Exercise with knee osteoarthritis: Why movement comes first in the guidelines

The fear of wearing out, and why it is so understandable

Wear and tear. Worn out. Bone on bone. Maybe you have heard these words before, at a doctor's appointment, from friends or on the internet. They create a clear image: a component that gets used up.

If you carry this image in your head, you move more carefully. That makes sense. Nobody wants to wear out a part faster than necessary.

You are not alone with this worry. Osteoarthritis is the most common joint disease of all, and the knee is the most commonly affected site.

The global figures come from the Global Burden of Disease study in the journal Lancet Rheumatology. There, around a fifth of the osteoarthritis burden was statistically attributed to high body weight. We will come back to this when we look at belly fat.

But how does the image of wear and tear form in our minds, and what does it do to us? Before we look at the studies, it is worth taking a quick look at the evidence labels in this article.

Osteoarthritis research has moved a good distance away from the pure wear-and-tear image. In a widely cited review, Berenbaum from the Sorbonne in Paris describes osteoarthritis as a disease partly shaped by inflammation. Not only cartilage is involved, but also bone and the joint lining. So the image of a tyre wearing thin captures only part of the picture.

And the crunching in the morning? There are data on that too. To be honest, they are not only reassuring, but they are helpful.

And now you know why the fear of wearing out is so powerful, and why it is worth examining the image behind it.

Cartilage has no blood vessels: how movement nourishes it

Slowly bend your knee. Between your thighbone and your shinbone, two smooth cartilage surfaces glide over each other. Almost without friction, thousands of times a day.

This tissue has a special feature that hardly anyone knows about. It has no blood vessels of its own, no nerves and no lymphatic vessels. Sophia Fox, Bedi and Rodeo from the Hospital for Special Surgery in New York describe this in a review of cartilage biology.

So how do the cartilage cells get their nourishment? According to this review, mainly through diffusion from the joint fluid. The authors also stress that regular joint movement and changing loads are important for healthy cartilage metabolism. At the same time, cartilage has only a limited ability to repair itself. So this is about smart dosing, not full throttle.

Imagine a sponge in a bowl of water. If you squeeze it, fluid flows out. If you let go, it soaks itself full again. Cartilage behaves in a very similar way under loading and unloading. However, this image is a simplification, as a laboratory study shows.

Even more important than the pumping effect could be the fact that load is a message. In a review, Sanchez-Adams and colleagues from Duke University describe how cartilage cells respond to pressure and deformation. According to them, mechanical, inflammatory and biochemical factors can either support or inhibit the maintenance of the cartilage matrix.

Conversely, can cartilage also respond to more movement? A small but remarkable study from Sweden suggests it can.

For your everyday life, this means: many small portions of movement spread across the day can be valuable for your knee. Long periods of sitting, on the other hand, take exactly this alternation away from the cartilage. And now you know why a tissue without blood vessels depends so much on movement.

Why movement comes first in the guidelines

When the knee hurts, many people first think of a pill, an injection or an MRI. That is natural. It feels more concrete than the advice to move.

That makes what the major guidelines say all the more surprising. Professional societies in rheumatology and orthopaedics have systematically assessed the evidence. Their common denominator: exercise and education form the foundation, and other treatments build on it.

The German S3 guideline of the German Society for Orthopaedics and Trauma Surgery was published in May 2025 and replaces the earlier S2k guideline. Its patient version names sport, movement and strengthening exercises as important building blocks. It states that they can ease pain, increase mobility and contribute to slower progression of osteoarthritis.

I find one detail particularly revealing. According to a summary by the Drug Commission of the German Medical Association, drug therapy for knee osteoarthritis is purely symptomatic. One of its goals is to make active exercise therapy possible in the first place. According to the patient version, a cortisone injection can also be useful if it is what makes taking part in exercise therapy possible.

These recommendations do not come from a niche. They come from rheumatology and orthopaedics themselves, backed by systematic analyses. Here, conventional medicine and an integrative view meet without contradiction. Above all, the integrative perspective adds a look at metabolism, inflammation and the nervous system. More on that shortly.

A practical tip from the patient version: self-help organisations such as the Deutsche Rheuma-Liga (German Rheumatism League) offer functional training that is prescribed by a doctor and funded by health insurers. Physiotherapy and rehabilitation sport are also named there as recommended. Feel free to ask about this specifically at your next appointment.

And now you know why movement is right at the top of all the major guidelines.

What studies on exercise for osteoarthritis show, and what they don't

Maybe you have already tried. A few weeks of exercises, then a bad day, and you stopped. Asking whether the effort is worth it is a fair question.

The most thorough answers come from Cochrane reviews. They pool all available randomised trials and rate their quality strictly.

Nine years later, the update was published. It is larger, and it is worded more cautiously. That is exactly what makes it valuable.

How does it compare with surgery? There is a widely noted Danish study on that.

The Danish care programme GLA:D shows how such programmes can work in everyday care. Skou and Roos analysed registry data from 9,825 participants who received education and supervised exercise from physiotherapists. Pain intensity fell by 12.4 points after three months and by 13.7 points after twelve months on a scale from 0 to 100. Afterwards, fewer participants took painkillers, and fewer were on sick leave. There was no control group, however, so part of the improvement may also have other causes.

And now you know why studies see exercise as a solid building block, even if they promise no miracles.

Does exercise wear down cartilage? What MRI and running studies show

Even if movement can ease the pain, a quiet question often remains: Will I end up paying for it with my cartilage?

Today, this question can be studied surprisingly directly. Modern MRI methods measure the thickness, volume and composition of cartilage. Breakdown products of cartilage and inflammatory markers can be measured in the blood.

A similar picture emerges in everyday life. In an analysis of the Osteoarthritis Initiative, Bricca and colleagues followed 689 people with knee osteoarthritis over four years. Overall, cartilage loss did not depend on activity level. However, women who moved little lost more cartilage than women with moderate activity. And a high activity level was not more harmful than a low one in either sex.

And what about running, the sport with the worst reputation for the knee?

What if osteoarthritis is already there? In the Osteoarthritis Initiative, Lo and colleagues studied 1,203 people aged 50 and over with knee osteoarthritis. Those who ran of their own accord had neither more frequent new knee pain nor faster progression on X-rays over four years. The authors conclude that self-selected running does not need to be discouraged as a matter of principle in knee osteoarthritis.

In 2025, Bricca, Juhl and Roos summarised the state of knowledge in a journal article. According to them, systematic reviews show that exercise therapy does not appear to damage the structure and quality of cartilage. Clinicians should therefore actively address the worry about wearing out the knee.

And now you know why experts today describe the fear that exercise wears out the knee as a widespread misconception.

Silent inflammation and belly fat: osteoarthritis is more than mechanics

Does this sound familiar? On some days your knee is warm and a little swollen, on others perfectly calm. And sometimes the pain does not fit at all with what you did the day before.

That is hard to explain with wear and tear alone. In a review, Robinson and colleagues from Stanford University describe osteoarthritis as a disease of the whole joint. Cartilage, bone, joint capsule and joint lining change together. A chronic, low-grade inflammation, sustained mainly by the innate immune system, could help drive many of these changes.

Don't picture this inflammation as a large blaze, as in rheumatoid arthritis. Think of it more as embers quietly smouldering on. Among the possible accelerants, Berenbaum names metabolic syndrome, meaning the interplay of belly fat, blood sugar, blood lipids and blood pressure, as well as inflammatory ageing processes.

A surprising clue comes from the hands. In a systematic review by Yusuf and colleagues, overweight was also associated with osteoarthritis of the finger joints, with about 1.9 times the risk. Fingers don't carry body weight. So more than load must be at play here.

For the knee, mechanics still matter. In the Dutch NEO study with 6,673 people, knee osteoarthritis was linked mainly to weight and fat-free mass, in other words to load. Hand osteoarthritis, on the other hand, was linked to metabolic syndrome, independently of weight. So in the knee, mechanics and metabolism seem to work together.

And belly fat itself? Visceral fat, the fat between the organs, is considered particularly metabolically active. It produces inflammatory messengers and hormones known as adipokines.

One messenger between fat tissue and joint could be the hormone leptin. Dumond and colleagues from Nancy found leptin in the joint fluid of people with osteoarthritis, and its concentration was linked to body mass index. In osteoarthritic cartilage, considerably more cartilage cells produced leptin than in healthy cartilage. Experiments in rats showed that leptin can strongly influence the metabolism of cartilage cells. Exactly what role this plays in humans is still open.

How well can this axis be influenced? There is a large randomised trial on that.

An earlier gait analysis by the same research group with 142 people shows how much mechanics count here. Each kilogram of body weight lost went along with around 40 newtons less compressive force in the knee per step. That is about four times the weight lost. Over thousands of steps a day, that adds up.

If you want to dig deeper into how blood sugar, insulin and belly fat are connected, read the article Recognising and reversing insulin resistance.

And now you know why knee osteoarthritis can also be a story about metabolism.

Muscle strength: the shock absorber you can train

Going down the stairs. Getting up from a low armchair. The first step out of the car. It is often exactly these moments when your knee makes itself felt.

In all these movements, one muscle group does most of the work: the knee extensors at the front of the thigh, known medically as the quadriceps. Imagine them as the shock absorbers of a car. They brake, stabilise and distribute the load before it reaches the joint.

Muscles are more than mechanics, though. Pedersen and Febbraio from Copenhagen describe skeletal muscle as an organ that releases messengers called myokines. Through them, muscle talks to fat tissue, liver, bone and brain. Many of these messengers depend on the muscle working. According to the authors, a lack of movement therefore probably changes this conversation.

Does that mean the heavier, the better? A large study tested exactly this question, and the result is a relief.

A Cochrane review by Regnaux and colleagues comes to a similar conclusion. In six studies with 656 people, more intensive programmes brought no clinically meaningful advantage in pain and function over less intensive ones. The data quality was low to very low. That is why the EULAR recommendations rely on individually tailored dosing with gradual progression.

You can find more on why muscles shrink over the years and what you can do about it in the articles Muscle loss with age and Strength training after 40.

And now you know why muscle strength means more for your knee than looks or performance.

Pain is a signal, not a damage report: the pain traffic light

You trained, and in the evening your knee is aching. Straight away the thought is there: Have I broken something now?

This is where a look at biology is worthwhile. Cartilage itself has no nerves, as the review by Sophia Fox and colleagues describes. What you feel arises in surrounding structures such as bone, joint capsule and joint lining. And it arises in your nervous system, which processes and evaluates these signals.

A review by Bedson and Croft from Keele University shows how loosely damage and pain can be connected. Depending on the study, 15 to 76 percent of people with knee pain had signs of osteoarthritis on X-ray. Conversely, 15 to 81 percent of people with X-ray changes had pain. The authors therefore advise against assessing X-rays in isolation.

With long-standing osteoarthritis, the nervous system can also change. In a meta-analysis of 15 studies, Fingleton and colleagues from Dublin found signs of pain sensitisation in knee osteoarthritis. People with osteoarthritis already responded with pain to lower pressure, and this was linked to the severity of their symptoms. Imagine an alarm system that has been set to be more sensitive over time and goes off even when a cat scurries past.

For training, this is important news. Pain during or after movement is feedback on the dose, not automatically a sign of new damage.

But how much pain is too much? In physiotherapy, the pain-monitoring model is often used for this. Thomeé originally described it for pain behind the kneecap. Silbernagel and colleagues later used it in a randomised trial in Achilles tendon problems. In that trial, people who kept running and jumping with the help of the model did no worse than the group that took a six-week break from loading.

The core rules of the model are simple: pain should not exceed 5 out of 10 during or right after exercise. And it should return to the pre-exercise level by the next morning. A traffic light can be derived from this.

The traffic light is a simplified guide. The limit of 5 out of 10 and the next-morning rule come from the pain-monitoring model, while the division into three colours is there for clarity. It does not replace individual guidance from a physiotherapist or doctor.

And now you know why you should take pain during training seriously, but don't have to read it as a stop sign for every movement.

What is established and what remains open

Three levers you can start with today

Maybe all of this feels like a lot right now. Then pick just one lever. That is plenty for a start.

In Rudolf Steiner's anthroposophic view of the human being, metabolism and limbs form one shared system, to which the will is assigned. This image accompanies me when it comes to osteoarthritis. Modern research links joint, muscle and metabolism in its own way. Movement starts right there, with doing rather than brooding. This is a perspective from anthroposophic medicine, not a finding from a study.

An important exception: if your knee suddenly becomes hot, red and swollen, if you have a fever, or if you can no longer put weight on your leg after an accident, the focus is not on training but on a medical assessment. Different rules apply to exhaustion that lasts for days after exertion. You can find more on this in the article Exercise with chronic fatigue.

And now you know why, with knee osteoarthritis, movement is less a risk than an invitation to living tissue.

Frequently asked questions about exercise with knee osteoarthritis

You might also be interested in