Discussion about this post

User's avatar
Igor Eckert's avatar

Blinding is important as pain is very much possibly affected by patient expectation -- however, it would have been nice to see a more contextualized analysis. What reductions in WOMAC do we typically see for each 1 kg of weight loss in the literature? Is there a clear dose-response relationship? Is the relationship consistent across different weight loss interventions?

We could (should, actually) look at such data to aid our interpretation of the STEP-9 trial. If the observed magnitudes of weight loss in both placebo and semaglutide groups are consistent with their respective reductions in WOMAC in a similar pattern observed in trials of other weight loss interventions, we could have a reason to believe that the placebo effect was not sufficiently large to offset the results towards a clinically important benefit for semaglutide.

In other words: it is possible that the hypothetical elimination of the (also hypothetical) bias due to unblinding could give us a smaller effect size, but not smaller enough so that the results would be _not_ clinically important. And there's a study that may help us with it: https://pubmed.ncbi.nlm.nih.gov/33855769

On a side note: I agree with the comment from Patrick. It may be misleading to say that Semaglutide is effective for pain. Semaglutide causes weight loss, which in turn is effective for pain. "Semaglutide-induced weight loss reduces knee pain" would probably be more accurate (although not so interesting for the manufacturer).

Expand full comment
Patrick Dziedzic's avatar

Did the authors state or the study find that semaglutide, the drug, decreased the knee pain or weight loss because of the drug decreased the knee pain? I ask, because the headlines in the media that, “Semaglutide cures knee pain”, maybe a slight exaggeration, will lead people who are at a healthy weight to ask their physician for a prescription.

Expand full comment
32 more comments...

No posts