outcome (absolute difference in pain scores versus placebo) over 22 months or more:
- glucosamine −0.4cm (95% credible interval [analogous to 95% confidence interval]) −0.7 to −0.1 cm)
- chondroitin −0.3 cm (95% credible interval −0.7 to 0.0cm)
- combination of glucosamine and chondroitin −0.5 cm (95% credible interval −0.9 to 0.0cm)
When the results were stratified by glucosamine salt, glucosamine sulphate was found to have a statistically, but not clinically significant effect: -0.5cm (95% credible interval -0.7 to -0.2cm). Glucosamine hydrochloride did not produce a statistically significant effect: -0.1cm (95% credible interval -0.8 to +0.7cm).
Changes in minimum radiographic joint space between baseline and the end of treatment
- glucosamine −0.2mm(95% credible interval −0.3 to 0.0 mm)
- chondroitin −0.1 mm (95% credible interval −0.3 to 0.1 mm)
- combination 0.0mm (95% credible interval −0.2 to 0.2 mm)
Withdrawals or drop-outs because of adverse events vs placebo
- glucosamine odds ratio (OR) 0.99 (0.61 to 1.50)
- chondroitin OR 0.92 (0.56 to 1.51)
- combination OR 0.90 (0.43 to 1.85)
Patients experiencing any adverse event vs placebo
- glucosamine OR 0.94 (0.59 to 1.47)
- chondroitin OR 0.99 (0.49 to 2.00)
- combination: no data
The study was funded by grants from the Swiss National Science Foundation’s National Research Program 53 on musculoskeletal health. Authors declare no competing interests.
This is an amended version of this MeReC Rapid Review. In the original version available on this website from October 21st to 25th 2010, we incorrectly stated “The Cochrane review also found a benefit from glucosamine when only studies which used a particular brand of glucosamine sulphate not widely available in the UK were included, but no benefit in studies which did not use this brand”. In fact this brand has been available from July 1st 2010 from HFA Healthcare (www.hfa-healthcare.com)
Include if appropriate
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you