Knee pain is one of the most common reasons people with obesity seek treatment. STEP 9 is the first large randomized trial to measure what a GLP-1 does to that pain. It worked — but placebo worked too.
STEP 9 was a double-blind, placebo-controlled phase 3 trial. It enrolled 407 adults who had both obesity and clinically and radiographically confirmed knee osteoarthritis with moderate-to-severe pain. Participants were randomized 2:1 to once-weekly subcutaneous semaglutide 2.4mg or placebo, and everyone in both arms received counseling on a reduced-calorie diet and physical activity. The trial ran 68 weeks.
The population skewed heavily toward the people most affected by this problem. Mean age was 56, mean BMI was 40.3, and 81.6% of participants were women. Baseline pain was substantial — a mean WOMAC pain score of 70.9 on a 0-to-100 scale, where higher is worse.
Two co-primary endpoints. The trial measured percentage change in body weight and change in WOMAC pain score, both from baseline to week 68. Physical function was assessed separately using the SF-36 physical-function score. Running pain and weight as co-primaries is unusual, and it is what makes this trial worth reading rather than skimming.
Safety collection was deliberately narrowed. Investigators recorded serious adverse events, events leading to discontinuation, events requiring invasive knee procedures, acute pancreatitis, medication errors, COVID-19, and pregnancy-related events — not the full symptom log used in most STEP trials.
Look at the placebo column. The placebo group improved by 27.5 points on a 100-point pain scale while losing only 3.2% of body weight. That is a large effect from diet counseling, activity counseling, trial participation, and regression to the mean combined. Osteoarthritis pain fluctuates, and people enroll in trials when it is bad.
Semaglutide still separated clearly from placebo — roughly 14 additional points of pain reduction, statistically significant at P<0.001 — and it did the same on physical function. But the honest framing is that the drug roughly doubled an improvement that was already happening, rather than producing relief out of nothing.
The trial cannot tell you the mechanism. Losing 13.7% of body weight takes real mechanical load off a knee joint, and that alone would be expected to reduce pain. Whether semaglutide also acts on joint inflammation independently of weight is a reasonable hypothesis that STEP 9 was not designed to test. Do not let anyone tell you it proved an anti-inflammatory effect.
Discontinuation for adverse events ran at 6.7% on semaglutide versus 3.0% on placebo, with gastrointestinal problems the leading cause — consistent with the rest of the STEP programme and with the drug's label.
Over 68 weeks in adults with obesity and knee osteoarthritis, semaglutide 2.4mg reduced WOMAC pain by 41.7 points versus 27.5 on placebo, alongside 13.7% versus 3.2% weight loss — a real but roughly two-fold improvement over an already-large placebo response.
Sources: Bliddal H, Bays H, Czernichow S, et al. Once-Weekly Semaglutide in Persons with Obesity and Knee Osteoarthritis. New England Journal of Medicine, Oct 31 2024;391:1573-1583 (doi:10.1056/NEJMoa2403664) · Novo Nordisk STEP 9 results announcement, Oct 30 2024. NEJM full text · PubMed record.
Semaglutide is not approved to treat osteoarthritis. STEP 9 is a single trial in a specific population, and knee pain is not an indication on any semaglutide label. If joint pain is your main problem, this is evidence to discuss with a clinician, not a reason to seek a prescription on your own.
Sixty-eight weeks is the unit of measurement. The pain curve in STEP 9 improved gradually alongside the weight curve — there was no early switch-flip. Benefit of this size assumes you are still on the drug more than a year later, which means roughly 68 consecutive weekly injections without extended gaps.
Track pain the way the trial did. WOMAC is just a structured pain-and-function questionnaire scored over time. A simple weekly note on knee pain, stiffness, and what you could physically do gives your prescriber a trend line instead of a snapshot on the day of your appointment.
Expect the side-effect profile you already know. Nothing in STEP 9 suggests a different safety picture for this population. Gastrointestinal symptoms clustered around the titration schedule, and they were the main reason people stopped.
This article summarizes published clinical trial results for general education. It is not medical advice. Semaglutide is not approved for the treatment of osteoarthritis. Individual results vary, and any decision about starting, switching, adjusting, or stopping any GLP-1 medication should be made with a prescriber.