1. What osteoarthritis really is
When people are given the diagnosis “osteoarthritis”, what many of them hear first is: worn out, cartilage gone, bone on bone. That language is catchy – and it is misleading.
Osteoarthritis affects the whole joint organ. It is not only the joint cartilage that changes, but also the bone underneath (it becomes denser and forms the typical bony outgrowths at the margins, the osteophytes), the joint lining (which can become inflamed at times), the joint capsule, the ligaments and the surrounding muscles [2]. Together these structures form a system that continuously responds and adapts to loading.
And that is the decisive point: joint tissue is living, adaptable tissue. It has a metabolism, it breaks down and it builds up. In osteoarthritis this balance tips – breakdown outweighs build-up. That is something fundamentally different from the passive wear of a tyre or a brake disc.
The practical consequence matters: a tyre inevitably gets worse the more you drive on it. A joint, by contrast, needs loading to stay healthy. Cartilage has no blood vessels of its own; it feeds itself by taking up and releasing joint fluid like a sponge as load comes on and off. If you protect a joint, you deprive it of exactly the stimulus it needs. Immobilisation is known to lead to cartilage deterioration, muscle loss and loss of strength.
Osteoarthritis is also common. According to the Global Burden of Disease analysis, around 595 million people worldwide were living with osteoarthritis in 2020, about 7.6 % of the world's population – more than double the number of cases since 1990, mainly because of the ageing population and rising body weight [1]. For the knee alone, a large US survey found radiographic signs in 37 % of people aged 60 and over [3]. So you are in very large company with this diagnosis.
1.1 How a joint is lubricated – and what movement has to do with it
A healthy joint glides with remarkably little friction. The joint fluid is responsible for this, and not through a single substance but through an interplay: hyaluronic acid, certain fat molecules (phospholipids) and a protein called lubricin work together. A review of the topic concludes that there is no single lubrication mechanism – the effect only arises in combination [4]. Lubricin itself is well studied and central to the gliding capacity of the cartilage surface [5].
The next step is the interesting one for you: physical activity intervenes in exactly this system. Reviews describe how movement influences the production of lubricin and of irisin – irisin being a messenger substance released by muscle during exertion – and that part of the favourable effect of exercise on cartilage might run through this route [6].
And now the qualification we owe you: this is a plausible explanation, not a proven mechanism. A systematic review gathered every study that measured lubricin after joint injury or in osteoarthritis. Of 38 studies, 19 found an increase or no change – and 19 found a decrease. What happened next is telling: studies reporting a decrease were cited roughly four times more often than those reporting the opposite [7]. The simple story “osteoarthritis = too little lubricin” is well known mainly because it tells well.
We mention the mechanism anyway, because it shows that a joint responds biologically to movement. But it carries no therapeutic promises.
2. Why scans can mislead
This is perhaps the most important section of this article – and the one that surprises most people.
The relationship between what is visible on an X-ray or MRI scan and what a person experiences in terms of pain and limitation is astonishingly loose.
Scans of people without symptoms. In a population-based study within the Framingham cohort, the knees of adults without radiographic osteoarthritis were examined by MRI. In 89 % at least one abnormality was found; marginal outgrowths in 74 %, cartilage damage in 69 %, bone marrow signals in 52 % [8]. Most of these people had no knee pain.
A synthesis of 63 studies covering more than 5000 knees of symptom-free, uninjured adults arrived at similar figures: cartilage defects in around 24 % overall – about 11 % in those under 40, about 43 % in those aged 40 and over. Meniscal tears were found in around 10 % overall and in around 19 % of those aged 40 and over [9]. All in people who had no knee problems whatsoever. A study of 230 knees of symptom-free adults using a 3 tesla scanner paints the same picture [10].
A note on the meniscus. The meniscus is not a spare part but a load-bearing element: it distributes load and absorbs shock. In osteoarthritis, meniscal changes are therefore very common – and they are part of what is happening in the joint rather than a separate accident. Whether the meniscus is cause or consequence often cannot be determined in the individual case [11]. A “degenerative meniscal tear” in a report is therefore rarely what it appears to be: damage in need of repair.
The reverse holds just as well. There are people with marked changes on the scan and little pain – and people with considerable pain and an almost normal scan.
What does that mean for you?
- A finding such as “grade 3 cartilage damage” or “degenerative meniscal tear” describes a state, not necessarily the cause of your pain. From middle age onwards such findings are so common that they belong more to the normal picture of ageing than to disease.
- A scan says very little about how well you will be doing in two years' time. Muscle strength, activity level, sleep quality, worries and expectations often predict the course better than the X-ray report.
- In typical osteoarthritis, additional imaging usually changes nothing about the treatment. The diagnosis is made clinically – specialists expressly describe the widespread use of imaging in osteoarthritis as overuse [2]. In people over 45 with activity-related joint pain and morning stiffness lasting no more than 30 minutes, clinical assessment is as a rule sufficient.
This expressly does not mean that imaging is useless. Where other causes are suspected – inflammatory rheumatic disease, infection, fracture, tumour, unexplained locking or an acute injury – it is important. But a scan taken “just to be safe” can do more harm than good if it creates fear of movement.
3. Why does osteoarthritis hurt at all?
Joint cartilage contains no pain fibres. When cartilage thins, you do not feel that directly. The pain comes from other sources:
- The joint lining. It is rich in nerve endings and can become inflamed in phases. That explains a large part of pain flares and swelling.
- The bone beneath the cartilage. This is where so-called bone marrow lesions are found – zones of increased turnover and fluid. They are relatively closely linked to pain and can come and go.
- Capsule, ligaments and tendon insertions, which are overtaxed when loading patterns change.
- The muscles. Weak muscles, loaded in unfamiliar ways, make themselves felt.
- The nervous system itself. With longer-lasting pain, pain processing can become more sensitive. Stimuli that were previously harmless are then reported more strongly as pain. This is not “imagined” pain – it is a real, measurable change. And it is reversible.
That muscle and pain sensitivity are connected is also shown by a more recent study: in 42 people aged between 50 and 85 with knee pain, pain sensitivity was measured alongside the mechanical properties of the thigh muscles – tone and stiffness. The profiles were related to one another and to reported pain [12]. This is a small pilot study and proves no cause and effect. But it fits what we see in practice: muscles, nervous system and pain are not three separate building sites.
In addition, lack of sleep, psychological strain, stress and the expectation that movement is harmful demonstrably influence pain intensity. That does not make the pain any less real. But it opens up points of attack that reach far beyond the joint.
4. How does osteoarthritis progress?
One of the most persistent fears runs: “From now on it will get worse every year until nothing works any more.”
That does not match what longitudinal studies show. In a considerable proportion of those affected, pain and function remain largely stable over years. In some they even improve. Only a minority experience continuous deterioration.
What is typical is a fluctuating course: phases with few symptoms, interrupted by flares that last days to weeks and then subside again. A flare does not mean that the joint is “another notch more broken”. It usually means that the joint is irritated for a short time.
This fluctuation matters not only for you but also for how research is read – we come to that in section 6.3.
Important: the scan and the symptoms do not develop in step either. X-ray changes can slowly increase while the pain decreases – or the other way round.
5. What influences the risk
Osteoarthritis arises through an interplay of several factors.
5.1 Not modifiable
- Age. The strongest single factor.
- Genetics. Particularly pronounced in hand and hip osteoarthritis.
- Sex. Women are affected more often and more severely after the menopause.
- Joint shape. Congenital variations at the hip or knee increase the risk.
5.2 Modifiable or partly modifiable
- Previous joint injuries. A cruciate ligament or meniscal tear considerably increases the risk of later knee osteoarthritis. Injury prevention in sport is therefore also osteoarthritis prevention.
- Body weight. High body weight is a recognised risk factor for knee and hip osteoarthritis. But mechanics explain only part of it: hand osteoarthritis is also linked to body weight – and the hands carry no body weight at all. Fat tissue produces messenger substances that promote inflammatory processes.
- Muscle weakness. Weak thigh muscles in particular often precede the onset of knee symptoms.
- Occupational loading. Decades of heavy lifting, kneeling and squatting increase the risk.
- Physical inactivity. It is not protective; it is a risk factor in its own right.
5.3 A special case: the torn cruciate ligament
After a cruciate ligament tear many people are told: “Have it operated on – otherwise you will get osteoarthritis later.” That reasoning does not hold.
A systematic review with meta-analysis examined five studies that followed operated and non-operated people for at least ten years after a cruciate ligament tear. The result: in the operated group the risk of radiographic osteoarthritis was higher, not lower. In return, those who had surgery needed later meniscal surgery less often. For repeat cruciate tears and follow-up procedures there was no difference, knee stability was better after surgery – and the outcomes reported by the patients themselves were the same in both groups [13].
The authors expressly urge caution in interpretation, because the underlying studies are methodologically difficult. For you it still means: cruciate ligament surgery can make good sense – stability, sport, occupation, accompanying injuries. Preventing osteoarthritis is not among those reasons.
6. What is proven to help: core treatment
Here the international guidelines agree to an unusual degree. The European professional society EULAR updated its recommendations in 2023 [14]; there are also recommendations from OARSI [15] and from the American College of Rheumatology, and for the hip a revised physiotherapy guideline is available [17]. A systematic review of the highest-quality guidelines summarised it as follows: what is consistently recommended is information, exercise and weight management – for everyone affected, at every stage, as the foundation of any further treatment [16].
A sobering aside: a large share of people with osteoarthritis never receive this core treatment at all [2].
6.1 Information and understanding
Understanding what is going on in your joint – and what is not – is itself an active ingredient. People who know that loading does not harm cartilage move more. People who move more have less pain. People who are less afraid of pain experience it less intensely.
That is why this article exists. And that is why every good osteoarthritis treatment begins with a conversation, not with a prescription.
6.2 Movement and exercise
This is the measure every guideline puts first. How large its benefit actually is, however, can be answered more honestly than it often is – see section 6.3 straight after this.
Knee. The most comprehensive analysis is the 2024 Cochrane review [18]. It took an important methodological step: it separated the comparison groups cleanly instead of lumping them together, as earlier analyses did. The result: exercise very probably improves pain and function. The magnitude is around 8 to 10 points on a 100-point scale for pain, similar for function, less for quality of life. The certainty of the evidence was rated low to moderate, because many individual studies are small and methodologically weak [18].
Hip – and here it gets uncomfortable. Since 2026 the hip has had its own Cochrane review, covering 18 studies and 1368 participants [19]. Its findings are considerably more reserved:
- Compared with a sham treatment or attention control, exercise may make little to no difference to pain (2 studies, 123 participants, low certainty); physical function may improve slightly.
- Compared with no treatment or usual care, exercise probably improves pain and function slightly – but the review states expressly that these improvements are unlikely to be clinically meaningful.
- Exercise added to another treatment probably achieves little to nothing over that treatment alone.
This does not mean you should not exercise with hip osteoarthritis. It means expectations must stay realistic – and that the evidence for the hip is weaker than for the knee. One incidental finding of the same review is encouraging: exercise added to another treatment tended to reduce the number of adverse events rather than increase it [19].
Which type of exercise? The Cochrane analysis found no meaningful differences between forms of exercise – and no clear relationship between the number of sessions prescribed and the outcome either [18][21]. Strength training, endurance training, balance training, aquatic exercise, tai chi, a group class or a home programme: all of them show an effect.
One distinct variant is neuromuscular exercise, as used in structured osteoarthritis programmes. It does not aim at maximum strength but at motor control and joint stability; the exercises are mostly performed standing, across several joints, with simple equipment. Important for you: it is considered safe – neither frequent pain flares nor serious adverse events – and it affects pain, function and quality of life much as other forms of exercise do [22].
In practice that means: the best exercise is the one you actually do regularly. If you like water, go into the water. If you like strength training, go to the gym. If you hate both but enjoy hiking, then hike.
How much? As a guide: targeted strength and functional training two to three times a week for at least 12 weeks, supplemented by everyday activity on as many days as possible. After 12 weeks you should notice a change. A large analysis of individual participant data did, however, find diminishing additional returns beyond 12 weeks [20]. And the guidelines expressly stress that the effect only lasts for as long as you keep going [14] – exercise is not a cure, it is a lasting habit.
Who benefits most? The same analysis of 31 trials with 4241 people looked specifically for characteristics that predict success. Of twelve characteristics examined, only two survived: those who started with more pain and worse function were more likely to benefit [20]. Age, weight and radiographic stage predicted nothing.
And pain while exercising? That is the most frequent question. A workable rule of thumb from structured osteoarthritis programmes: discomfort up to about 5 out of 10 during and after the exercise is acceptable, as long as it settles back to the starting level by the next morning. If the discomfort stays elevated beyond 24 hours, cut back next time – less weight, fewer repetitions, a smaller range of movement. You do not damage your joint by loading it.
Please discuss this rule with your physiotherapist – in certain situations (a fresh injury, an inflammatory disease, a joint replacement, a pronounced flare) different rules apply.
6.3 How certain are these numbers? An honest interjection
You have read numbers above: 8 to 10 points, probably a slight improvement, low to moderate certainty. An uncomfortable debate about them has been running in the field for some years, and we think it is right not to keep it from you.
The charge. A widely noted commentary in a specialist journal carries the title “The emperor's new clothes?” [23]. The core of it: the strong recommendation for exercise rests mostly on studies that compared exercise with nothing or with a minimal treatment – not with a sham treatment. Other approaches such as TENS or acupuncture were downgraded precisely because they performed poorly against a sham. Where exercise has been tested against a genuine sham – for instance against saline injections into the joint – the result has largely been negative: both groups improved to a similar degree, and lastingly so.
Two effects that must be kept apart. For the common knee osteoarthritis treatments, contextual factors explain on average about 75 % of the observed improvement [23]. They consist of two parts:
- The placebo effect – a real improvement, even if a contextual one.
- Regression to the mean – no improvement at all, but a statistical illusion.
What is regression to the mean? Osteoarthritis pain fluctuates (see section 4). People usually sign up for a treatment when it happens to hurt most. Afterwards things move back towards their personal average – all by themselves. Anyone comparing the moment of enrolment with the moment three months later is therefore also measuring this return to normal, and readily credits it to the treatment.
How large is it? An analysis of 547 people from a large observational study put a figure on it: regression to the mean accounts for about 1 point on a pain scale from 0 to 10 – against typical reported improvements in knee osteoarthritis trials of 1.2 to 2.5 points [24]. The same was estimated for the large national osteoarthritis programmes in Denmark and Sweden: between 15 and 60 % of the reported improvement is likely to be down to this effect [25]. That second piece of work is a conference abstract, so it has not been peer reviewed as a full publication – but the order of magnitude matches the first.
So what does this mean for you? Four things:
- Figures from programmes without a comparison group (“our participants improved by X points”) are not proof of effectiveness. That applies to programmes we think highly of, too.
- Part of what improves after a treatment would have improved without it. That is not a disgrace, it is statistics.
- The comparison between two groups of a randomised trial is not affected by this problem – which is why the 8 to 10 points from the Cochrane review are more robust than any before-and-after figure.
- Despite all of this, exercise remains recommended. Not because the effects are large, but because they are real and broad, because the risk is small, and because none of the alternatives stands up better. A joint that is moved does the muscles, the circulation, the bones and the mood good anyway – and no pain scale measures that.
For context: the effects of exercise are comparable to those of anti-inflammatory medication, but without its risk of side effects for the stomach, the kidneys and the cardiovascular system. No drug acts at the same time on blood pressure, blood sugar, bone density, fall risk and mood.
6.4 Weight
Where there is excess weight, weight reduction is one of the most effective measures in knee and hip osteoarthritis [14][16]. The magnitude at which studies show clear improvements is around 5 to 10 % of body weight. Combined with exercise, the effect is greater than with either measure alone.
New to the discussion are the GLP-1 medications. In the STEP 9 trial, 407 people with obesity and moderate knee osteoarthritis received semaglutide or placebo over 68 weeks, both in addition to dietary and exercise counselling. Weight fell by 13.7 % compared with 3.2 % on placebo. The WOMAC pain score improved by 41.7 points compared with 27.5 points on placebo [26].
These figures are remarkable – but read them carefully: the placebo group also improved by 27.5 points. The added benefit of the drug was around 14 points. And both groups received exercise and dietary counselling. A fine illustration of everything in section 6.3: the difference between the groups is the dependable number, not the improvement within one group. Such medications require a prescription, have side effects (mainly gastrointestinal) and do not replace core treatment. Whether they are an option for you is a medical decision.
One point that matters to us: if weight reduction is not realistic for you right now, that is no reason to do nothing at all. Exercise works without weight loss too.
7. Can I still do sport? And run?
Yes. And it is very probably good for you.
The widespread assumption that running “uses up” the knees does not stand up to scrutiny. A synthesis of 25 studies with more than 125,000 people found the following rates of hip and knee osteoarthritis: 3.5 % in recreational runners, 10.2 % in non-running controls and 13.3 % in competitive runners [27].
Recreational runners therefore fared better than the inactive comparison group. Only very high volumes and intensities over many years – elite-sport level – went along with a higher rate. Note that these are observational data: people who already have pain are more likely to stop running, which also influences the figures. But the direction is unambiguous: recreational running is not a risk factor.
For you in practice:
- Increase volume and intensity slowly. The problem is not loading as such, but increasing it too quickly.
- Vary the forms of loading – cycling, hiking, swimming, strength work.
- Avoid longer complete breaks. A joint loses load tolerance faster than you can rebuild it.
- During a flare: do not stop, adapt. Shorter, slower, flatter – and build back up again within a few days.
8. What helps less than commonly assumed
This section is uncomfortable, but it is part of honest information.
Supplements for “cartilage rebuilding” (glucosamine, chondroitin, collagen): despite decades of research and intensive marketing, the evidence is weak and contradictory. In the better-quality studies the effects shrink towards zero. Several guidelines advise against them or make no recommendation [15][16].
Hyaluronic acid injections: for the hip, consistently not recommended [16][17]. For the knee the recommendations are contradictory – the effects in good studies are small and the placebo component of an injection into a joint is considerable.
PRP (platelet-rich plasma) and stem cell therapies: the better-quality guidelines advise against them [16]. A five-year analysis found no long-term advantage over standard treatment in early knee osteoarthritis. These treatments are expensive and are often marketed with promises the data do not support.
Corticosteroid injections into the joint: these can help in the short term (weeks to a few months) and are an option in the knee during an acute flare. But they are not a lasting solution, and frequent repetition is viewed with caution.
Arthroscopy for degenerative complaints: consistently not recommended in the better-quality guidelines [16]. Several studies using sham surgery as a comparison showed no additional benefit. The case is different with genuine mechanical locking or fresh injuries – that is a medical judgement.
Ultrasound, electrotherapy, TENS: the evidence is weak; TENS is explicitly not recommended in several guidelines [15][16].
Passive measures in general (massage, heat, cold, tape, manual therapy): they can be pleasant in the short term and relieve pain – and that is exactly where they have their place, for instance to open a window for active exercise. As a treatment on their own they are not sufficient. For hip osteoarthritis, however, there are guideline recommendations to use manual techniques alongside exercise [17].
Pain medication: anti-inflammatory drugs (NSAIDs) work, both orally and as a gel. But they carry relevant risks of side effects, particularly with prolonged use and with pre-existing conditions. Paracetamol is viewed with caution. Opioids are clearly advised against in osteoarthritis [15]. Medication is medical territory – discuss it with your family doctor, not with us and not with the internet.
8.1 One exception: the knee brace
This section is otherwise about disappointments. The knee brace is the counter-case – here the evidence has recently shifted in favour of the measure.
Until now the guidelines contradicted one another on bracing. A large British trial added to the picture in 2026: 466 people aged 45 and over with knee osteoarthritis received either advice, written information and exercise instruction – or the same plus a knee brace, selected according to the affected compartment of the joint, with a follow-up consultation to support wearing it. After six months the brace group did better, most clearly for pain (around 6 points on a 100-point scale, a small to moderate effect). The advantage diminished again over the course of the year; adverse effects were minor and expected [28].
For you this means: a brace is not a treatment for everyone and does not replace exercise. But as a time-limited addition – correctly chosen and correctly fitted – it is a serious option worth discussing.
9. When does surgery make sense?
Joint replacement at the hip and knee is one of the most successful procedures in modern medicine. For many people with severe, persistent symptoms and a clearly reduced quality of life it is the right decision.
Three things are worth knowing:
First: core treatment comes first. Studies show that a considerable proportion of people who arrive for an orthopaedic assessment or on a surgical waiting list have never received structured exercise and information treatment [2]. In programmes such as GLA:D, a relevant proportion of participants decide against the planned operation after a few months of training. Two qualifications: such programme figures come from analyses without a comparison group and are therefore not proof of effectiveness [25] – and declining surgery is itself a decision, not a measurement. The principle nevertheless holds: an operation you do not need is an avoidable risk.
Second: the X-ray finding alone is not a reason to operate. What matters is your pain, your function and your quality of life – not the joint space on the image.
Third: the outcome is good, but not perfect. A notable proportion of those operated on – more after knee replacements than after hip replacements – are not fully satisfied with the result. Good preparation (strength and mobility before the procedure) and consistent aftercare improve the prospects.
Surgery is a shared decision between you and your doctor – on the basis of realistic expectations.
10. Your everyday life: what else counts
Sleep. Poor sleep intensifies pain, and pain worsens sleep. Breaking that circle is one of the most effective and most overlooked measures.
Managing flares. A flare is unpleasant, but it is not an emergency and not a sign that something has broken. Reduce the load in the short term, but do not stop altogether. Heat or cold – whichever you find more pleasant. And take the activity up again after a few days, rather than waiting until “everything is fine”.
Smoking and accompanying conditions. Diabetes, high blood pressure and cardiovascular disease occur more often alongside osteoarthritis – partly through shared causes, partly through the reduced activity. Exercise acts on all of them at once.
Diet. There is no “osteoarthritis diet” with robust evidence. A balanced, Mediterranean-style diet makes sense for many reasons and may support weight regulation. Be sceptical of products that promise cartilage regeneration.
Psychological well-being. Worries, low mood and catastrophic thoughts (“it will never get better”) measurably influence pain intensity. That does not mean the pain is “in your head”. It means it is worth taking this side seriously too – with professional support if needed.
11. A realistic start
If you have read this article to the end and would like to change something tomorrow, here is a deliberately modest suggestion:
Weeks 1–2: establish a baseline
- Note down for one week: what do you do on a normal day? How strong is the pain (0–10)?
- Choose two or three exercises for the affected joint – better a few that you really do than a programme you give up after four days.
- Set fixed times. Not “when it fits”.
Weeks 3–6: build up
- Increase slowly: more repetitions first, then more resistance.
- Add an endurance activity you enjoy.
- Keep the 24-hour rule in view.
Weeks 7–12: keep going
- Now the effect shows. Compare with your notes from week 1.
- Extend into everyday life: stairs instead of the lift, longer distances on foot.
From week 12: habit
- Look for a format you can sustain over years – a group, a club, a training partner, a studio.
- Plan for setbacks. They are part of it.
A word on judging your own progress: if you feel better after twelve weeks, that was probably partly your training – and probably partly the natural fluctuation from section 6.3. Both at once. That is no reason to be less proud, but a good reason not to throw the whole programme overboard after a setback.
If you are unsure which exercises suit your situation, get guidance. That is exactly what an initial physiotherapy assessment is for.
12. Frequently asked questions
Does resting help in osteoarthritis?
In the short term during a flare: yes, in moderation. As a strategy: no. Rest leads to muscle loss, falling load tolerance and more pain.
I can hear my knee cracking. Is that bad?
Joint noises without pain have no established disease value. They are very common, including in completely healthy joints.
Should I avoid stairs?
No, except during an acute flare. Climbing stairs is strength training in everyday life. If it is too much: more slowly, using the handrail, one floor less – but do not avoid it altogether.
Can cartilage rebuild itself?
Complete restoration is not to be expected in adults. But – and this is decisive – your symptoms and your function can improve considerably without the scan changing at all. The goal of treatment is not a nicer X-ray, but a better life.
Does cold or heat work better?
Whichever you find more pleasant. Both are harmless and relieve symptoms in the short term; neither changes the course.
Do I need cruciate ligament surgery so that I do not get osteoarthritis later?
No. That particular argument does not stand up to scrutiny (see section 5.3). There can be other good reasons for cruciate ligament surgery.
Can physiotherapy prevent a joint replacement?
In many cases it can delay it or make it unnecessary – but not with any guarantee. And if a replacement does come, you go into the operation with better muscles.
In closing
Osteoarthritis is a common, often painful, but rarely as threatening a condition as it feels and as it is often communicated. The images are less informative than they appear. The course is more open than it seems. And the most effective lever lies not in an injection or a supplement, but in what you do regularly.
We have deliberately given you the uncomfortable figures as well – that on average exercise gains around 8 to 10 points on a 100-point scale, that the evidence for the hip is weaker than for the knee, and that a considerable part of any observed improvement is down to contextual factors and statistical regression to the mean. That is more honest than a promise we cannot keep.
What remains is still a great deal: a treatment with small risk, broad benefits on the side, and no alternative that stands up better. And the certainty that you will not harm your joint by loading it.
If you would like to explore the topic playfully: in our game Understanding osteoarthritis you steer a knee joint over four weeks – movement feeds it, rest starves it, overload harms it.
Legal note
This text is general information and does not replace an individual medical assessment. With acute, severe or unusual symptoms – in particular fever, marked swelling, redness, injuries, sudden loss of strength or symptoms unrelated to loading – please consult a doctor.
References
The following works are the foundations of this article. A Digital Object Identifier (DOI) is given only where it could be verified; otherwise the PubMed ID (PMID) is shown, and where neither exists we say so. The links lead to the publishers' pages or to PubMed. Some of these are located outside Switzerland and the EU. When you click, your IP address is transmitted to the provider in question – this does not happen on our own site.
[1] GBD 2021 Osteoarthritis Collaborators. Global, regional, and national burden of osteoarthritis, 1990–2020 and projections to 2050. The Lancet Rheumatology. 2023;5(9):e508–e522. https://doi.org/10.1016/S2665-9913(23)00163-7
[2] Hunter DJ, Bierma-Zeinstra S. Osteoarthritis. Seminar. The Lancet. 2019;393:1745–1759. Kein DOI auf dem vorliegenden Sonderdruck – bitte über den Titel suchen.
[3] Sharma L. Osteoarthritis of the Knee. New England Journal of Medicine. 2021;384(1):51–59. https://doi.org/10.1056/NEJMcp1903768
[4] Daniel M. Boundary cartilage lubrication: review of current concepts. Wiener Medizinische Wochenschrift. 2014;164(5–6):88–94. https://doi.org/10.1007/s10354-013-0240-2
[5] Li H, Yi G, Zhou D, et al. Lubricin in osteoarthritis: functions and therapeutic prospects. Frontiers in Immunology. 2026;17:1790804. https://doi.org/10.3389/fimmu.2026.1790804
[6] Roggio F, Petrigna L, Trovato B, Di Rosa M, Musumeci G. The Role of Lubricin, Irisin and Exercise in the Prevention and Treatment of Osteoarthritis. International Journal of Molecular Sciences. 2023;24(6):5126. https://doi.org/10.3390/ijms24065126
[7] Watkins AR, Reesink HL. Lubricin in experimental and naturally occurring osteoarthritis: a systematic review. Osteoarthritis and Cartilage. 2020;28(10):1303–1315. https://doi.org/10.1016/j.joca.2020.05.009
[8] Guermazi A, et al. Prevalence of abnormalities in knees detected by MRI in adults without knee osteoarthritis (Framingham Osteoarthritis Study). BMJ. 2012. PMID 22932918
[9] Culvenor AG, et al. Prevalence of knee osteoarthritis features on magnetic resonance imaging in asymptomatic uninjured adults: a systematic review and meta-analysis. British Journal of Sports Medicine. 2019. PMID 29886437
[10] Horga LM, et al. Prevalence of abnormal findings in 230 knees of asymptomatic adults using 3.0 T MRI. Skeletal Radiology. 2020. PMID 32060622
[11] Englund M, Guermazi A, Lohmander SL. The Role of the Meniscus in Knee Osteoarthritis: a Cause or Consequence? Radiologic Clinics of North America. 2009;47:703–712. https://doi.org/10.1016/j.rcl.2009.03.003
[12] Johnson AJ, Murillo BC, Cruz-Almeida Y. Pain Sensitivity Profiles Are Associated With Muscle Biomechanics and Clinical Pain in Persons With Symptomatic Knee Osteoarthritis. ACR Open Rheumatology. 2026;8(8):e90118. https://doi.org/10.1002/acr2.90118
[13] Lien-Iversen T, Morgan DB, Jensen C, Risberg MA, Engebretsen L, Viberg B. Does surgery reduce knee osteoarthritis, meniscal injury and subsequent complications compared with non-surgery after ACL rupture with at least 10 years follow-up? A systematic review and meta-analysis. British Journal of Sports Medicine. 2020;54(10):592–598. https://doi.org/10.1136/bjsports-2019-100765
[14] Moseng T, et al. EULAR recommendations for the non-pharmacological core management of hip and knee osteoarthritis: 2023 update. Annals of the Rheumatic Diseases. 2024;83(6):730–740. https://doi.org/10.1136/ard-2023-225041
[15] Bannuru RR, et al. OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis. Osteoarthritis and Cartilage. 2019. PMID 31278997
[16] Systematische Übersicht zu Leitlinien für Hüft- und Kniearthrose. Osteoarthritis and Cartilage. 2023. ScienceDirect S1063458423008324
[17] Hip Pain and Mobility Deficits – Hip Osteoarthritis: Revision 2025. Journal of Orthopaedic & Sports Physical Therapy. 2025;55(11):CPG1–CPG31. https://doi.org/10.2519/jospt.2025.0301
[18] Lawford BJ, et al. Exercise for osteoarthritis of the knee. Cochrane Database of Systematic Reviews. 2024, Issue 12, Art. No.: CD004376. https://doi.org/10.1002/14651858.CD004376.pub4
[19] Hall M, Lawford BJ, Hinman RS, et al. Exercise for osteoarthritis of the hip. Cochrane Database of Systematic Reviews. 2026, Issue 7, Art. No.: CD007912. https://doi.org/10.1002/14651858.CD007912.pub3
[20] Hunter DJ, Hall M. Time to revisit the therapeutic benefits of exercise for osteoarthritis? The Lancet Rheumatology. 2023;5(7):e365–e366. https://doi.org/10.1016/S2665-9913(23)00135-2
[21] Marriott KA, et al. Are the Effects of Resistance Exercise on Pain and Function in Knee and Hip Osteoarthritis Dependent on Exercise Volume, Duration, and Adherence? Arthritis Care & Research. 2024. https://doi.org/10.1002/acr.25313
[22] Roos EM, Kroman S, Ageberg E. An Exercise Therapists' Guide to Neuromuscular Exercise for People With Knee or Hip Osteoarthritis. Journal of Orthopaedic & Sports Physical Therapy. 2025;55(8):518–526. https://doi.org/10.2519/jospt.2025.13041
[23] Englund M, Turkiewicz A. The emperor's new clothes? Osteoarthritis and Cartilage. 2023;31(5):549–551. https://doi.org/10.1016/j.joca.2023.02.001
[24] Englund M, Turkiewicz A. Pain in clinical trials for knee osteoarthritis: estimation of regression to the mean. The Lancet Rheumatology. 2023;5(6):e309. https://doi.org/10.1016/S2665-9913(23)00090-5
[25] Englund M, Turkiewicz A. Regression to the mean in first-line interventions for osteoarthritis: an illusion of effectiveness. Kongressabstract 251. Osteoarthritis and Cartilage. 2022;30(Suppl 1):S191. Kongressabstract, nicht als Vollpublikation begutachtet – kein DOI abgedruckt.
[26] Bliddal H, et al. Once-Weekly Semaglutide in Persons with Obesity and Knee Osteoarthritis (STEP 9). New England Journal of Medicine. 2024;391(17):1573–1583. https://doi.org/10.1056/NEJMoa2403664
[27] Alentorn-Geli E, et al. The Association of Recreational and Competitive Running With Hip and Knee Osteoarthritis: A Systematic Review and Meta-analysis. Journal of Orthopaedic & Sports Physical Therapy. 2017. https://doi.org/10.2519/jospt.2017.7137
[28] Holden MA, Nicholls E, Abdali Z, et al. Provision of knee bracing for knee osteoarthritis (PROP OA): multicentre, parallel group, superiority, statistician blinded, randomised controlled trial. BMJ. 2026;392:e086005. https://doi.org/10.1136/bmj-2025-086005