Doc defends pain supplements, slams study

By LAURA GILCREST, UPI Health Business Editor

WASHINGTON, Feb. 23 (UPI) -- The latest government study of a popular dietary supplement combination glucosamine and chondroitin - taken by millions of Americans to relieve osteoarthritis knee pain -- which showed no overall benefit in reducing pain, compared to Celebrex and placebo, was fundamentally flawed, and obscured the fact that the supplements do benefit patients with moderate to severe pain, a medical expert said Thursday.

The GAIT (Glucosaminechondroitin Arthritis Intervention Trial), showed that, overall, there was no significant pain reduction in patients taking the supplements alone or in combination, compared to patients taking Pfizer's COX 2 inhibitor Celebrex and those taking placebo. The six-month research was sponsored by the National Institutes of Health and is the first of two parts, with a second X-ray study still pending.


The findings were published this week in the New England Journal of Medicine, along with an editorial that questioned the supplements' viability as a treatment option for osteoarthritis patients.


"On the basis of the results from GAIT, it seems prudent to tell our patients with symptomatic osteoarthritis of the knee that neither glucosamine hydrochloride nor chondroitin sulfate alone has been shown to be more efficacious than placebo for the treatment of knee pain," the editorial said.

The author added, "Furthermore, there is no evidence that these agents prevent osteoarthritis in healthy persons or in persons with knew pain but normal radiographs."

The article also noted that patients with severe knee pain on glucosamine sulfate or chondroitin sulfate could become addicted and that patients should stop taking the supplements if there's no improvement after three months.

But Jason Theodosakis, an arthritis and sports medicine specialist and author of the book, "The Arthritis Cure," attacked the study and the NEJM editorial's conclusions.

In an interview with United Press International, he also suggested that medical journals like NEJM are beholden to drug companies -- whose drug advertising helps support their operations - and thus are more inclined to view studies in the companies' favor.

"The (GAIT study) abstract, when this was first presented, was very positive, and the New England Journal turned it more negative. My concern is that those guys hate (dietary) supplements of any kind," Theodosakis said.


"(NEJM) has no ads for supplements, they have ads for Celebrex and Vioxx. I'm very concerned about their bias because all of those people work for the drug companies," said Theodosakis, who said that he does consulting work for drug companies and firms making dietary supplements, but that he has no financial ties to the maker of the particular supplements used in the GAIT study.

UPI did not receive NEJM's prepared statement in response by press time, but a spokesperson for the journal quoted NEJM's Editor-In-Chief Jeffrey Drazen that "If this person has concerns (about the GAIT study), he should submit them in a letter to the editor."

The NEJM spokesperson explained that the journal reviews letters to the editor and if "legitimate concerns" about a particular study are raised, it publishes the letter and the study's authors are given an opportunity to respond.

Regarding the GAIT study, Theodosakis charged that there were multiple flaws in its design, so the findings obscured the fact that, in a single subgroup of patients with moderate to severe pain who were taking the supplements, 79 percent experienced significant pain relief and improved function, compared to Celebrex and placebo, he told UPI.


"That's a huge finding. In patients with high pain, the supplements beat everything," said Theodosakis, who is also a member of the steering oversight committee for the GAIT study.

But despite the study's design problems, he said, "it still had a tremendously significant -- highly statistically significant-- finding for the combination glucosamine/chondroitin."

What's more, the study must be put in context with numerous other trials of the supplement duo, Theodosakis said.

The NIH study is "just another" of 30 human clinical trials done on the supplements, he pointed out. "It's not like this NIH study is the answer, yes or no, it's just part of the body of literature."

The GAIT study used 14 outcome measures and separated patients into different groups according to pain severity, as is normally done in such trials. However, "unfortunately, unlike in most studies, 78 percent of the subjects only had moderate to mild pain," Theodosakis noted. "People with more pain have more room for improvement, so when you have a study with patients who are mildly affected, you don't see a statistically significant difference."

Another problem with the research, he added, was what Theodosakis called "diluted data." Specifically, Celebrex did no better than placebo in relieving knee pain in 12 out of the study's 14 outcome measures, he noted.


"Celebrex was being used as a positive control, and it wasn't positive in most of the outcomes. That's why we know that the results were diluted and people need to read into the results carefully. Because the comparator failed most of the time, you have to be cautious about your conclusions about the supplements. The most important finding is (the supplements) beat Celebrex in patients who had moderate to severe pain," Theodosakis said.

"That shows us that this study understated the results of all of the treatments. That was a bellwether for the study design," he said.

Finally, the study had a "very high placebo rate," so that the supplements' overall pain reduction rate of 68 percent -- which would normally look impressive -- isn't considered significant when compared to a placebo rate of 60 percent, he explained.

While a placebo effect of about 35 percent is presumed in most studies, the GAIT study had a placebo rate of 60 percent, partly because the study patients were checked too frequently for treatment compliance. "The more interaction you have (with patients), the more likely you are to get a placebo effect," Theodosakis said. Second, because there were so many treatment arms in the study (four treatment groups, and one placebo group) patients "knew they had an 80 percent chance of being randomized to a treatment group. They had high expectations," he explained, "so these were design flaws."


He said the optimal use of the supplement combination for osteoarthritis knee pain might be to prescribe them in combination with short-term use of a drug like Celebrex, as is typically done in Europe.

Some patients might still need to take a drug intermittently, however, for pain flare-ups, he added.

Theodosakis noted that, in a three-year study of the supplement duo, where patients were followed for eight years, the need for joint replacement was cut by 75 percent. What's more, the products have not been linked to a single death in 15 years, based on European data, he said.

Theodosakis also criticized the media for its coverage of the GAIT study. "I think they just read the conclusion, they don't read the results," he said. "Why did Celebrex fail 12 out of 14 times (in the GAIT study)? That's incongruent with all of the other Celebrex studies. That should raise a huge red flag that this is an inconclusive study. That should be the headline."

Latest Headlines