There is a reasonable chance someone has told you that a weight-loss injection helped their knee. There is an equally reasonable chance someone else has told you it is a fad. Both conversations are happening because the evidence has genuinely shifted over the last two years, and because that evidence is being stretched a long way beyond what it actually supports. Here is where things really stand, and where a medicine like this sits alongside the things that do most of the work in osteoarthritis: loading, strength, weight, and occasionally a well-placed injection.
If you have knee osteoarthritis alongside significant obesity or type 2 diabetes, a licensed weight-management medicine may be a reasonable part of your plan, inside a proper rehabilitation programme, not instead of one. If you are hoping for a tiny "microdose" that treats arthritis directly without changing anything else, that evidence does not exist yet. Strength, progressive loading and, when needed, a precisely placed injection remain the foundation.
The knee is not always the whole problem
Osteoarthritis was taught for decades as simple mechanical wear: load exceeds tissue tolerance, cartilage thins, pain follows. That story is true but incomplete.
In a substantial proportion of people, osteoarthritis behaves like a metabolic condition that happens to present in a joint. Excess adipose tissue is not inert; it produces inflammatory signals that circulate and act on joint tissue directly. This is why hand osteoarthritis is more common in obesity, despite the hands carrying no body weight, and why two people with identical X-rays can have entirely different amounts of pain. If part of the driver is metabolic, part of the treatment can be metabolic too. This is the door that GLP-1 medicines walked through.
This metabolic picture is also why the menopause matters. The fall in oestrogen around menopause is associated both with a greater tendency to gain weight and with a higher risk of knee osteoarthritis, so for many women the joint changes and the metabolic changes arrive together. We look at this overlap in more detail in our guide to menopause and musculoskeletal health.
What a GLP-1 receptor agonist actually is
GLP-1 is a hormone the gut releases after eating. It prompts insulin release, slows gastric emptying, and signals to the brain that you have had enough. GLP-1 receptor agonists imitate that hormone with a much longer duration of action.
They are prescription-only medicines, licensed in the UK for type 2 diabetes and for weight management in adults meeting defined body mass index criteria. They are not licensed anywhere for the treatment of osteoarthritis. Any use aimed at the joint itself, rather than at weight and metabolic health, is off-label.
What the evidence shows
The landmark study is the STEP 9 trial, published in the New England Journal of Medicine in 2024. It recruited 407 adults with obesity and moderate, radiologically confirmed knee osteoarthritis with at least moderately severe pain, and randomised them to weekly semaglutide 2.4 mg or placebo for 68 weeks, both alongside dietary advice and increased physical activity.
Body weight: −13.7% with semaglutide versus −3.2% with placebo. WOMAC pain (0–100 scale): −41.7 points versus −27.5 points. Physical function significantly improved in the medicine group.
To put that pain improvement in perspective, its size was in the broad range that has been reported after knee replacement surgery. The two are not directly comparable, and the trial was not designed to compare them, but it gives a sense of why the result drew so much attention.
Two details deserve more attention than the headline. First, the placebo group still improved by 27.5 points, a large, clinically meaningful improvement from diet and activity advice alone. The medicine added to a good baseline; it did not replace it. Second, an earlier trial of a different GLP-1 medicine, liraglutide 3 mg daily for 52 weeks, produced only modest weight loss and no significant improvement in knee pain. Same drug class, very different result. The most parsimonious explanation is that the amount of weight and metabolic change achieved is what drives the joint benefit, not the molecule itself.
Observational data add a suggestive layer: people with knee osteoarthritis and type 2 diabetes taking these medicines long-term have shown slower cartilage loss on imaging and a lower rate of progressing to knee surgery. Those studies cannot establish cause and effect, because people who take and tolerate a medicine for years differ in many ways from those who do not. They are a reason to run better trials, not a reason to change practice on their own.
Is it more than just weight loss?
Part of the benefit is simply weight loss, and that is worth saying plainly, because weight loss can be achieved without any medicine at all. Diet and progressive exercise remain the cornerstone of osteoarthritis care, and the placebo group in STEP 9 is the proof of it: they lost weight and improved their pain substantially on advice alone.
But the picture now looks like more than weight. Osteoarthritis is increasingly understood as partly a metabolic and inflammatory condition rather than simple wear and tear, a view that is now well established in rheumatology. Body fat is not an inert store: it releases inflammatory signalling proteins called cytokines that travel through the body and can act on joint tissue, nudging it towards inflammation and cartilage breakdown. Reducing that inflammatory load may help the joint in its own right, over and above the mechanical relief of simply carrying less weight.
There may also be a direct effect. GLP-1 receptors, the docking points these medicines act on, have been identified in the synovium (the lining of the joint) and in cartilage itself. That raises a genuine possibility that these medicines quieten inflammation signalling inside the joint directly, independent of weight. Laboratory and animal work supports the idea, and observational data hint at slower cartilage loss over time. In humans, though, no trial has yet separated the joint effect from the weight effect, so a weight-independent benefit remains a well-supported hypothesis rather than an established fact.
This is the thinking behind one of the most interesting research directions in the field: giving GLP-1 directly into the joint, as an intra-articular injection, rather than as a systemic weekly dose. The aim is to place the medicine exactly where the receptors are, on the cartilage cells and the joint lining, to capture a local pain-relieving, anti-inflammatory, cartilage-protective and possibly cartilage-repairing effect, while avoiding whole-body weight loss and its side effects. If it works, it would matter to people with osteoarthritis whether or not they are overweight, and it is being explored as a possible first disease-modifying treatment, one that slows the disease itself rather than only easing symptoms.
The evidence so far is early. Most of it is preclinical: laboratory studies and animal models in which intra-articular liraglutide has reduced pain and joint-lining inflammation and helped preserve cartilage. The first human studies have only just begun, a small feasibility study followed by an early proof-of-concept trial starting in 2025, and this work is being driven by a commercial developer, so healthy caution applies. It is not something any clinic can offer today, and it is likely years away from routine use. It is a genuinely promising idea to watch, not a treatment to seek out yet. Readers who want the science will find it in the reviews listed at the end.
One honest limitation frames all of this: there is currently no licensed disease-modifying drug for osteoarthritis, nothing proven to halt or reverse the underlying process. Everything we have either manages symptoms or addresses the drivers, such as weight and metabolic health. A GLP-1 medicine, where it is appropriate at all, sits in that second group.
What these medicines are not
Not a treatment for osteoarthritis. No regulator has licensed them for it.
Not cartilage regeneration. Nothing currently available rebuilds a worn joint surface.
Not a replacement for exercise. Rapid weight loss without resistance training costs muscle, and muscle is what protects an arthritic knee. Losing weight while losing quadriceps strength can leave a knee no better off, which is precisely why this should never be prescribed in isolation.
Not a treatment for inflammatory arthritis. In rheumatoid or psoriatic arthritis these medicines may improve metabolic health and lower inflammatory markers, but no trial has shown they reduce disease activity or joint damage. They must never displace DMARD or biologic therapy.
Not appropriate for everyone with a sore knee. A lean 55-year-old with a stiff, painful knee and a normal metabolic profile has nothing obvious to gain and a list of possible side effects to accept.
Where it fits in a multidisciplinary plan
Osteoarthritis management works as four strands that reinforce each other. A GLP-1 medicine, where it is appropriate at all, supports one of them.
1. Loading and strength, the foundation. Supervised exercise therapy remains the single most evidence-supported intervention in osteoarthritis, and UK national guidance places it first for everyone. It is also the strand people abandon first, usually because pain makes the early sessions feel impossible.
2. Weight and metabolic health. Weight reduction improves pain and function roughly in proportion to how much is lost, and it improves the inflammatory environment the joint sits in. For most people this is diet and activity. For some, particularly those with significant obesity, type 2 diabetes, or a long history of unsuccessful attempts, a licensed weight-management medicine prescribed and monitored properly is a legitimate part of that strand.
3. Targeted injection, buying a window. An injection does not change the disease. What a well-placed, ultrasound-guided corticosteroid injection can do is reduce pain enough, for long enough, that someone can actually begin loading the joint and moving more. That window is the entire point; an injection given without a rehabilitation plan attached is a wasted opportunity. For knee osteoarthritis specifically, other injectables such as hyaluronic acid and Arthrosamid are also used in selected cases. Ultrasound guidance matters because blind injections around the knee, hip and shoulder miss their target more often than most clinicians expect.
4. Education, expectations and medicines review. Understanding that pain does not equal damage, that flares settle, and that a joint can be sore and safe at the same time changes behaviour more than most interventions.
Sequencing usually matters more than ingredients: injection to open a window, exercise to use it, metabolic change to keep it open.
Muscle matters: why strength training goes with the medicine
Here is the single most important caveat if you take one of these medicines, and the reason I will not prescribe one without an exercise plan attached. GLP-1 medicines are very effective at reducing body weight, but not all of what is lost is fat. Across several meta-analyses, roughly a quarter to a third of the weight lost is lean body mass, which includes muscle. One recent systematic review put lean tissue at about 28% of the total weight lost, and its authors specifically recommended monitoring body composition and working to preserve muscle.
That matters for anyone, but it matters more when the target is an arthritic knee, because the muscle most at risk, the quadriceps at the front of the thigh, is exactly the muscle that protects and controls that knee. Weak quadriceps are linked to more pain, poorer function, reduced walking ability, greater disability and a higher risk of falls. Lose weight but also lose quadriceps strength, and the knee can end up no better off. Weight loss on its own is not the goal; losing fat while keeping, or building, muscle is.
Resistance training is the main defence. Supervised strengthening carried out for more than about ten weeks can meaningfully increase muscle mass and strength, which is why current expert reviews recommend resistance exercise as a companion to GLP-1 therapy rather than an optional extra. Exactly how much muscle it preserves has not yet been pinned down in high-quality trials, but there is already more than enough evidence to make it a standard part of the plan.
Protein supports it. Pairing that training with enough dietary protein gives the body the raw material to hold on to muscle while the fat comes off.
The combination is worth more than the sum of its parts. The medicine reduces weight and the load on the knee, resistance training preserves the muscle and improves function, and together they can put someone in a far better position to take part in rehabilitation than either could alone. And since strengthening exercise is, on its own, among the most effective treatments for knee osteoarthritis pain and function in the latest evidence, none of that effort is ever wasted.
How this works in this clinic
Where it is clinically indicated, I prescribe a licensed weight-management medicine as one part of a multidisciplinary plan for knee osteoarthritis. In practice this is for people whose knee osteoarthritis is bound up with excess weight they have not been able to manage through diet and exercise alone, typically with a body mass index (BMI) over 30. It is always one strand of a wider plan, alongside rehabilitation and, where appropriate, a targeted injection, and never a treatment on its own. A few boundaries are worth stating plainly, because they are the difference between this being good medicine and being a weight-loss shop.
Diet and exercise come first, always. A licensed medicine is only ever added to a plan built on progressive loading, strength work and, where relevant, dietary change. If you are not doing that work, there is nothing for the medicine to add to, and for many people resistance training and a well-placed injection will do more.
Injection or oral, where suitable. These medicines are best known as weekly injections, but we can also offer an oral GLP-1 option for people who would prefer not to inject.
Common side effects are mostly gut-related. Nausea, vomiting and diarrhoea are the usual ones, typically worse when starting or increasing the dose and often settling with time. If side effects are troublesome, or the medicine is simply not tolerated, that is reviewed promptly.
Long-term prescribing is shared with a doctor. Where treatment is appropriate I would initiate it. For ongoing use, any tolerance issues and long-term review, a review with a doctor is recommended, and I work alongside a doctor I can refer you to for exactly that, so your care is properly supervised over time.
Not everyone is eligible or suitable. Prescribing requires meeting the licensed criteria and passing a full screen for contraindications and cautions. A personal or family history of medullary thyroid cancer, or of the genetic condition MEN 2, is a contraindication, and is one of the things that screen checks for.
This is not a weight-loss service. I do not treat obesity as a condition in its own right, and I do not take referrals or self-referrals for weight management alone. If weight rather than your knee is your primary concern, your GP or a specialist weight management service is the right route.
I do not offer microdosing. See below.
The microdosing question
"Microdosing" refers to doses well below the licensed range, often a fraction of the lowest pen increment, with the stated aim of capturing an anti-inflammatory effect on the joint while avoiding significant weight loss and gastrointestinal side effects. It has become popular online, particularly for joint pain, stiffness and generalised aching. An honest assessment:
The hypothesis is genuinely promising. If part of the joint benefit is a direct anti-inflammatory action rather than weight loss, a low dose might capture some of it without the weight change, which would make it relevant to a much larger group of people with osteoarthritis, including those who are not overweight at all. That is a serious idea, and it is why rheumatology is paying attention.
It has not been tested. There are no randomised controlled trials of low-dose GLP-1 for osteoarthritis, rheumatoid arthritis, or any other arthritis. None. The supporting evidence is laboratory work at pharmacologically active doses, plus uncontrolled patient reports. Joint pain fluctuates enormously on its own and responds strongly to expectation, which makes anecdote an unusually poor guide here.
The mechanism cuts both ways. The anti-inflammatory effects seen in the laboratory are dose-dependent, strongest at high exposures or with injection directly into the joint. Whether a sub-therapeutic systemic dose reaches a meaningful concentration in joint tissue is unknown.
The supply route is the immediate risk. Licensed pens are calibrated to fixed increments. Achieving a true microdose generally means a compounded or repackaged product from outside the licensed supply chain. Regulators have taken repeated enforcement action against suppliers of these products, citing inaccurate potency, sterility failures, false labelling and misleading marketing. What is actually in the vial is a more urgent question than whether a low dose works.
Side effects do not scale away neatly. Gastrointestinal effects, gallbladder disease and rare pancreatitis are class risks. A lower dose reduces the likelihood of some of them, but does not abolish them, and comes with no established monitoring framework.
So: promising, off-label, and unproven. I do not offer it, and I will not prescribe it on request. If that changes it will be because trial evidence has arrived, and this page will be updated to say so.
The bottom line
If you have knee osteoarthritis alongside significant obesity or type 2 diabetes, the evidence that substantial weight and metabolic change improves your knee is now strong, and a licensed weight-management medicine may be a reasonable part of how you get there, inside a proper plan, not instead of one.
If you are hoping for a small injection that treats arthritis directly, at a dose low enough that nothing else changes, that evidence does not exist yet. It might one day. Building a plan on it now means paying for hope with money, side effects, and time you could have spent on the things that reliably work.
The unglamorous answer remains the right one. Get strong, load the joint progressively, address metabolic health where it is genuinely relevant to you, and use a precisely placed injection when pain is the thing blocking progress rather than as a substitute for it.
References and further reading
STEP 9 trial: once-weekly semaglutide in adults with obesity and knee osteoarthritis, New England Journal of Medicine, 2024. Earlier randomised trial of liraglutide 3 mg for knee osteoarthritis, 52 weeks. Observational cohort data on GLP-1 receptor agonist use, cartilage change on imaging, and rates of knee replacement.
On muscle, exercise and protein: Bittel et al., Diabetes Care (2024) review resistance training alongside GLP-1-based therapy; the 2025 BMJ network meta-analysis confirms strengthening exercise as one of the most effective treatments for knee osteoarthritis; and NICE guideline NG226 (2022) places exercise first in osteoarthritis care. Systematic reviews and meta-analyses of body composition during GLP-1 treatment show that a substantial share of the weight lost is lean tissue.
Further reading on GLP-1 receptors in joint tissue and the intra-articular research direction: a systematic review in Frontiers in Pharmacology (2025); a mechanistic review in Signal Transduction and Targeted Therapy (Nature, 2024); and an overview in the European Medical Journal, Rheumatology.
This article is general information about published research and is not medical advice, nor a recommendation to use any specific medicine. GLP-1 receptor agonists are prescription-only medicines and are not licensed for the treatment of osteoarthritis. Prescribing decisions are made individually at consultation, only where licensed criteria and clinical suitability are met, and in shared care with your GP. If weight management rather than joint pain is your primary concern, speak to your GP or a specialist weight management service.