I have to chime in here. The full text is paywalled but if you have a subscription or institutional access the devil is in the details.

The "pilot study" the authors performed was an n=20 with 30% dropout, meaning 14 finished the study, 6/10 in the non-semaglutide arm and 8/10 in the semaglutide arm. They did not include participants who dropped out in their final analyses which is a big methodological flaw. I could go on about their choice in and measurement of outcomes, but suffice to say with only 14 participants these results should be taken with a grain of salt.

Further, they did extensive molecular biological studies in mice. There have been MANY (dozens) of murine (mouse) studies showing beneficial effects of various compounds / interventions on mouse knee osteoarthritis. None of them have translated into a therapy for the human disease. Mice live 2-3 years and have very different knee biomechanics than humans.

Appreciate the authors investigative efforts, but more confirmatory studies are needed before I'll be injecting semaglutide into my knee.