training-an-arthritic-knee-blog-banner
5.4 million people in the UK have osteoarthritis in at least one knee.
That figure comes from Arthritis UK, the charity formerly known as Versus Arthritis, in its State of Musculoskeletal Health 2025. The same report puts the median age at which symptoms begin at around fifty-five. Not eighty-five. Fifty-five, which is likely the average age of most trainers’ client.
Almost every one of those people has been told the same two words. Wear and tear. It is a phrase that carries a whole mechanical story inside it, and the story tells the client they should probably do less. This article is about why that is the wrong message, what the research says to do instead, and how much of it we as personal trainers can actually deliver.
The phrase implies there is a finite amount of cartilage being used, with every squat wearing it down even more. If that were true, protecting the joint by using it less would be sound advice.
There is now enough evidence for a rebuttal. A 2025 clinical commentary in the Journal of Orthopaedic and Sports Physical Therapy took the belief on directly and concluded that exercise therapy does not appear to harm cartilage structure or quality, either in people who already have knee osteoarthritis or in people at risk of it. The authors go further, recommending that clinicians ask clients directly whether they believe their condition is caused by “wear and tear”, rather than waiting for them to raise the belief themselves (Bricca et al., 2025).
Osteoarthritis is better understood now as a condition of the whole joint. Cartilage, bone, the synovial lining and the muscles that control the joint are all involved, with load sensitivity and low-grade inflammation both playing a part. Cartilage behaves like most other tissue in the body. Remove the load entirely and it can actually be counter-productive.
The other half of the wear and tear story is that running destroys knees, and it is the question you will be asked in the first session by anybody who used to run.
The largest look at this pooled twenty-five studies covering 125,810 people. Prevalence of hip and knee osteoarthritis came out at 3.5 per cent in recreational runners against 10.2 per cent in sedentary controls, with competitive elite runners higher again at 13.3 per cent. Under fifteen years of running exposure carried an odds ratio of 0.60 against non-runners (a 2017 meta-analysis in JOSPT).
Be careful with that. It is observational, so it cannot show cause, and it carries an obvious selection problem. People whose knees hurt stop running, which loads the sedentary group with exactly the people the study is trying to compare against. The defensible line is that there is no good evidence recreational running causes knee osteoarthritis, not that running protects the joint.
One of the clearest tests of whether exercise loading could worsen the joint came from the START trial, published in JAMA in 2021. Three hundred and seventy-seven adults aged fifty and over with radiographic knee osteoarthritis were randomised to eighteen months of high-intensity strength training, low-intensity strength training or an attention control group. Alongside pain, the researchers measured knee joint compressive force.
After eighteen months, compressive force in the high-intensity group was 2453 newtons. In the control group it was 2512. The adjusted difference was minus 58 newtons, with a confidence interval running from minus 282 to 165 (Messier et al., 2021).
Eighteen months of heavy lifting left arthritic knees under fractionally less compressive load than an attention-control programme did, and the difference sat well within normal variation. Whatever heavy loading does to an arthritic knee, grinding it away faster is not on the list.
One note, from the same trial. There were eighty-seven non-serious adverse events, fifty-three of them in the high-intensity group against four in the control group. Heavy work is safe for the joint and still produces more sore backs, shoulders and knees. That is a coaching problem, not a pathology problem.
START’s primary outcome was WOMAC pain on a nought to twenty scale, where a change of two points is roughly the smallest difference a person tends to notice. The high-intensity group finished at 5.1. The control group finished at 4.9. High intensity against low intensity was 5.1 against 4.4, a difference of 0.7 that did not reach significance (Messier et al., 2021).
Everybody improved. Nobody improved because of the load.
Load selection remains important, but not for the reason we might once have assumed. Pick heavy because your client wants to be stronger, because they have a set of stairs or a grandchild or a bag of compost that needs lifting, or because they enjoy it. Do not pick it expecting extra pain relief, and do not avoid it out of a fear that the joint will suffer.
UK research provides one of the clearest estimates of the effect of exercise on knee and hip osteoarthritis, and it suggests the average benefit is meaningful but relatively modest. A team led from Keele University pooled individual participant data from thirty-one randomised trials covering 4,241 people. Set against a non-exercise control, therapeutic exercise reduced pain by 6.36 points on a nought to a hundred scale at twelve weeks, by 3.77 points at six months and by 3.43 points at twelve months. Function followed the same shape and the same decay. The authors describe the effect as of questionable clinical importance, particularly in the medium and long term (Holden et al., 2023).
That is a research group funded by the Chartered Society of Physiotherapy publishing a number that undersells its own profession.
Larger pooled figures do exist. A 2025 network meta-analysis in the BMJ covering 217 trials and 15,684 people produced far bigger effect sizes and ranked aerobic exercise highest overall. The gap between the two is mostly about what exercise gets compared against and how the numbers are pooled, and the most reliable estimate appears to be closer to the individual-participant data.
So what you say to a client is not that exercise abolishes knee pain. It is that exercise produces a modest, real and repeatable improvement in pain and function, at no known cost to the joint, while making them stronger at the same time. Nothing else does all three.
NICE NG226, the osteoarthritis guideline, is short and unusually usable.
For everybody with osteoarthritis, offer therapeutic exercise tailored to their needs, for example local muscle strengthening and general aerobic fitness. Consider supervised sessions. Consider combining therapeutic exercise with an education programme or behaviour change approaches in a structured treatment package. That is the core of it, and exercise is the first thing on the list rather than the last resort.
Diagnosis is clinical. Somebody aged forty-five or over with activity-related joint pain, and either no morning stiffness or morning stiffness lasting no longer than thirty minutes, has osteoarthritis. Imaging is not routinely needed to say so.
There is also a list of things the guideline says not to offer, which is useful to know because clients arrive having bought several of them. Acupuncture and dry needling. TENS, ultrasound, laser and interferential therapy. Glucosamine. Injections of hyaluronan into the joint. Arthroscopic washout or debridement.
The guideline’s best advice is about what to say when it hurts.
NICE asks that people are told joint pain may increase when they start therapeutic exercise, and that doing regular and consistent exercise, even though this may initially cause pain or discomfort, will be beneficial for their joints. It adds that sticking with the plan over the long term increases the benefit.
A client whose knee is a little sorer in week two has not broken anything and does not need the programme abandoned. What they need is us to educate them on what may happen, and let them know that it may happen.
Say it before it happens rather than after. Their experience and perception will then be totally different.
The guideline’s information about pairing exercise with education is more than just a “recommend-you-do”, because it describes the design of the two best-known osteoarthritis programmes in Europe.
The Danish GLA:D programme puts supervised neuromuscular exercise sessions alongside structured education about the condition itself. The UK has ESCAPE-pain, which does the same job across six weeks of twice-weekly sessions. Read on…
ESCAPE-pain has been delivered in community venues by 482 trained exercise professionals rather than by physiotherapists in hospital departments. A 2024 evaluation followed 1,492 participants with a mean age of seventy through the programme and reported improvements in pain, function, quality of life, mental wellbeing and objectively measured physical performance.
Before the programme, 24 per cent were doing at least thirty minutes a week. Immediately afterwards, 78 per cent. Six months later, 69 per cent were still there. The evaluation also put a healthcare saving of £326.16 per participant on it, and the authors state that community delivery was as effective as physiotherapist-led delivery in hospitals (Hurley and Thompson, 2024).
Two things to remember. There was no control group and no randomisation, so natural history and regression to the mean are both uncontrolled, and the activity figures are self-reported. The authors call the work preliminary in the title.
Even with that, it is the clearest published statement we have that a properly trained exercise professional working in a leisure centre can deliver this well. Set 5.4 million knees against an English trauma and orthopaedic waiting list of 852,975 and it becomes obvious that the NHS is not going to supervise all of it. We can help.
An individual participant meta-analysis published in 2024 pooled 1,767 people from ten trials and split them by how long they had had symptoms. Those with a year or less of symptoms were 8.33 points better off on pain at twelve months than those with longer histories, and the same pattern held when the cut-off was moved to two years (van Middelkoop et al., 2024).
This is a subgroup analysis, so treat it as a strong hint rather than absolute. The member who has just started noticing their knee on the stairs is a better prospect than the member who has been quietly managing it for six years, and the moment to act is the moment they mention it in passing.
NICE asks that people with osteoarthritis are told weight loss improves quality of life and physical function and reduces pain, that they are supported to pick their own goal, and that any weight loss helps while losing 10 per cent of body weight is better than losing 5.
The STEP 9 trial, published in 2024, gave weekly semaglutide to 407 people with obesity and knee osteoarthritis for sixty-eight weeks. The drug group lost 13.7 per cent of their body weight and improved WOMAC pain by 41.7 points. But the placebo group is the one that was really interesting. The placebo group received physical activity counselling and a reduced-calorie diet and nothing else, lost 3.2 per cent, and still improved pain by 27.5 points (Bliddal et al., 2024).
Roughly two thirds of the pain improvement in a major pharmaceutical trial came from the thing a personal trainer already does. The trial was funded by the drug’s manufacturer and involved participants with a mean BMI of 40, so caution is needed when applying the findings more broadly.
None of this changes the need to recognise when a client’s knee symptoms do not fit the typical presentation of osteoarthritis. Knowing when to refer on is an important part of safe practice.
| What You See | What to Do | Why |
|---|---|---|
| Hot, red, swollen knee, with fever or the client feeling generally unwell | Same-day medical advice. Do not train the joint | Septic arthritis and crystal arthropathy both look like this and both are urgent |
| True locking, or a knee that will not straighten, after a twisting injury | Refer to a GP or physiotherapist before loading it | Suggests a mechanical block inside the joint rather than osteoarthritis |
| Morning stiffness lasting longer than thirty minutes, several joints involved, night pain or unusual fatigue | Refer for medical assessment | NICE uses the thirty-minute mark to separate osteoarthritis from an inflammatory pattern |
| Joint pain in somebody under forty-five with no injury history | Refer for medical assessment | The clinical diagnosis of osteoarthritis starts at forty-five |
| Recent significant trauma, or an inability to put weight through the leg | Refer. Do not train | Fracture and ligament rupture need ruling out first |
| Unexplained weight loss, night sweats, or a history of cancer | Refer for medical assessment | Pain that is not related to movement needs a medical answer |
| A sudden and severe change in a previously stable knee | Pause loading and refer back | A change in pattern is always worth a second opinion |
With the screening done, the programme itself is less exotic.
Two to three sessions a week, and keep them. Consistency does more here than any single exercise choice, and the twelve-month data backs this up. The effect fades in people who stop.
Build the quadriceps and the hip. Knee extension strength is the variable most consistently linked to function, and the hip abductors control what the knee does underneath the client in single-leg work. Leg press, sit-to-stand, step-ups, split squats and a machine knee extension are all important.
Keep aerobic work in the plan. The BMJ analysis ranked it highest overall, and it is usually the easiest thing to get somebody to keep doing. A bike, a brisk walk, a cross trainer, whatever they will actually turn to do.
Focus on how the knee responds over the following 24 hours, rather than judging the session by pain experienced during the exercise itself. Some soreness can be expected. Soreness that settles by the next morning is a normal training signal. Soreness that is still there two days later, or swelling that was not there before, means the last session was too much and the next one comes down a notch. Let your client know that this is what you are watching out for.
Progress something every few weeks, even when you have backed off a bit. Load, range, tempo, single-leg work, the length of the walk. Osteoarthritis fluctuates and a programme that only moves forward on good weeks stops moving.
Educate, educate, educate. Programmes with the best track record pair the exercise with structured teaching about the condition. A client who understands that the joint is not being spent, that soreness in week two is expected, and that the benefit comes from years rather than weeks is more likely to continue to train on days when you’re not there.
Expect a twelve-week horizon before you judge anything. The pooled data show the largest effect at twelve weeks and a smaller one holding at twelve months, which means the first three months is where the greatest positive changes in behavior happen and the following nine are where it’s kept.
Many of us find our clients arrive with a diagnosis and a belief that wear and tear is the cause. It’s bone-on-bone. It’s age.
The evidence here, however, gives you something better to say. The joint is not being worn out by use. Eighteen months of heavy strength training did not increase compressive loading through the knee joint. Exercise produces a modest but real improvement in pain and function, it works better the earlier one starts, and it works whether the load is heavy or light, which means the load can be chosen around what the person actually wants to do.
The research does not suggest osteoarthritis can be “cured” through exercise. What it does show is that the condition can often be managed effectively over the long term, with the aim of keeping people active, capable and training consistently for years rather than weeks.
If you are someone who has been steering clients away from anything that loads a sore knee, this is the week to stop. There are five and a half million of them, most are still working, and almost nobody has told them they are allowed to get strong.
Bliddal, H., Bays, H., Czernichow, S., Uddén Hemmingsson, J., Hjelmesæth, J., Hoffmann Morville, T., Koroleva, A. and Skov Neergaard, J. (2024). Once-Weekly Semaglutide in Persons with Obesity and Knee Osteoarthritis. New England Journal of Medicine, 391(17), pp.1573-1583. Click here to review the full research article.
Bricca, A., Juhl, C.B. and Roos, E.M. (2025). Exercise Therapy ‘Wears Down’ My Knee Joint: Myth or Reality? Journal of Orthopaedic and Sports Physical Therapy, 55(7), pp.1-5. Click here to review the full research article.
Holden, M.A., Hattle, M., Runhaar, J., Riley, R.D., Healey, E.L., Quicke, J., van der Windt, D.A. and Dziedzic, K. (2023). Moderators of the Effect of Therapeutic Exercise for Knee and Hip Osteoarthritis: A Systematic Review and Individual Participant Data Meta-Analysis. The Lancet Rheumatology, 5(7), pp.e386-e400. Click here to review the full research article.
Hurley, M. and Thompson, F. (2024). Community-Based Care for People with Chronic Knee and Hip Pain: Preliminary Clinical Outcomes and Healthcare Utilisation for ESCAPE-Pain. Musculoskeletal Care, 22(1), e1847. Click here to review the full research article.
Messier, S.P., Mihalko, S.L., Beavers, D.P., Nicklas, B.J., DeVita, P., Carr, J.J., Hunter, D.J. and Lyles, M. (2021). Effect of High-Intensity Strength Training on Knee Pain and Knee Joint Compressive Forces Among Adults with Knee Osteoarthritis: The START Randomized Clinical Trial. JAMA, 325(7), pp.646-657. Click here to review the full research article.
van Middelkoop, M., Schiphof, D., Hattle, M., Simkins, J., Bennell, K.L., Hinman, R.S., Allen, K.D. and Knoop, J. (2024). People with Short Symptom Duration of Knee Osteoarthritis Benefit More from Exercise Therapy Than People with Longer Symptom Duration: An Individual Participant Data Meta-Analysis from the OA Trial Bank. Osteoarthritis and Cartilage, 32(12), pp.1620-1627. Click here to review the full research article.
Osteoarthritis affects millions of people in the UK, and exercise is a core part of its long-term management. But many people need more support than a brief appointment can provide to turn that advice into a safe, practical programme they can follow consistently.
Our Level 3 Exercise Referral qualification prepares you to work with clients referred with a range of common health conditions. You’ll learn how to screen and risk-stratify clients, adapt exercise around medication and comorbidities, and communicate appropriately with referring health professionals. Available through distance study, live-virtual or in-person learning.
Great Discounts: Get started on the UK's leading Exercise Referral course and save £100 when you pay in full upfront. Use Promo Code EXREF20 at checkout. Priced from £399, or £83/month. Offer Ends 4th August!
Level 3 Exercise Referral – In-Person, Live-Virtual & Distance Study
A great group exercise class delivers far more than a workout. There is the exercise…
Few beliefs in fitness are held as firmly as the idea that a tight muscle…
Twenty-nine systematic reviews. Eight hundred and sixty-three separate trials. Twenty-four different treatments, all pointed at…
Many of us have a client who has started moving differently. They take the stairs…
Few moments in a client's life involve as much physical change, and as much uncertainty…
Most of us have worked with a client who was having a hard time with…