Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Glucosamine and arthritis update

Towheed et al, 2001 [1]
McAlindon et al, 2000 [2]
Long-term randomised trial [3]

As long ago as Bandolier 46 we performed a swift systematic review of glucosamine for arthritis. Our conclusion, after reviewing eight randomised trials, was that there existed a body of evidence supporting the use of glucosamine in arthritis. Since that time there have been two more systematic reviews published, and one superb three year study, all of which support the original conclusion. Now is a suitable time to update our knowledge, in part because the efficacy of glucosamine remains one of those questions asked most often by professionals and by the public.


There are a few points that need to be borne in mind when examining trials of glucosamine for arthritis. First is the problem that there is no international pharmaceutical standard for glucosamine (or at least none that we know of). Glucosamine is usually in the form of the sulphate, but this formulation will contain other salts and some water of crystallisation, the actual amount of glucosamine may vary from preparation to preparation, stability may be a problem, and glucosamine may be in combination with chondroitin.

Doses used also vary, as do the number of times a day oral tablets are taken. Usual oral doses are about 1 to 1.5 grams a day. Though oral administration is most common, it has also been given by injection into a joint, and by intramuscular or intravenous injection. There is no substantial evidence that this makes any difference.

Outcomes used in trials vary enormously, as is the norm in arthritis trials, especially older ones. This inhibits pooling of data in meaningful ways. Trials are often short, again as has been the case for older trials in arthritis.

Finally there is the control. Frequently this is placebo, but some studies compared glucosamine with an NSAID.

Despite these problems, two systematic reviews have been able to come to conclusions about the evidence.

Towheed et al, 2001 [1]

This Cochrane review was notable because the authors found a number of unpublished studies. In all they identified 16 randomised double-blind studies, with 992 patients randomised to glucosamine and 1037 randomised to placebo or NSAID. The mean age of patients was 61 years. Twelve of the trials could be included in the review. All had quality scores of 3 or greater out of 5, indicating that substantial bias was unlikely.

A number of analyses of efficacy were undertaken, depending on outcome or comparator. The broad conclusion was that glucosamine had a clinically significant benefit compared with placebo based on standardised mean differences. In the review this was 1.4, and standardised mean differences (or effect sizes) of 0.8 or greater are defined as large.

The review also looked at adverse events. It found only 14 patients treated with glucosamine withdrew from treatment because of suspected toxicity, with only 61 reporting any adverse effects. Mean trial duration was only six weeks, so long-term safety might be an issue. Otherwise they concluded that glucosamine was safe and effective in treating osteoarthritis.

McAlindon et al, 2000 [2]

This review sought randomised trials of glucosamine and chondroitin in the treatment of osteoarthritis. In the analysis it pooled outcomes reported as the primary outcomes by authors of the original papers, and pooled pre-defined outcomes at four weeks, arguing that outcomes before that time may be spurious in osteoarthritis. It also pooled oral and intramuscular or intra-articular administration.

Using effect size, it found moderate to large efficacy for glucosamine (effect size 0.4) and chondroitin (effect size 1.0) with the outcomes used by authors of the original papers. Using their own hierarchy of outcomes at four weeks, effect sizes were modest (glucosamine 0.3, chondroitin 0.4). Large trials had smaller effect sizes than small trials.

Despite some concern over study quality, and of publication bias (probably misconceived as it used funnel plots, see Bandolier 81 ) that might exaggerate efficacy, it reached an overall positive conclusion about both glucosamine and chondroitin.

Long-term randomised trial [3]

A three-year study comparing glucosamine with placebo has just reported [3]. Patients aged over 50 years and with primary knee osteoarthritis were randomised to 1500 mg oral glucosamine sulphate once daily or placebo. The mean age was 66 years, and the mean duration of their osteoarthritis was eight years.

The primary outcome was the mean joint space width of the medial compartment of the tibiofemoral joint, with X-rays taken with patients standing, and using a validated measuring system using digitised images. Pain, functioning, stiffness and consumption of analgesics were also measured, and measurements were taken at baseline, and one and three years. Two hundred and twelve patients were randomised, and 71/106 on placebo completed three years and 68/106 with glucosamine.

The average joint space width was about 5.4 mm at baseline. With placebo there was a mean narrowing of 0.3 mm over three years. With glucosamine there was no narrowing. After three years, 32/106 patients (30%) on placebo had a significant joint space narrowing of more than 0.5 mm, compared with 15/106 patients (15%) with glucosamine. The relative risk of significant joint space narrowing with glucosamine was 0.5 (0.3 to 0.9), and the number of patients needed to be treated for three years to prevent one patient having significant joint space narrowing compared with placebo was 6.6 (3.8 to 25).

With placebo there was no overall change in pain or functioning. With glucosamine there was a significant improvement of 20-25%. Stiffness was not affected, and the consumption of analgesics or NSAIDs was not different. Patients used rescue medicines on average once every six days.

Adverse events were reported by 95% of patients over the three years. Most were transient and mild, not clearly related to treatment, and there were no differences between glucosamine and placebo. Adverse event withdrawals occurred in 21/106 patient on glucosamine and 18/106 on placebo (relative risk 1.2; 95% confidence interval 0.7 to 2.1).


Evidence that glucosamine (and chondroitin) is effective in osteoarthritis continues to build. We now have two top class reviews of older, short, studies that come to this conclusion, and a new randomised trial of some quality that demonstrates a clear disease-modifying effect, as well as showing improvements in pain and functioning and an absence of long-term harm. Added to this is the accumulating volume of anecdotal evidence from professionals who prescribe glucosamine with good effect, and of individual who use it and report the same good effects.

We might still argue that we are lacking a large randomised outcome study using a standardised glucosamine preparation, and independent of manufacturers. That is on the way. The National Institutes of Health in the USA is funding considerable research in the complementary therapy area, and glucosamine is close to, if not at the top of the list.

One practical point that emerges from several studies is that glucosamine takes about a month to exert its full effects. Another is that we now have some evidence that chondroitin is also effective. A third is evidence on formulation and stability is notable by its absence. There is no help, as yet, in choosing which of the many different preparations available is best (if any). The cheapest might be a good place to start. Glucosamine is not black-listed in the UK, but checking with health authority prescribing advisers regarding the status of reimbursement is advisable.


  1. TE Towheed et al. Glucosamine therapy for treating osteoarthritis (Cochrane Review). In: The Cochrane Library, Issue 1, 2001. Oxford: Update Software.
  2. TE McAlindon et al. Glucosamine and chondroitin for treatment of osteoarthritis. A systematic quality assessment and meta-analysis. JAMA 2000 283: 1469-1473.
  3. JY Reginster et al. Long-term effects of glucosamine sulphate on osteoarthritis progression: a randomised, placebo-controlled clinical trial. Lancet 2001 357: 251-256.
previous or next story in this issue