By David Kiefer, MD, Editor

Dr. Kiefer reports no financial relationships relevant to this field of study.


  • Researchers randomized 604 people with mild-moderate knee osteoarthritis to 800 mg chondroitin sulfate, 400 mg celecoxib, or placebo for six months.
  • Two surveys for pain and function showed similar improvement for the two treatment groups compared to the placebo group, although in one survey the benefits from celecoxib occurred sooner than that from chondroitin.
  • Neither of the treatment groups displayed adverse effects more than the placebo group.

SYNOPSIS: Six months of 800 milligrams of pharmaceutical-grade chondroitin sulfate daily relieved knee pain as much as 400 milligrams of celecoxib.

SOURCE: Reginster JY, Dudler J, Blicharski T, Pavelka K. Pharmaceutical-grade chondroitin sulfate is as effective as celecoxib and superior to placebo in symptomatic knee osteoarthritis: The ChONdroitin versus Celecoxib versus Placebo Trial (CONCEPT). Ann Rheum Dis 2017; May 22. doi: 10.1136/annrheumdis-2016-210860.

The search continues for even partial solutions to symptomatic knee osteoarthritis. Near the top of the dietary supplement list for this is glucosamine hydrochloride (or sulfate), which is usually paired with chondroitin sulfate. Rarely is chondroitin mentioned in its own sentence. That is, until this clinical trial.

In this double-blind trial, 604 people with knee osteoarthritis were randomized to 800 mg of chondroitin sulfate (and a placebo celecoxib capsule; n = 199), 200 mg of celecoxib (and a placebo chondroitin capsule; n = 199), or two placebo capsules (n = 205) for six months. The patients were selected from several European countries, and needed to be older than 50 years of age and with medial or lateral femorotibial compartment arthritis as per the American College of Rheumatology clinical and radiologic criteria. The “target” knee was the side with a greater score on a 0 to 100 visual analog scale (VAS) for pain. Exclusion criteria were referenced from a prior trial (but not detailed in the current study), or having grade 4 knee osteoarthritis, having intra-articular injections in the last six months, taking any dietary supplements for knee pain in the last three months, taking nonsteroidal anti-inflammatory drugs (NSAIDs) in the last five days, or taking paracetamol in the 10 hours before study enrollment. The primary endpoints were patient VAS pain rating and the Lequesne Index (LI), a pain and function score from 0 to 24. Also, several secondary endpoints were recorded, as were adverse events and abnormal laboratory values. Paracetamol 500 mg tablets were allowed up to a maximum of 3 grams daily for breakthrough pain control.

The authors made a point to draw a stark contrast between pharmaceutical-grade chondroitin, and what is available as a nutraceutical, the latter being less reliable with respect to label claims and product content. They cited more favorable research for pharmaceutical-grade chondroitin, hence the reason for including the form in this research trial. They described pharmaceutical-grade chondroitin as being the 4&6 sulfate form, in no less than a 95% purity.

The researchers completed an intention-to-treat analysis on the 603 patients who actually took the study medication. The demographics between the three groups were similar at baseline, and there were not any significant differences between the dropouts in any of the three groups. For the VAS, all three groups improved compared to baseline at day 30, 91, and 182. Compared to the placebo group, the chondroitin and celecoxib groups improved more than the placebo group, but were similar to each other at six months. (See Table 1.) A similar general trend was observed for the LI score, although the celecoxib group reached significance earlier (day 30) than the chondroitin group (day 91) when compared to placebo.

All three groups had the same rate of adverse reactions, most commonly abdominal pain and discomfort. The rate of adverse effects was 2.5% in the chondroitin group, 4.5% in the celecoxib group, and 2.9% in the placebo group.

Table 1: Chondroitin vs. Celecoxib Visual Analog Scale and Lequesne Index Scores





VAS (baseline)




VAS (30 days)




VAS (91 days)




VAS (182 days)

28.6 (P = 0.001 vs. placebo)

30.5 (P = 0.009 vs. placebo)


LI (baseline)




LI (30 days)


9.1 (P = 0.045 vs. placebo)


LI (91 days)

8.1 (P = 0.050 vs. placebo)

8.0 (P = 0.027 vs. placebo)


LI (182 days)

7.1 (P = 0.023 vs. placebo)

7.0 (P = 0.015 vs. placebo)


Results from the two treatment groups and one placebo group for visual analog scale (VAS) pain testing and Lequesne Index (LI) scores for each of the time points. Note: Only the statistically significant P values are shown.


This is a favorable study for a dietary supplement for knee osteoarthritis, not always a common outcome in this field. The authors studied chondroitin sulfate, as they described, “… a sulfated glycosaminoglycan composed of chains of alternating D-glucuronic acid and N-acetyl-D-galactosamine.” But, this wasn’t just your old garden-variety chondroitin; this was pharmaceutical-grade chondroitin sulfate of at least 95% purity. When we send patients to the pharmacy shelves to purchase a dietary supplement, unless you (or they) have checked a third-party certifier’s database for accuracy of product content and label claims, our patients may or may not be receiving the necessary bang for their buck. The authors aligned themselves with past research on this pure form of chondroitin to maximize the effect on symptoms of knee osteoarthritis. It appears that their strategy worked.

This study was focused on symptomatic improvement, as documented by the VAS and LI scoring systems, the latter involving both pain and function parameters. The study was considered too short term to capture structural change, so the decision was made not to analyze for bone nor cartilage markers. Therefore, we don’t know whether chondroitin is “building up cartilage” or simply serving as an anti-inflammatory, two of the proposed mechanisms of action. At the end of the day, it may or may not matter, as long as our patients notice some benefit to their symptoms. Does chondroitin sulfate lead to symptomatic improvement immediately, as our NSAIDs of choice might? It is unlikely. In fact, the pharmaceutical in this study, celecoxib, led to benefits in the LI scale a full two months before chondroitin. Not unlike other subtly acting dietary supplements, when we nudge our patients down the chondroitin trail, we encourage them to commit to that treatment in the medium-to-long term, not expecting short-term benefit.

Perhaps this study is the beginning of clinicians considering chondroitin in its own right. Many prior studies examined chondroitin in combination with its more well-known “cousin,” glucosamine. Now that we know that a very pure form of chondroitin sulfate also may be efficacious, which one do we choose? Perhaps both. Perhaps what our patient’s budget would allow. Perhaps what will be safe with their pharmaceutical list after we consult our resources for supplement-pharmaceutical interactions (which we always should do). Such personalized decision-making should lead, hopefully, to optimal outcomes. And for our patients with symptomatic knee osteoarthritis, there is little reason not to steer them toward a pharmaceutical-grade chondroitin sulfate for at least a three-month trial.