Glucosamine and Chondroitin for Knee Arthritis - Still No Evidence of Effectiveness

In the most recent edition of our book Worst Pills, Best Pills, we listed the dietary supplements glucosamine and chondroitin as Do Not Use, primarily because most of the studies testing them were poorly designed (e.g., too small, too short, too unstandardized in their diagnoses of arthritis). In better-designed studies, the supplements generally proved ineffective. However, we noted that the National Institutes of Health was conducting a major randomized trial that might resolve the controversy once and for all. The results are now in.

In that study, published in the New England Journal of Medicine (Feb. 23, 2006, pp. 795-808), researchers randomly assigned patients with osteoarthritis (the most common form of arthritis in older persons) of the knee to glucosamine, chondroitin, both, neither (an inactive placebo) or celecoxib (Celebrex). Patients were followed for 24 weeks and the impact upon the severity of their knee pain was measured.

The segment of patients who experienced a 20 percent decrease in their pain scores was 60 percent in the placebo group, 64 percent in the glucosamine group (i.e., only 4 percent better than placebo), 65 percent in the chondroitin group, 67 percent in the combined glucosamine/chondroitin group and 70 percent in the Celebrex group. Only Celebrex proved statistically superior to placebo. (Because of its association with heart attacks and strokes, we list Celebrex as a Do Not Use drug and instead recommend generic ibuprofen or aspirin, which are not toxic to the heart.)

While that should be the death knell for these supplements, some still find reason to believe. Supplement defenders point out that glucosamine hydrochloride was used in the study while most supplements on the market contain glucosamine sulfate, and so the unfavorable study findings might only apply to the hydrochloride. There is no evidence for this proposition. Proponents had long argued that it was glucosamine itself, a component of cartilage, that would be effective. But, regardless of the form of the chemicals, both glucosamine and chondroitin are so heavily metabolized after being eaten that essentially no drug even makes it to the joints. The reason the researchers used the hydrochloride form was that the sulfated forms on the market were not sufficiently pure or consistently manufactured.

The other potential lifeline for the supplements was the finding that, among the minority of patients with more severe pain, the combination of glucosamine and chondroitin was effective, although the individual components and Celebrex were not. This was an odd finding because Celebrex was effective in the full study. However, this analysis appears to have been what researchers call post hoc: it was not planned before the study. Such analyses are notoriously difficult to interpret and often lead to findings that are subsequently debunked.

The bottom line is this: in the most rigorous study to date, there was no evidence of effectiveness for glucosamine, chondroitin or the two together. We continue to recommend against the use of these unregulated supplements.