Ozempic and Knee Pain: What the 2024 NEJM Study Actually Shows

Why Everyone Is Asking About Ozempic and Knee Pain

"The question worth answering is not whether the drug works. It is what the drug is actually doing to the knee."

Knee osteoarthritis is one of the most common causes of chronic joint pain in adults, and body weight is one of its biggest modifiable drivers. So when a wave of patients on GLP-1 weight-loss medications started saying their knees hurt less, researchers had a clear question to test. The STEP 9 trial, published in the New England Journal of Medicine in 2024, was built to answer it directly (Bliddal and colleagues, 2024).

Here is why this matters for anyone searching "ozempic knee pain" at 11pm with an aching joint. The internet is full of drug-company press releases and reblogged headlines, and almost none of them explain the biology in plain terms or tell you what the study cannot prove. I am a chiropractor. I do not prescribe semaglutide, I do not sell it, and I have no financial stake in whether you ever take it. That is exactly why I can give you a straight read of the evidence.

The short version: the study is legitimate and the result is real, but the headline most outlets missed is the mechanical one. The knee felt better largely because the load on it went down. That principle applies whether the weight comes off through a medication, a diet, or a barbell.

The Study: What STEP 9 Actually Measured

Let me lay out the design so you can judge the weight it carries. STEP 9 enrolled 407 adults who had obesity and a clinical and radiographic diagnosis of knee osteoarthritis, specifically Kellgren-Lawrence grade 2 or 3, which is moderate structural change on X-ray. Participants were randomly assigned in a 2 to 1 ratio to receive either once-weekly injectable semaglutide at 2.4 milligrams or a matching placebo, and both groups also got lifestyle counseling. The trial ran for 68 weeks (Bliddal and colleagues, 2024).

"A randomized, placebo-controlled, double-blind trial is about the strongest single-study evidence a question like this can get."

That design matters. Randomized means chance decided who got the drug, so the groups start out comparable. Double-blind means neither the patients nor the researchers knew who got what, which strips out the placebo-driven bias that plagues pain research. Placebo-controlled means the comparison is against an inactive injection, not against nothing. This is a well-built trial, and it is the reason its result deserves to be taken seriously rather than dismissed as hype.

The two co-primary outcomes were the percentage change in body weight and the change in knee osteoarthritis pain as measured by the WOMAC pain score. WOMAC is a validated questionnaire that scores knee pain from 0 to 100, where a higher number means worse pain. A key secondary outcome was physical function, measured with the SF-36 questionnaire. So the trial was asking two linked questions at once: does the drug take weight off, and does the knee feel and work better when it does.

What STEP 9 Found

"The knees improved on placebo too. They just improved a lot more when the weight came off."

The weight result was large. People on semaglutide lost an average of 13.7 percent of their body weight over the 68 weeks, compared with 3.2 percent for those on placebo (Bliddal and colleagues, 2024). For a person starting at 250 pounds, 13.7 percent is roughly 34 pounds. That is not a trivial change, and it is squarely in the range where the knee-load research predicts a real benefit.

The knee result tracked the weight result. WOMAC pain scores improved by 41.7 points in the semaglutide group and by 27.5 points in the placebo group. Notice that both groups got better, which is normal in a supported trial where everyone is getting lifestyle attention. The point is that the semaglutide group improved by about 14 points more. Physical function, measured on the SF-36, also improved more with semaglutide, a mean gain of about 12 points versus 6.5 on placebo.

One number to keep in the front of your mind: semaglutide is not approved to treat knee osteoarthritis. STEP 9 studied people who had both obesity and knee arthritis, and the trial itself frames the finding as the effect of weight reduction on the joint. It is evidence about what happens to a knee when a person loses substantial weight. It is not a recommendation to take the drug for a knee problem, and I am not making one.

Does Losing Weight Actually Help Knee Pain?

This is the part the drug headlines skip, and it is the part that actually helps you. The knee is a load-bearing joint. When you walk, the compressive force passing through it is several times your body weight because of how the muscles and levers around the joint work. That force multiplier is why weight matters so much at the knee specifically.

"Lose one pound and the knee feels about four pounds lighter with every step you take."

Put a number on it. A foundational 2005 study found that each single pound of body weight lost was associated with roughly a fourfold reduction in the compressive load through the knee per step (Messier and colleagues, 2005). Multiply that across the thousands of steps you take in a day, and a modest weight loss becomes an enormous reduction in cumulative joint stress over weeks and months. That is the engine underneath the STEP 9 result. Take weight off the frame and you take load off the joint.

How much do you need to lose before the knee notices? A 2024 systematic review and network meta-analysis of weight-loss trials in knee osteoarthritis found that significant pain relief can be anticipated once a person achieves at least about a 7 percent reduction in body weight (Shahid and colleagues, 2024). And the well-known IDEA trial showed that a weight loss of 10 percent or more, achieved through diet plus exercise, produced greater reductions in pain, better function, and measurably lower knee compressive loads than exercise alone (Messier and colleagues, 2013). The STEP 9 group cleared both of those thresholds comfortably at 13.7 percent.

What This Means Clinically

Here is how I think about a painful knee when someone walks into my clinic. There are two separate things going on. One is the structural change inside the joint, the cartilage thinning and the bony change you see on an X-ray. I cannot reverse that, and neither can anyone else with a conservative treatment, and any chiropractor who tells you otherwise is selling something. The other thing is everything around that structure: how much load passes through it, how well the joint glides, how strong and balanced the muscles that support it are. That second bucket is where real improvement lives, and it is a big bucket.

"You cannot un-wear the cartilage. You can change almost everything around it, and that is usually enough to hurt less and move more."

Weight management is the load lever, and STEP 9 is a vivid reminder of how powerful it is. Whether a patient gets there through a medication their physician prescribes, through diet, or through building an exercise habit, the payoff at the knee is the same reduction in mechanical stress. My job is not to weigh in on the drug. My job is to help with the movement side while that load work happens, and to be honest that the two work best together.

The knee is one of well over a hundred conditions I see, and it rarely travels alone. A stiff hip, a locked-up lower back, or a guarded gait all pile extra work onto a sore knee. Addressing the whole chain, not just the loudest joint, is the point of a full evaluation. You can see the range of what conservative care covers on our guide to the conditions we treat, and the knee sits inside that larger picture of how the body loads and moves.

There is also a bigger metabolic thread worth naming honestly. Excess weight, high blood sugar, and musculoskeletal pain tend to cluster together, which is one reason weight-focused treatments ripple out to the joints. I dug into the metabolic side of that overlap in a separate commentary on what a 2026 meta-analysis shows about diabetes and spine pain. The knee and the spine are different joints, but the lesson rhymes: metabolic health and mechanical pain are connected, and the honest fix works on both.

What Chiropractic Can and Cannot Do for an Arthritic Knee

Let me be plain about the boundary, because honesty is the whole point of this review. Chiropractic care does not regrow cartilage, does not cure osteoarthritis, and has nothing to do with prescribing or managing GLP-1 medications. If a patient asks me whether they should go on Ozempic, my answer is that it is a conversation for them and their physician, full stop. I report the evidence here. I do not steer that decision.

"The strongest thing conservative care offers an arthritic knee is this: it keeps you moving while the load comes down."

What conservative care can honestly offer is meaningful. Hands-on work can improve how the knee and the joints above and below it move, so the knee is not fighting a stiff hip or a guarded ankle. Targeted, progressive strengthening of the muscles that support the knee changes how force is shared across the joint. And getting pain down to a tolerable level is not a small thing, because pain is what stops people from walking, and walking is exactly the activity that both manages weight and feeds the joint. The goal is a loop that turns in the right direction: less pain, more movement, better load management, less pain again.

None of that is a cure, and I will not dress it up as one. For a knee with real structural arthritis, the realistic aim is better function and less pain, not a new joint. When the pain is a recurring mechanical pattern rather than a one-off flare, ongoing periodic care can help hold the gains, and I am straight about the time course involved. It is always a flat 40 dollars a visit, no contract, and you stop whenever you want. The medication question stays with your doctor. The movement and load question is where I can help.