Review of Controversial BMJ (British Medical Journal) article on Glucosamine and Chondroitin

Jason Theodosakis, M.D.

(Draft version- more to follow)

 

This is the eighth meta-analysis or review article on glucosamine and/or chondroitin. Not sure why it got so much press.

A group of (mostly) Swiss researchers seemed to set out with good intent, to answer the question of whether or not glucosamine and/or chondroitin relieves pain and slows the loss of cartilage. No new patients were studied, this is just a gathering of some (10 out of 58 studies) of the previously published studies on these supplements. The study looks impressive and has enough medical and scientific jargon to seem legitimate. The trouble is - the study is not legitimate, the researchers made some cardinal mistakes not allowed in a meta-analysis, which I will explain, below. It's such a bizarre interpretation, that it appears to be very suspicious. Perhaps there are underlying political motivations to try and discourage the British government from covering the cost of these supplements, especially since the supplements (sold as drugs) are covered in other EU countries. The result is an improper conclusion and improper damage to the supplement's reputation. Glucosamine and Chondroitin remain as the oral therapies with the best risk/benefit ratio of any treatments for osteoarthritis. {You'll enjoy the challenge I propose to those whose opinion differs}.

 

-----------------------------------------------------------

If you follow the same reasoning the authors used in this review article, and applied it to runners, you could come up with the following conclusion:

Usain Bolt is no better than the average track and field runner. Just take some of his times in 100 meter sprints, and some of his times in running longer races, like 10Ks and Marathons. Average his speed (time/distance) for all of these races. See, he's pretty slow! This is essentially what the foolish authors did for the glucosamine/chondroitin review article. They somehow made the decision to combine short and long-term information (with different outcomes, mind you) and then said the outcomes were no better than placebo. Sorry- this is nonsense.

-----------------------------------------------------------

 

What did the researchers do wrong?

A meta-analysis is supposed to allow one to combine (or pool) the data from several small studies (of the same type) to create one bigger study. When combining study results, you can't just average the results of studies designed to answer different questions. Just like you can't mix apples and oranges, you can't just pick and choose any studies to do this, or the results will be useless - you could come up with the opposite conclusion to the true nature of the treatments in question.

Studies combined for a Meta-analysis have to be of the same type:

- Same types of people (similar proportion of age, race, sex, disease severity, etc)
- Studying the same area (knee, back, hip)
- Utilizing the same form and dosage of treatment
- Performed for the same duration
- Looking at the same outcome measures

The researchers FAILED in each of these areas. They just seemed to pick and choose a fraction of the 58 studies on these supplements, and hand pick their own analysis, to create their own conclusion.

Imagine if you said that Tiger Woods was a terrible golfer because you averaged 10 of his worst golf scores from the past 58 tournaments. The BMJ study is even worse than this. It used studies that are invalid, and thus should not be part of any evaluation.

The biggest and most obvious flaw in the BMJ review is that they mixed togther studies designed to evaluate pain outcomes with those designed to look at x-ray changes in cartilage (a surrogate marker for cartilage loss). This easily explains the false negative results they obtained for their review. Long-term studies are required to look for changes on x-ray indicative of cartilage loss. These studies must use subjects who have very little pain to start, ortherwise, the subjects on placebo would drop out (who could stay in pain for 3 years?), or muddy the results by taking (or sneaking) pain pills of some type. It's no wonder that there's very little difference, if any between the mild starting pain of people in a 2 or 3-year study and the mild pain (if any) of those who complete the study. The BMJ reseachers used the pain data from these long-term structure sutdy and averaged it with the pain data from short-term studies. This gives us results which are completely invalid and actually nonsense.

 

How we Know Glucosamine and Chondroitin "Work"

"Work" can mean a lot of different things. Each one of these is important:

- Pain relief (compared to placebo, or compared to known, FDA-approved drugs such as acetaminophen or anti-inflammatory drugs)

- Slowing of the disease process (e.g. loss of cartilage or reduction in bone spurs) without regard to any effect on pain

- Reduction in the need for drugs such as acetaminophen, anti-inflammatories, or opiates

- Reduction in the need for physical therapy or surgery

- Saving overall health care costs

- Improved safety over existing treatments

 

Pain relief (compared to placebo, or compared to known, FDA-approved drugs such as acetaminophen or anti-inflammatory drugs)

 There is only one published study comparing glucosamine to acetaminophen. It's called GUIDE (Arthritis Rheum. 2007 Feb;56(2):555-67). Very well designed, it met the European guidelines for doing studies on osteoarthritis treatments. In this 6 month study, 1,500 mg of glucosamine was compared to 3,000 mg of acetaminophen and placebo (3 arms). Results were analyzed for the two most validated measures of osteoarthritis pain and function (WOMAC and Lesquene Index). In both outcomes, glucosamine users had statistically significant improvement. Acetaminophen, despite its high dose, was not significant from placebo.

 

Slowing of the disease process (e.g. loss of cartilage or reduction in bone spurs) without regard to any effect on pain

 

Reduction in the need for drugs such as acetaminophen, anti-inflammatories, or opiates

 

Reduction in the need for physical therapy or surgery

 

Saving overall health care costs

 

Improved safety over existing treatments

 

 

The BMJ completely contradicts previous reviews or Meta-analyses that followed the world-wide rules for performing such reviews:

The latest, and most credible review for Chondroitin and structure analysis is shown here:
The analysis showed, with an almost unheard of 99.99% certainty, that 800 mg of high quality chondroitin each day slowed the loss of cartilage in people with osteoarthritis

The most appropriate analysis on Glucosamine is to look at a review of the two long-term studies with the same design: