I have to confess that a few years ago I was advising patients to go and get glucosamine for their dodgy knees. There can't be many GPs that haven't dropped it into the discussion with a patient with intractable osteoarthritic pain. The therapeutic options for worsening osteoarthritis seem to dwindle rapidly. Unfortunately, the evidence for any effect from glucosamine has been eroded nearly as quickly as the joints involved.

My fondness for glucosamine had been partly motivated by the fact Mrs ND is a vet and she advised me of the potential benefits of glucosamine for arthritis. She was adamant that a glucosamine-chondroitin supplement was good for dogs with arthritis and she gave testimony to having witnessed some near miraculous effects. Apparently this combo could perform wonders for portly labradors, or 'coffee tables' as vets occasionally refer to them, with knackered knees.

The GAIT trial was reported a year ago by NCCAM and Dr*T blogged about glucosamine way back last October. I know that NCCAM is not hugely popular amongst anti-quackers in the USA but there are some helpful Q&A here on their $12.5 million GAIT trial.

What were the key results of the study?

Researchers found that:

Participants taking the positive control, celecoxib, experienced statistically significant pain relief versus placebo-about 70 percent of those taking celecoxib had a 20 percent or greater reduction in pain versus about 60 percent for placebo.

Overall, there were no significant differences between the other treatments tested and placebo.

For a subset of participants with moderate-to-severe pain, glucosamine combined with chondroitin sulfate provided statistically significant pain relief compared with placebo-about 79 percent had a 20 percent or greater reduction in pain versus about 54 percent for placebo. According to the researchers, because of the small size of this subgroup these findings should be considered preliminary and need to be confirmed in further studies.

For participants in the mild pain subset, glucosamine and chondroitin sulfate together or alone did not provide statistically significant pain relief.

Essentially this huge and very expensive study involving around 1800 patients showed no benefits of glucosamine other than a small effect in a sub-group. If you go raking through all the subgroups in a large trial it is an inevitability of the way statistics are analysed that eventually you will turn up a significant result. Torture the data and it will tell you anything.

The possibility remains that perhaps glucosamine and chondroitin supplements are an effective treatment to reduce pain in arthritis in dogs. This could be an effect which is confined to canines and can't be replicated in humans.

Or maybe there is a subtle placebo effect at work manifesting itself through the owners. It's an intriguing thought. However, the most likely explanation is that the advice that vets give the owners when clients consult about a dog with dodgy knees is having the effect. Vets will give advice on lifestyle changes for the animal and are likely to give non-steroidal analgesia. It is this that makes all the difference - not the glucosamine and chondroitin supplements.

I've had a dig around in the vet literature and I found one systematic review of treatments for osteoarthritis in dogs in the Journal of the American Veterinary Medical Association. Incidentally, in the abstract when they talk about 'comfort' they are referring to the reliability of the evidence - not how little pain the animals were in. This isn't clear until one reads the paper. The trial with glucosamine and chondroitin showed no effect above placebo. They commented:

Presently, the strongest evidence available for the medical treatment of clinical signs associated with osteoarthritis in dogs is mostly limited to nonsteroidal anti-inflammatory drugs. Additional controlled studies in all groups of medical treatments are needed.

There is possibly an argument that dogs will present later than their human counterparts. Dogs won't turn up at the surgery whining at the first twinges in their knees. Perhaps vets are more likely to be treating severe osteoarthritis in dogs and this might fit with the GAIT trial showing an effect in their subset. Perhaps. But I'm grasping at straws.

However, there was another strand to GAIT which showed that there was no change in structural changes on X-ray. They went on to cast doubt on these findings as the overall changes in joint narrowing were smaller than expected. This week the American College of Rheumatology Annual Scientific Meeting presented an abstract at their conference. This was the Joints on Glucosamine (JOG) Study: A Randomized, Double-Blind Placebo-Controlled Trial to Assess the Structural Benefit of Glucosamine in Knee Osteoarthritis based on 3T MRI study. After 6 months of treatment they found no structural changes using MRI or any evidence of new cartilage formation when measuring a urinary biomarker of cartilage synthesis. It seems likely to be an important trial with negative findings confirming the GAIT findings.

Glucosamine limps on in high street stores but it seems to be running out of steam. One of the GP practices in which I locum has taken the decision to stop prescribing glucosamine. I've not been involved in the discussions but I suspect they have taken the view that it is a fundamentally an unremarkable food supplement with a miserable evidence base.

It would have been great to have identified a cheap, well-tolerated remedy to reduce the pain of osteoarthritis. Sadly, wishful thinking isn't enough and for those with knackered knees it remains a dog's life.

This post is also published at Northern Doctor's Antidote.