Glucosamine sulphate and osteoarthritis

Jean Yves Reginster and colleagues (Jan 27, p 251)

claim to have shown long-term combined structure-modifying and symptom-modifying effects of glucosamine sulphate on osteoarthritis. However, as we understand their findings, they show no clear correlation between the modifying effects on structure and symptoms of glucosamine, since patients with relief of symptoms are not identical to those experiencing a structural improvement.

As Tim McAlindon states in the Commentary (Jan 27, p 247),

the clinical relevance of the primary outcome measure for structure modification–ie, joint-space narrowing—is not clear. In our opinion, the difference in favour of glucosamine becomes even more doubtful given themean difference of 0·24 mm after 3 years and, even more importantly, its lower 95% confidence limit of 0·01 mm (intention-to-treat analysis) or 0·02 mm (per-protocol analysis).

Symptom assessment is reported only at baseline and at 3 years. Indeed, symptoms were scored every 4 months, as reported at the meeting of the American College of Rheumatology in Boston 1999.

Thus, crucial data for the assessment of glucosamine’s clinical efficacy in the course of this chronic disease are withheld. Furthermore, meaningful baseline imbalances between treatment groups are neglected. The total Western Ontario and McMaster Universities (WOMAC) index, as well as the subscales for pain and function are about 10% lower in the placebo group than in the glucosamine group. Absolute score indices at study end are not presented. Hence, the impact of this potential bias cannot be assessed.

Moreover, table 3 and figure 2 apparently display discordant study results: the WOMAC score is reported as worsening by 10% in the placebo group in table 3 as well as in the text, but figure 2 shows improvement in the subscales for pain and function in this group by 5% and 10%, respectively. Those two subscales, however, account for 90% of the total WOMAC score. Finally, calendar dates are missing for the whole study period. Overall, from our point of view, the reported data are scarce, incomplete, and inconsistent.