In adults with obesity and knee osteoarthritis, STEP 9 found greater improvement in knee-pain scores with semaglutide than placebo over 68 weeks. It was not a trial showing that the medicine rebuilds cartilage or replaces every other part of arthritis care.
Pain improved in both groups, with a larger change on semaglutide
STEP 9 compared weekly semaglutide with placebo in adults with obesity and clinically and radiologically diagnosed moderate knee osteoarthritis. The mean WOMAC pain-score change at week 68 was −41.7 points with semaglutide and −27.5 with placebo. Novo Nordisk funded the study.[1]
A negative change means a lower pain score. Subtracting the two reported changes gives a 14.2-point larger improvement in the semaglutide group. That is a difference in a symptom scale, not a percentage of cartilage restored and not the proportion of people who became pain-free.
| Group | Mean pain-score change | Interpretation |
|---|---|---|
| Semaglutide | −41.7 points | Lower average reported pain at follow-up. |
| Placebo | −27.5 points | Improvement also occurred with the comparison program. |
| Difference in rounded changes | 14.2 points | Additional average improvement, not a personal prediction. |
Including the placebo result prevents a misleading before-and-after story. Trial participation, background lifestyle support and the course of symptoms can all matter. Randomization helps estimate the added effect of the assigned treatment rather than crediting it with every change observed.
A knee-pain score answers a particular question
Pain is important even when it is not a structural measurement. Less pain can make ordinary activities easier and change how someone experiences their condition. But a symptom questionnaire and an image of a joint do not measure the same thing. Evidence for one should not be relabeled as evidence for the other.
This is why claims about “reversing arthritis” go further than the result summarized here. The trial headline does not establish new cartilage growth, avoidance of future joint replacement or a benefit for every cause of knee pain. An injury, inflammatory arthritis and osteoarthritis are not interchangeable diagnoses.
Participants also had obesity. The study therefore does not directly answer the same treatment question for people without obesity or for pain in another joint.[1]
Why weight and pain cannot be casually separated
The medication changed more than one outcome during the trial. That makes it tempting to attribute the entire pain benefit to a single mechanism. A consumer reading the main result should resist that shortcut. A randomized treatment comparison can establish a treatment effect without identifying exactly how much each biological or behavioral pathway contributed.
It is equally unhelpful to treat the pain result as merely cosmetic weight loss. The trial measured an outcome that matters to daily life. The stronger interpretation is that the assigned treatment improved average pain in this population; the narrower interpretation is preferable to an unsupported claim about how every participant improved.
For an individual, the next decision can include existing arthritis treatment, ability to be active, tolerability and access to follow-up. A response in a study does not require abandoning physical therapy or another established plan. It gives the treating team additional evidence to consider within that plan.
Compare providers on coordination, not an arthritis promise
A weight-management service advertising knee benefits should explain what it actually provides. Does it assess the medication decision only, or does it also communicate with the clinician managing the joint? Can existing records be reviewed? Does a change in pain trigger a clinical conversation rather than an automatic refill?
The Form Health review and PlushCare review describe different service models. Their value here is helping distinguish a focused weight program from broader primary-care access. Our provider comparison keeps those differences visible.
Do not budget as though a weight prescription will eliminate every arthritis expense. Visits, rehabilitation, imaging or other treatment may remain relevant. Include only costs the actual care plan expects to change. The first-year calculator can compare medication and program charges without assuming speculative savings elsewhere.
Pain improved more in the semaglutide group. That finding can inform a treatment discussion, but it does not establish joint repair or eliminate the need to compare the ongoing care and its cost.