Cartilage Injuries in Runners: Recovery and Rehabilitation

From Fellrnr.com, Running tips
Jump to navigationJump to search

"Cartilage damage" sounds like the end of a running career, but the phrase covers several different problems with different outlooks. A small, stable defect in the smooth cartilage that lines a joint, the gradual changes of osteoarthritis, a torn meniscus, and an injury that involves the bone beneath the cartilage each need a different approach. Many runners with cartilage problems keep running or return to it. Some first need a period of protection, meaning loading that a clinician has restricted: for example, crutches to limit how much weight goes through the leg, avoiding impact or particular movements, and occasionally a brace to limit how far the joint bends. Others need specialist treatment, and a few need surgery. Pain alone cannot tell you which group you are in.

The main points are:

  • Find out what you are dealing with. The diagnosis matters more than the word "cartilage". See which situation are you in? and diagnosis, scans and injury grades.
  • Know the warning signs. Get help the same day if, after an injury, your knee becomes stuck and will not straighten, if you cannot put weight on the leg, if the joint swells rapidly and markedly after an injury, if a joint is hot and swollen, or if worsening hip or groin pain when you put weight on the leg raises the possibility of a stress fracture. See when to seek medical help.
  • Feeling better is not the same as regrowing cartilage. Pain, function and the ability to run can all improve a great deal without the damaged tissue being rebuilt. Adult joint cartilage adapts to loading but has very little ability to regrow. See can cartilage heal?
  • Running is not simple wear and tear. Recreational running has not been shown to cause knee osteoarthritis, and the changes seen in cartilage straight after a run are small and temporary. A previous significant joint injury is a clearer risk factor. See does running wear out cartilage? and activity, inactivity and osteoarthritis.
  • Change the load; don't default to complete rest or to carrying on regardless. Reduce what provokes the joint, keep what it tolerates, and watch swelling, movement, walking and your running form as well as pain. See can I keep running? and monitor the joint's response.
  • Rebuild in stages. Comfortable movement first, then strength, then impact, then running, then full training. See the practical guide.
  • Cycling is useful, but it is not a proven repair treatment. Gentle cycling is a sensible way to keep a knee moving and to stay fit, and it featured in some of the meniscus rehabilitation programmes that have been studied. No trial was found showing that a particular cycling dose heals a torn meniscus or a knee or hip cartilage defect. See cycling.
  • Be sceptical of anything sold as regenerating cartilage. Injections and supplements may ease symptoms for some people; none has been shown to reliably regrow joint cartilage. Surgery helps selected injuries. See additional treatments.
  • I regained the ability to run after a meniscus tear and returned to ultramarathons. I describe what I did in my experience after a meniscus tear. It is one person's story, not a tested programme.
How to use this article. This article is educational information, not individual medical advice or a diagnosis. Use it as a starting point for your own research and as food for thought when you discuss your options with a suitably qualified clinician. It combines published evidence with clearly labelled personal experience and hypotheses. What helped me may not suit your injury, and feeling better does not necessarily mean that damaged tissue has healed. Reading about an injury does not replace having it assessed: this guide cannot stand in for an examination of persistent, severe or unusual symptoms, and it does not override the instructions you have been given after surgery.

Disclosure: this article is the result of multiple iterations of ChatGPT, Claude, and my analysis/review. Any mistakes are mine alone.

How the evidence is labelled. Where a label helps, claims in this article are marked as follows. Each label applies to the particular claim, not to a whole treatment.

Strong evidence
Consistent, good-quality studies for that specific outcome.
Limited evidence
Real uncertainty about benefit, the size of the effect, who it applies to, or harms.
Clinical consensus
Agreement among experts, not necessarily tested in trials.
Mechanistic hypothesis
A plausible biological explanation that has not been shown to produce the claimed result.
Personal experience or opinion
My own account or view. It is not proof that something works.

When to seek medical help

Get help the same day if any of these apply:

  • After an injury, your knee becomes stuck and will not straighten. Do not force the movement.
  • You cannot walk or put weight on the leg.
  • The joint swells rapidly and markedly after an injury.
  • The joint is hot and swollen, whether or not you have a fever, or you have joint pain with a high temperature or feel shivery and unwell. An infected joint is an emergency[1][2].
  • Worsening hip or groin pain when you put weight on the leg raises concern about a stress fracture of the neck of the thigh bone (femoral neck), particularly if walking now hurts or you limp. Stop running, keep weight off the leg and seek assessment today. An early X-ray can look normal; if the suspicion persists, an MRI may be needed[3][4].

Go to an emergency department straight away if the joint looks deformed or out of place, or if you have tingling or loss of feeling in the leg or foot[1].

Where to go.

  • UK: ask for an urgent GP appointment or call NHS 111. Use A&E for the emergency signs above[1].
  • US: call your primary care doctor for a same-day appointment, or go to an urgent care centre that can take X-rays. Use the emergency room, or call 911, for the emergency signs, for a suspected joint infection, or if you cannot be seen the same day[5][2].
  • Elsewhere: the principle is the same. Contact your usual doctor or local urgent or after-hours service the same day, and use the emergency department or your local emergency number for the emergency signs.

Arrange a timely, non-urgent assessment if any of the following apply:

  • The joint swells repeatedly, or swelling does not settle.
  • The knee gives way, or catches briefly and then releases. If it stays stuck, use the urgent route above.
  • You are gradually losing movement, or cannot fully straighten the joint.
  • You had a significant injury and the knee has since caught or briefly locked. This can mean a repairable meniscus tear, and a prompt specialist referral is appropriate[6].
  • Walking, stairs or daily activities are getting worse rather than better.
  • Pain is present at rest or at night, or you have a tender spot on a bone. Bone stress injuries can mimic joint problems[7][8]. Hip or groin pain on weight bearing belongs on the urgent list above.
  • Symptoms have not improved after a few weeks of sensible changes.

You do not need a scan to start sensible rehabilitation of an ordinary sore knee. Most people with knee pain and no warning signs, recent significant injury or locking do not need an MRI at the outset[6].

Who to see

  • A GP (UK) or primary care doctor (US) is the usual starting point for an undiagnosed painful or swollen joint, and can arrange investigations or referral.
  • A sports physiotherapist (UK) or physical therapist (US) can examine the joint, guide rehabilitation of an uncomplicated problem, and recognise when you need a medical opinion. In many parts of the UK you can refer yourself to NHS musculoskeletal services[1]. Every US state allows some direct access to a physical therapist without a referral, although some states set limits and insurers may still require one[9].
  • A sports and exercise medicine doctor (UK) or primary care sports medicine physician (US) is a non-surgical specialist. They are helpful when the diagnosis is uncertain, when imaging or injections are being considered, or when you need advice that balances the injury against your sport.
  • An orthopaedic surgeon is the right person for a knee that is stuck, a suspected loose or unstable fragment, significant injury to the bone beneath the cartilage, a meniscus tear that may be repairable, or disability that continues despite good rehabilitation.

My view on when to see a surgeon. For a non-urgent joint problem, I think it is better to see a good sports doctor or experienced sports physiotherapist first, and to see a surgeon when they advise it. My concern is that a surgeon may be more likely to recommend surgery than is ideal. There is some evidence behind this concern, although none of it tests my approach directly:

  • Doctors tend to recommend the treatment they deliver. When urologists (surgeons) and radiation oncologists each saw the same 108 men with localised prostate cancer, 79% of the urologists' recommendations were for surgery and 68% of the radiation oncologists' were for radiation; the two specialists' recommendations agreed for only 31% of the men[10]. No equivalent study of knee surgery was found.
  • Surgeons often disagree about whether to operate on the same knee, particularly when osteoarthritis is present[11][12].
  • Keyhole surgery for degenerative knees remained common long after trials showed it was no better than exercise. An international panel now makes a strong recommendation against it for nearly all degenerative knee disease[13]. In England, rates of keyhole meniscus surgery rose for many years before falling, with wide variation between regions[14].

On the other side, physiotherapists who triage orthopaedic referrals largely agree with surgeons about who needs an operation (70% to 93% agreement in cohort studies, with low to moderate certainty)[15], and some problems need a surgeon's opinion promptly: a knee that is stuck, a suspected displaced or repairable meniscus tear (including root tears in younger people)[16], an unstable fragment or a fracture. Waiting for a sports doctor should not delay those. So the evidence gives my preference indirect support for non-urgent problems (limited evidence); it is not a reason to avoid a surgeon when one is needed. If a surgeon recommends an operation for a non-urgent problem, it is reasonable to ask what is likely to happen with rehabilitation alone, and to seek a second opinion.

My own experience. In June 2016 a surgeon proposed removing the inner part of my torn meniscus, and I declined. The surgeon did not suggest that the operation would probably lead to a full recovery. My tear was worst in the outer part, and I was concerned that removing the inner part would increase the stress on the damaged outer edge. I was not convinced that the tear was the main cause of my pain, which was further forward in the knee than the tear. My knee was already improving slowly, and I thought the tear might settle down on its own. And any operation carries risks, including the small but serious risks of an anaesthetic. Six months later I was glad I had not had surgery, and I went on to run ultramarathons. That is one person's decision about one tear. It is not evidence that surgery would have been the wrong choice, and the trials described in what about the meniscus? are a better guide (personal experience).

Practical guide to rehabilitation

This guide is a practical synthesis. Very little of it has been tested in trials of runners with cartilage injuries, so where a number is given it is a starting point or an example, not a validated dose. If you have had surgery, your surgeon's and physiotherapist's instructions override everything here.

Which situation are you in?

Flowchart. Any urgent warning sign leads to same-day assessment. Otherwise, without a diagnosis, reduce what provokes the joint and get assessed if it is not improving. With a diagnosis: a stable focal defect follows staged rehabilitation; degenerative change or osteoarthritis means staying active and building strength; bone involvement, an unstable fragment or recent surgery means a specialist-led plan.
Which pathway? Urgent warning signs come first; the pathway then depends on the diagnosis.
Situation Starting approach Running Get help sooner if
Symptoms, but no diagnosis Check the warning signs. Reduce whatever provokes the joint and keep comfortable activity going. Arrange an assessment if it is not clearly improving within a few weeks A short reduction is reasonable. Do not run with swelling, a limp, a joint that will not move as usual, or hip or groin pain when you put weight on the leg There was a significant injury, swelling keeps returning, pain feels as though it is in the bone, or the joint catches or gives way. A knee that becomes stuck, or hip or groin pain on weight bearing, needs the urgent route
Confirmed stable focal defect (a localised area of cartilage damage, with the fragment and underlying bone judged stable) Agree any loading limits with your clinician, then follow the staged programme below An individual decision. Some runners continue with modified running; others need a period without impact Swelling or restriction returns, or you are no better after a few months of consistent rehabilitation
Degenerative changes or osteoarthritis Keep active, build strength, and address anything relevant such as excess body weight Running that you choose and tolerate may continue; follow the trend over weeks, not single days Symptoms change character, function declines, or pain no longer settles
Osteochondral injury (cartilage plus bone), osteochondritis dissecans, an unstable fragment, or after surgery Specialist-led plan for movement, weight bearing and impact Do not clear yourself to run on symptoms alone Follow the healing and procedure-specific criteria you have been given

A meniscus tear depends on its type; see what about the meniscus?

Can I keep running?

There is no pain score that tells you a cartilage problem is safe to run on. Pain can be mild with a significant bone or fragment problem, and it can be quite marked in long-standing osteoarthritis without meaning that harm is being done. Base the decision on the diagnosis and on how the joint behaves.

Once your diagnosis permits running, useful checks include:

  • Walking normally, including on stairs.
  • Movement back to your usual range, or to the range agreed with your clinician. Someone with long-standing osteoarthritis may not have had equal movement in both knees for years, so matching the other side is not always the right test.
  • No new or increasing swelling.
  • Enough strength and control for the running you plan.
  • Symptoms during and after running that are mild and are not building up from run to run.
  • Being able to hold your normal running form (see below).

Stop running and get assessed if the joint locks or gives way, if it swells noticeably, if you lose movement, or if you suspect a bone injury.

Running form. My view is that if an injury changes how you run, so that you cannot hold your normal form, that is a reason to stop the run and drop back, even if it does not hurt much. This is stricter than "no limp". A limp is obvious to anyone watching, but subtler changes, such as shortening the stride on one side or landing differently to protect the joint, are often noticeable to the runner first. The evidence here is indirect. In laboratory studies, experimentally induced knee pain changed walking and running mechanics in ways that may alter how the knee is loaded[17], and runners with kneecap pain changed how they moved once pain developed during a run[18]. No study has tested whether running with altered form causes harm, so this is personal opinion, not a tested rule. Treat it as a reason for caution: stop the run, walk, and try a shorter or easier run once your form is normal again. If your form keeps changing, get assessed.

The best evidence that continuing to run can be acceptable comes from people with knee osteoarthritis. In people aged 50 or older with knee osteoarthritis who chose whether and how much to run, runners showed no greater worsening on X-ray over four years[19]. Because people chose for themselves, this tells us little about starting running, high mileage or other diagnoses. If you have osteoarthritis, NICE advises that joint pain may increase when you start therapeutic exercise, and that regular, consistent exercise will nevertheless benefit the joints, with the benefits growing the longer you keep it up[20]. That applies to established osteoarthritis, not to a recently injured joint. Evidence for runners with a symptomatic focal defect is limited, so that decision deserves individual advice.

Monitor the joint's response

Check the joint at three points: during the activity, later the same day, and the next morning. Use the same few checks each time:

  • Swelling. Compare the outline of the joint with the other side. Swelling that appears or increases after activity is one of the most useful signs that the load was too much.
  • Movement. Can you straighten and bend the joint as far as usual? Loss of your usual straightening matters more than a pain score.
  • Walking. Is your walking normal, including on stairs?
  • Stiffness. How long does the joint take to loosen up in the morning?
  • Pain. Note the trend over days. A 0–10 score is useful for tracking change, but no particular number has been shown to mark the boundary of safety for cartilage.

These checks are practical feedback. They are not validated tests of what is happening inside the joint: a quiet morning does not prove that the cartilage was unharmed, and a sore one does not prove that it was damaged. Expert consensus on rehabilitating both meniscal and cartilage injuries relies on the same kind of signs (swelling, range of movement, strength and function), not on a pain threshold[16][21].

Modify training and maintain fitness

You have several things you can change: how far, how often, how fast, and on what terrain. Nobody has established a correct order in which to change them for cartilage problems, so start with whatever most reliably provokes your joint.

  • Shorten runs before abandoning them, if the joint tolerates shorter ones.
  • Add recovery days if symptoms build up across consecutive days. See How Often To Run.
  • Drop the provoking sessions. For many people these are downhills, fast running or uneven ground, but your joint may differ.
  • Use flat, predictable surfaces while you are testing what the joint tolerates, so that you change one thing at a time.
  • Keep fit another way. Cycling, pool running, swimming and the elliptical trainer can maintain fitness while impact is reduced. Choose by what your joint tolerates and what you have access to. Kicking hard in the pool or pedalling against high resistance can still aggravate a joint.

Fitness and joint tolerance are separate things. Cross-training can keep your heart and lungs race-fit while the injured joint is still only ready for short walks. Let the joint, not your fitness, set the pace of your return.

Restore comfortable movement

A joint that is kept still tends to stiffen and weaken. Unless you have been told to protect the joint, move it regularly through the range that is comfortable.

  • Bend and straighten the joint gently several times a day.
  • Easy cycling on a stationary bike is a convenient way to move a knee through a repeated, controlled range. See cycling.
  • Walking within comfortable limits keeps normal loading going.
  • Do not force the joint. A knee that has become stuck after an injury and will not straighten needs urgent same-day assessment. Recurrent brief catching that releases, or a gradual loss of movement, also deserves assessment.

Protection takes priority when the bone is involved, when a fragment may be unstable, or after surgery. In those cases follow the limits on weight bearing, movement and impact that you have been given, including any brace.

Build strength and control

Strong muscles do not regrow cartilage, but they are one of the few things you can improve that reliably helps how a joint functions. Work on the front of the thigh (quadriceps), the hips, the hamstrings and the calves.

  • Choose exercises the joint tolerates. Leg press, squats to a comfortable depth, step-ups, knee extensions, bridges, hamstring curls and calf raises are all reasonable options.
  • Use the comfortable range. The part of the joint surface under load changes with the joint angle, so an exercise that is fine in one range may be sore in another. A shallower range is a reasonable starting point; it does not need to be permanent.
  • Start modestly and progress. As an example, two sessions a week of one or two sets of controlled repetitions at a moderate effort, building up as the joint allows. These numbers are a convenient starting point, not a cartilage-specific dose.
  • Add balance and single-leg control once basic strength is returning.
  • No method has been shown to be best for cartilage. Isometric holds can be a comfortable way to begin, and heavier lifting is fine if the joint tolerates it, but neither has been shown to be superior for cartilage problems. In a large trial in knee osteoarthritis, high-intensity strength training was no better than low-intensity training for pain[22]. The specific loading recipes used for tendon injuries should not be assumed to apply.

Return to impact and running

Consider starting to run again once your diagnosis permits impact and the joint is coping with daily life. Useful checks include walking normally, movement back to your usual or clinician-agreed range, no new or increasing swelling, and sufficient strength and control for the activity. Some people use a comfortable brisk walk of around 30 minutes as a convenient benchmark; there is no evidence that 30 minutes is better than 20 or 40.

Experts agree that swelling, movement, strength and function should all be considered[21][16], but none of these checks has been validated as predicting a safe return to running for a cartilage injury managed without surgery. The cartilage consensus was developed mainly for people who have had surgery. If a bone injury is suspected, do not use hopping or jogging as a home test.

Then reintroduce running gradually:

  • Start with run–walk. For example, a few one-minute easy jogs separated by two minutes of walking, on a flat surface. This is only an illustration; for some joints even that is too much at first. See Walking Breaks.
  • Consider a day between runs at first, so that you can see the delayed response before the next one.
  • Change one thing at a time. Lengthen the running sections before adding speed, hills or rough ground.
  • Hold or step back if the checks in monitoring worsen, or if you cannot hold your normal running form.

Ready-made beginner programmes such as the NHS Couch to 5K, with three runs a week and a rest day between each, give a gradual structure[23]. They were designed to get beginners running, not to rehabilitate joints, so let the joint's response, not the programme's schedule, decide when you move on.

Handle flare-ups and setbacks

Flare-ups are common and usually mean the load got ahead of the joint, not that new damage has been done. Use the response during exercise, later that day and the next morning to decide what to do.

What you notice What to do What it means
Symptoms absent or mild; movement normal; no new swelling; running form normal; back to your usual state by the next morning Progress. Repeat the session, then make one small increase A reasonable sign of tolerance. It does not prove that the tissue is unharmed
Mild symptoms that you are unsure about; function unchanged Hold. Repeat the same level and watch a few more sessions Avoid progressing while you are uncertain
Clearly worse the next day; more stiffness; a limp or a change in your running form; new or increased swelling Reduce. Reduce the provoking activity and reassess the response The load got ahead of the joint. If this keeps happening, or does not settle, get reassessed. A new block, marked swelling, inability to bear weight or symptoms suggesting a bone injury change the action: see the next row
Knee stuck and will not straighten; giving way; marked swelling; cannot bear weight; hot joint; hip or groin pain on weight bearing Seek assessment. Stop, and see when to seek medical help This is beyond self-managed rehabilitation

Return to full training and racing

Being able to run again is the start of rebuilding, not the end.

  • Build repeatable easy running first. Aim to tolerate your normal easy runs, week after week, before adding anything harder.
  • Reintroduce one demand at a time. Long runs, speedwork, hills and rough terrain each load the joint differently. Add them separately so that you can tell which one causes trouble. See Safe Speedwork.
  • Race only after event-relevant training is repeatedly tolerated. That means training that resembles the race in distance, terrain and pace has gone well more than once, and any restrictions from your clinician have been lifted.
  • Expect your fitness to be ahead of your joint. If you have cross-trained well, your aerobic fitness will support more running than the joint has been prepared for. See Training Readiness.
  • Some runners need lasting changes to volume, terrain or racing. Others do not. A scan that still looks abnormal is not, by itself, a reason to stop for good.

Fellrnr's Rehab Approach: my experience after a meniscus tear

This is what I did to recover from a torn meniscus. It is a personal account; the sections above, not this one, are the general guidance. The dates come from my training log and activity records.

Read this first: what this account is, and who it is not for.

  • This is my experience with a meniscus tear that a sports doctor diagnosed using MRI. It is not a validated programme for meniscus tears, and it is not a cartilage-repair programme.
  • Symptoms cannot tell you what kind of tear or lesion you have, or whether an approach like this suits it. My tear was in the rear (posterior) part of the meniscus, worst in the outer part but extending through to the inner edge, and it never caused my knee to lock. Other details of the tear are not described here, which limits how far my experience applies even to other meniscus tears.
  • If you want to consider something similar, treat it as something to discuss with a clinician who has assessed you and judged that rehabilitation without surgery is appropriate. That does not mean everyone needs an MRI. It does mean that this is not a way to diagnose or clear yourself.
  • Do not use this on your own if your knee locks, catches repeatedly, or may have a displaced fragment; after a major injury or with major swelling; if you cannot bear weight; if infection, a fracture or a bone stress injury is suspected; if you have an unstable injury involving cartilage and bone; if the joint is significantly unstable; or if you have restrictions after surgery. See when to seek medical help.
  • Some tears deserve a specialist opinion before any plan is chosen, because early repair may matter: tears of the meniscus root, complete radial tears, bucket-handle tears and other tears that may be repairable[16]. Not all of these need surgery, but being able to exercise without pain does not show that this approach is suitable for them.
  • Hip pain: worsening hip or groin pain when you put weight on the leg needs urgent assessment for a stress fracture[3]. Do not use this knee story to clear yourself to run or to conclude that you have a cartilage problem; other causes of hip pain need different management.
MRI scans of the author's knee. The torn meniscus shows a light area within its normally dark, triangular outline.
My knee MRI from May 2016. The tear shows as a light area within the dark triangle of the meniscus.

What happened

  • The injury. I injured my left knee in November 2015, during a period of heavy training that included marathon-distance training runs, so my underlying fitness was far higher than most people's. The pain was on the inner (medial) side of the knee, and within days I had to abandon a run because the pain felt wrong.
  • A slow diagnosis. For several months I kept running at reduced volume, with setbacks. I suspected the inner ligament or the tendons on the inner side of the knee, and one sports doctor thought the most likely cause was an inflamed bursa there (pes anserine bursitis). In May 2016 a sports specialist doctor, recommended as an expert in diagnosing injuries rather than a surgeon, diagnosed a meniscus tear, and an MRI confirmed it. The tear was in the rear (posterior) part of the meniscus. It was worst in the outer part, which has a blood supply, but extended through to the inner edge. My knee never locked.
  • Declining surgery. In June 2016 I saw a surgeon, who proposed removing the inner part of the meniscus. I declined; my reasons are under who to see.
  • Other things I tried. In the early months I also did some strengthening exercises and worked on tight and sore muscles around the knee, including trigger points. Later I used a treadmill desk, because my knee seemed to do much better with movement than with sitting still.
  • Setbacks. Recovery was not smooth, and doing too much too soon made things worse more than once. In September 2016, after a trip abroad, my knee became painful and badly swollen after my first run at home; I cut back to short daily runs and plenty of walking for three weeks, then rebuilt. In March 2017 a deep squat upset the knee again, but it settled within a few weeks.
  • The outcome. By late 2016 my knee was no longer limiting my running, although it still did not like being fully bent. In April 2018 I completed a 50-mile race, and I have since run hilly ultramarathons. Since then the knee has given me only occasional minor twinges.

The cycling and walking approach

I did not record this part of my recovery, because the rides were so short and easy that I did not log them. I cannot date it or say how long each stage took. Neither my starting point nor my eventual outcome should be treated as typical.

  1. My starting point. At the worst, I could not run, and walking was painful.
  2. My symptom rule. I allowed mild "discomfort" but no pain. By discomfort I mean a change from normal sensation or from how the movement normally felt, such as a sense of pressure, without pain. That is my own vocabulary, not a validated pain scale.
  3. Gentle stationary cycling. I began with 15 minutes on a stationary bike, about three times a day, spread roughly evenly through the day, at about 1 watt per kilogram of body weight. These numbers were arbitrary, practical choices. A lower resistance than about 1 W/kg felt uncomfortable, and that level felt more natural; to me it felt almost effortless, like coasting. I had an indoor bike at home, and the effort was low enough that I did not need a shower afterwards. With a sedentary job, I wondered whether moving my knee several times during the day might help fluid exchange, and three short rides also fitted my working day.
  4. Why I chose this. I thought that gentle, repeated loading and unloading might encourage fluid to move in and out of the joint tissues. That was my reasoning, not an established explanation for my recovery.
  5. Why indoors. I preferred a stable stationary bike because I could control the resistance, avoid hills, traffic, uneven surfaces and unexpected exertion, and stop pedalling promptly.
  6. Introducing walking. Once I could walk around the house without discomfort, I began walking outdoors and reduced the cycling to once a day.
  7. My walking progression. I started at about half a mile (800 m), which felt small compared with my training before the injury. My first increase doubled the walk to roughly 1.6 km. After that, following a comfortable walk, I increased by roughly 800 m at a time, working towards about three miles (roughly 5 km). If I had only very minor discomfort I repeated the same distance. If symptoms increased I reduced the distance the next day. I took account of how the knee felt through the whole day, not just during the walk.
  8. My first running. Once I could walk about three miles comfortably, I added a quarter-mile (400 m) run in the middle of the walk. I gradually added more short running sections, increasing when it was comfortable, holding when there was only very minor discomfort, and backing off when symptoms increased.
  9. Building to 5 km. From there I followed a gradual run–walk progression similar in principle to Couch to 5K, increasing the running and reducing the walking until I could run 5 km comfortably.
  10. Returning to training. When I could run 5 km I stopped the rehabilitation cycling and cautiously rebuilt my normal running, allowing for the fitness I had lost.
  11. What this does and does not show. This is what I did and the recovery I experienced. It cannot show that the approach caused the recovery, or that the same approach would suit anyone else. My tear was worst in the part of the meniscus that has a blood supply but extended into the part that does not, and I was never sure the tear was the main source of my pain: much of the pain was further forward than the tear, and the MRI also showed some unexplained fluid (oedema) in the kneecap. So my recovery cannot show whether the tear itself healed; being able to run again does not tell us that. Different injuries and different people need different approaches.

Notes for readers on the steps above

  • Discomfort is not the only thing to watch. My rule was about sensation. A knee that gives way, locks, swells, loses movement or makes you limp is deteriorating in a way that matters even if it does not hurt, and none of those should be accepted as "just discomfort".
  • Strength training played only a small part. I did some strengthening exercises in the early months, but strength work was not a large part of my recovery. The general guidance includes it because progressive hip and knee strengthening is a standard part of exercise-based care for meniscus tears[16]. My experience does not show that it is unnecessary.
  • The numbers were mine. 800 m is not a trivial distance for an injured joint, and my running background does not make my increments right for anyone else. I am not suggesting that you add 800 m a day. Likewise, what felt like coasting at 1 W/kg to me may not suit another knee; choose the resistance by comfort, not by a number.
  • A stationary bike is not automatically safe. Even indoors, bike fit, getting on and off, and many repetitions can aggravate a knee. Set the saddle so that the knee is not forced into deep bending. A bike on balance rollers is not the same as a stable stationary bike.
  • Keeping walks close to home is worth considering, so that you can stop without a long walk back. Short loops near home or a treadmill are two ways to do that. My own walks in the progression above were outdoors; the treadmill desk I mention was for gentle movement during the working day. A treadmill has a moving belt and its own balance demands, so it is not automatically the safer choice.
  • Next-morning checks are a sensible addition. See monitoring.
  • Stopping the cycling when I could run 5 km was my choice. It is not a necessary milestone, and it is not evidence that the tissue had healed.

How this compares with the evidence

The two tables below set each thing I did against what has been studied. "Not established" and "no direct study located" mean that no study showing it was found; they do not mean that it has been shown not to work.

Overall approach and cycling
What I did, and why Direct human clinical evidence Mechanistic reasoning Limits, and what it means for you
Rehabilitation without surgery (I declined the operation a surgeon proposed) What trials tell us about exercise-based care for selected meniscus tears: for degenerative tears without locking, physiotherapy was not inferior to keyhole surgery for knee function over five years[24] (strong evidence for that group). In adults aged 18 to 40 who were candidates for surgery, early surgery was not superior to exercise and education, with surgery available later, on a combined score of pain, symptoms, sport and quality of life at one year; 26% of the exercise group later had surgery[25]. In 100 adults aged 18 to 45 with traumatic tears, excluding locked knees and tears that looked repairable, function scores were identical at two years[26] (limited evidence) The outer part of the meniscus has a blood supply and some capacity to heal; the inner part has limited healing potential[27]. My tear was worst in the outer part but extended to the inner edge Does not apply to every tear. Root, complete radial, bucket-handle and other repairable tears may need earlier surgery[16]. My recovery says nothing about whether the tear itself healed
"Discomfort but no pain" rule Not established. No pain or discomfort threshold has been validated for meniscus or cartilage rehabilitation. Consensus statements rely on swelling, movement, strength and function[16][21] Cartilage itself has no nerve supply, so sensation is an indirect guide to what the tissue is experiencing[28] A reasonable personal brake, not a safety test. Watch swelling, movement, walking and running form as well
Low-load stationary cycling Limited evidence that stationary cycling reduces pain in knee osteoarthritis[29]. Cycling featured in some studied meniscus rehabilitation programmes, such as a five-minute warm-up on a stationary bike in one trial[30], but the studies cannot isolate its contribution. No trial was found showing that about 1 W/kg for 15 minutes three times a day heals a torn meniscus or a knee or hip cartilage defect Measured forces inside the knee are lower during easy cycling than during walking[31] A sensible way to move a knee and keep active. Whether it speeds recovery from a meniscus tear is not established
About 1 W/kg No direct study located. In older adults with knee osteoarthritis, ten weeks of cycling at 40% or 70% of heart-rate reserve both improved pain and function, with no significant difference between them[32]. That cannot show the two intensities are equivalent, or that a minimum or an optimum exists In the conditions tested, knee force rose with pedalling power[31]. Watts measure the power produced at the pedals, not the stress on a particular patch of cartilage. Nothing indicates that a minimum power is needed for fluid to move Chosen because lower resistance felt uncomfortable and about 1 W/kg felt natural, almost like coasting. That was a comfort choice, not a dose. Choose resistance by comfort
15 minutes at a time No direct study located comparing durations for a meniscus or cartilage injury. One surgical team describes 10 to 15 minutes of daily cycling during a phase of recovery from cartilage surgery; that is their practice, not a trial of the dose[33] None that identifies a particular duration Arbitrary. Shorter is fine, and may be necessary at first
Three sessions a day, because my job was sedentary Not established. No study was located comparing three short sessions with one session of the same total length for any joint-structure outcome. In a single-visit study of 37 older adults with knee osteoarthritis, five two-minute bouts of moderate cycling, separated by two minutes of seated rest, led to less pain and fatigue immediately afterwards than ten continuous minutes[34]. That does not test rides spread across a working day Long periods of joint immobility are generally unhelpful, and movement spread through the day is plausible. That is a long way from showing that three sessions heal better than one Split sessions are reasonable if they suit you and the joint tolerates them. Do not expect them to heal the joint faster
Stationary, not outdoor, cycling One trial in 82 people with knee osteoarthritis and type 2 diabetes compared supervised outdoor cycling with stationary cycling indoors. Two intervention-related falls with minor pain and bruising occurred outdoors, and one intervention-related event indoors[35]. With so few events, in a different population, the trial cannot tell us whether either is safer for an injured runner A stationary bike removes traffic, road surfaces, hills and unplanned efforts, and lets you stop at once. Bike set-up changes the forces and movement at the knee[36] For gentle rehabilitation, a stable stationary bike gives better control of effort and removes traffic and road-surface hazards. That makes it a sensible practical choice, although it can still aggravate a joint and has not been shown to make this programme safe for every injury
Walking, running and outcome
What I did, and why Direct human clinical evidence Mechanistic reasoning Limits, and what it means for you
Starting to walk outdoors once walking indoors was comfortable No direct study located for this particular trigger. Restoring normal walking before running is standard practice Walking loads the knee more than easy cycling[31], so it is a logical next step A sensible sequence, with nothing special about the exact trigger
Increasing walks: a first doubling to 1.6 km, then about 800 m at a time No direct study located None. The step size reflected my running background Not a recommendation. Choose steps that suit your own joint, and do not increase automatically each day
Walking about three miles (5 km) before running Not established as a readiness test. Consensus favours several criteria, such as swelling, movement and strength, over a single distance[16][21] Comfortable sustained walking shows that the joint tolerates repeated low-level loading A personal milestone, not a test or a threshold
First runs of about 400 m within a walk No direct study located Short running sections limit the total impact while you find out how the joint responds A personal milestone, not a minimum or a validated test. For some people a shorter first run is more appropriate
Couch-to-5K-style run–walk progression Not tested for meniscus tears. Couch to 5K is a beginners' running plan[23] Gradual, repeated exposure with recovery between runs is a reasonable way to rebuild tolerance I followed the principle, not the official programme. Its aim is 30 minutes of continuous running, which is not 5 km for everyone
Reducing, then stopping, the cycling No direct study located None needed; walking and running were by then providing the loading My choice. Continuing to cycle would also have been reasonable
Returning to hilly 50-mile ultramarathons Not applicable: this is one person's outcome None Shows that a good outcome is possible. It does not show that the meniscus healed, that the approach caused the recovery, or that the result will last for life
Could this help cartilage problems in the knee or hip? (my hypothesis) Not established. No trial has tested this approach for any cartilage condition Gentle cyclic movement has a mechanistic rationale that may be relevant to articular joints as well as to this personal knee experience. In pieces of human femoral-head cartilage loaded to simulate walking, cyclic loading did not significantly change the transport of two small tracers, urea and iodide, but increased the rate at which albumin, a large protein, left the tissue[37]. The meniscus also differs from joint-surface cartilage: part of it has a blood supply[27] Whether this particular programme improves outcomes in a given knee or hip cartilage condition, and especially whether it repairs cartilage, has not been established. My meniscus recovery does not demonstrate that cartilage can be healed this way

Understanding cartilage and its injuries

What is cartilage?

Schematic front view of a knee showing articular cartilage on the ends of the femur and tibia, the menisci between them with a shaded outer zone that has a blood supply, the subchondral bone beneath the cartilage, and the joint capsule and synovial lining containing synovial fluid.
A schematic front view of the knee. The meniscus is a different tissue from the articular cartilage on the bone ends, and only its outer zone has a blood supply.

What this means for runners. The cartilage on your joint surfaces is living tissue that is designed to be loaded. It has no nerves and no blood supply, which is why it can be damaged without hurting and why it repairs poorly.

The evidence and its limits. The ends of the bones in a joint are covered with articular (hyaline) cartilage: a smooth, slippery layer a few millimetres thick. It is mostly water, held in a framework of collagen fibres and water-attracting molecules called proteoglycans. When you load the joint, fluid is squeezed through this framework and the pressure in the fluid carries much of the load; when you unload it, the fluid returns. The cells that maintain the tissue (chondrocytes) are sparse, and the tissue has no blood vessels or nerves, so nutrients reach the cells by moving through the joint fluid and the cartilage itself[28]. Pain from a joint therefore comes from other structures, such as the bone beneath the cartilage, the joint lining and the capsule. The menisci of the knee are a different tissue: two C-shaped pads of tougher fibrocartilage that sit between the thigh bone and the shin bone. See what about the meniscus?

Different problems need different treatment

What this means for runners. Work out which of these you have before deciding what to do. The same word on a scan report can mean an incidental finding or a problem that needs surgery.

Problem What it is Why it matters
Focal chondral defect A localised area of damage to the joint-surface cartilage; the bone beneath is intact Depth, size, position and stability all affect the outlook. Some are found by chance and cause no trouble
Osteochondral injury Damage to the cartilage and the bone beneath it Bone healing and the stability of the fragment come first. May need protected weight bearing or surgery
Osteochondritis dissecans A disorder of the bone beneath the cartilage that can loosen a fragment of bone and cartilage Needs specialist assessment; management depends on age and whether the fragment is stable
Degenerative change Gradual thinning or roughening of cartilage Very common on scans as people age, often without symptoms
Osteoarthritis A condition of the whole joint, involving cartilage, bone, joint lining and other tissues Managed by treating symptoms and function, staying active, and addressing risk factors
Chondromalacia patellae Softening or damage of the cartilage behind the kneecap Not the same as pain around the kneecap, which often occurs with normal cartilage. See Knee Pain
Meniscus tear A tear in one of the fibrocartilage pads of the knee The type of tear decides the treatment. Loss of meniscus function increases the load on the joint-surface cartilage

The evidence and its limits. These categories overlap, and more than one can be present. Whether a lesion is stable is as important as how deep it is: a loose flap or fragment, or a displaced meniscus tear that blocks movement, is a mechanical problem that exercise cannot fix. After surgery, the repaired tissue needs protection on a timescale set by the procedure, whatever the symptoms.

How does cartilage become injured?

What this means for runners. Significant cartilage injuries usually follow a definite injury to the joint, not ordinary running. An increase in training can bring an existing problem to light without having caused it, and the absence of a memorable injury does not rule out a bone stress injury or degenerative change.

The evidence and its limits. Recognised causes include a twisting or impact injury, a dislocated kneecap, ligament injuries such as a torn anterior cruciate ligament, loss of meniscus tissue, and poor alignment that concentrates load on one area. A previous significant knee injury is a clearer risk factor for later osteoarthritis than recreational running: a meta-analysis estimated that the odds of knee osteoarthritis were roughly four to six times higher after cruciate ligament or meniscus injury[38]. Those are relative odds from varied studies, not a prediction for any individual. Meniscus problems are divided into traumatic tears, caused by a definite injury, and degenerative lesions, which develop slowly[16].

Diagnosis, scans and injury grades

What this means for runners. A scan is a tool for answering a specific question, not a verdict. Many people have cartilage and meniscus abnormalities on MRI and no symptoms at all, and the grade on a report does not tell you how much running the joint will tolerate.

The evidence and its limits.

Symptoms such as pain, swelling, stiffness, catching and reduced tolerance of load are reasons to be assessed, but none of them reliably identifies cartilage damage or its depth. A painless click is very different from a knee that physically cannot straighten.

Examination can detect swelling, loss of movement, instability and other causes of pain. It cannot measure the depth of a cartilage lesion.

X-rays show bone, alignment and the features of osteoarthritis. They do not show cartilage directly. Osteoarthritis can usually be diagnosed without imaging in people aged 45 or over with typical activity-related joint pain and little or no morning stiffness[20].

MRI shows cartilage, menisci, ligaments and bone. It is most useful after a significant injury, when a locked knee or a repairable tear is suspected, when mechanical symptoms persist after about three months of rehabilitation, or when surgery is being planned[6]. Its weakness is that it finds a great deal that does not matter. In a pooled analysis of people with no knee symptoms and no history of injury, MRI showed cartilage defects in about 11% of those under 40 and about 43% of those aged 40 and over[39]. A finding on a scan has to be matched to the symptoms and the examination before it is blamed for them.

Five diagrams of cartilage on bone showing ICRS grades 0 to 4: normal; surface softening or shallow cracks; a lesion less than half the depth; a lesion more than half the depth that can reach but not go through the bone; and a lesion through the cartilage into the bone.
ICRS grades describe how deep a cartilage lesion goes. Kellgren–Lawrence grades of osteoarthritis on X-ray are a different scale.

Grades are a common source of confusion because three different scales are in use, and their numbers are not interchangeable:

Scale What it describes Grades
ICRS (International Cartilage Regeneration & Joint Preservation Society)[40] Depth of a cartilage lesion 0 normal; 1 superficial softening or fissures; 2 lesion less than half the cartilage depth; 3 deeper than half the depth; 4 through the cartilage into the bone
Outerbridge Appearance of the cartilage surface, originally at surgery I softening and swelling; II and III increasing fissuring and fragmentation; IV exposed bone. Modified versions are used for MRI
Kellgren–Lawrence Osteoarthritis on an X-ray (bony spurs and narrowing of the joint space) 0 to 4. It is not a measure of the depth of a focal cartilage lesion

"Grade 4" on one scale does not mean the same as "grade 4" on another, so always check which scale a report is using. Depth is only part of the picture: the area of the lesion, where it is, whether its edges are intact, the state of the bone beneath, the alignment of the leg and the condition of the menisci and ligaments all influence what should be done.

Can cartilage heal?

What this means for runners. Separate five things that are often lumped together as "healing": less pain, better function, a return to running, a better-looking scan, and regrowth of normal cartilage. The first three are realistic goals and often achieved. The last is not something that current rehabilitation, injections or supplements have been shown to deliver, and even surgery produces repair tissue that is not identical to the original.

Four boxes: less pain and better function and running are realistic goals; a better-looking scan should be interpreted with care; regrown normal cartilage has not been shown for exercise, injections or supplements, and surgery produces repair tissue rather than the original.
Less pain, better function, a better-looking scan and regrown cartilage are different outcomes that can change independently.

The evidence and its limits.

  • The collagen framework is close to permanent. Radiocarbon dating of human cartilage found virtually no replacement of collagen after skeletal maturity, in healthy and osteoarthritic joints alike[41]. (Strong evidence for the specific finding.) Other components, including the proteoglycans and the water they hold, do turn over and respond to loading. The finding applies to the collagen that was sampled; it cannot rule out every local repair response.
  • Cartilage composition can respond to exercise. In a small randomised trial of people who had previously had part of a meniscus removed, four months of supervised exercise improved an MRI measure related to proteoglycan content compared with no exercise; only 30 of the 45 participants had usable scans[42]. (Limited evidence.) This shows a change in cartilage composition on MRI. It does not show that a defect fills in.
  • Function can improve without regrowth. In a study of 48 people with full-thickness knee cartilage lesions, three months of combined aerobic, strength and neuromuscular training produced meaningful improvements in function, strength and hopping, although four people could not continue because of pain or swelling[43]. (Limited evidence: there was no comparison group, so the improvement cannot confidently be attributed to the programme, and cartilage structure was not the outcome.)
  • Untreated defects can progress. A review of observational studies found that untreated focal defects were associated with further cartilage damage over time; it did not show that osteoarthritis was inevitable[44]. (Limited evidence.) Some people remain well with a persistent defect.
  • Depth changes the repair response. A defect confined to the cartilage has no access to blood or bone marrow cells. One that reaches the bone can fill with repair tissue, but this is largely fibrocartilage, which is less durable than the original hyaline cartilage. This is the basis of microfracture.

There is no reliable figure for the proportion of runners with a focal cartilage defect who recover without surgery. Figures from osteoarthritis, or from other joints, should not be borrowed.

What about the meniscus?

What this means for runners. A meniscus tear is commonly called "torn cartilage", but the meniscus is a different tissue from the joint-surface cartilage, and the type of tear determines the treatment. Many tears, particularly degenerative ones, do well with exercise-based rehabilitation. Some need a surgeon's opinion early.

The evidence and its limits. The menisci spread load across the knee and help to stabilise it. In adults only the outer 10% to 30% of the meniscus has a blood supply; the inner part is poorly supplied, is nourished from the joint fluid, and has limited healing potential[27].

  • Degenerative tears develop gradually, usually in middle age and later, and are common in knees without symptoms. A European and American consensus recommends non-operative treatment, including physiotherapy, as the first approach[16]. In the ESCAPE trial of 321 adults with degenerative tears and no locking, exercise-based physiotherapy was not inferior to keyhole removal of the torn part for knee function over five years[24]. (Strong evidence for this group.)
  • Tears in younger adults are less clear-cut. In the DREAM trial of 121 adults aged 18 to 40 who were candidates for surgery, with traumatic or non-traumatic tears, early surgery was not superior to supervised exercise and education, with surgery available later, on a combined score of pain, symptoms, sport and recreation, and quality of life at 12 months; 26% of the exercise group went on to have surgery[25]. In the STARR trial of 100 adults aged 18 to 45 with traumatic isolated tears, which excluded locked knees and tears judged repairable on MRI, function scores were the same in both groups at 24 months (a mean of 78 out of 100)[26]. In a secondary analysis of DREAM participants who reported mechanical symptoms, measured with a single questionnaire item, early surgery relieved those self-reported symptoms more often, but did not improve pain, function or quality of life more than exercise[45]. (Limited evidence.) The consensus regards both surgery and exercise as reasonable options for acute tears, and could not define criteria for choosing between them[16].
  • Tears that may need early surgery. The same consensus states, on the basis of expert opinion, that bucket-handle tears, complete radial tears, extensive "ramp" lesions and root tears in younger patients may need earlier surgery[16]. (Clinical consensus.) A tear of the meniscus root detaches the meniscus from its anchor; in a cadaver study this raised contact pressure in the knee to the same level as removing the whole meniscus[46]. In a matched comparison of 45 people with tears of the posterior root of the medial meniscus, those whose tears were not repaired had more progression of osteoarthritis and more knee replacements than those who had a repair[47]. In a series of 94 people (average age 67; 90 of them women) whose medial posterior root tears were treated without surgery and followed for an average of 46 months, 36% showed progression of osteoarthritis and 13% had a knee replacement[48]. These figures describe an older population and cannot predict an individual runner's outcome. (Limited evidence: these were not randomised studies, and surgeons choose who is suitable for repair.) A displaced tear that leaves the knee stuck is a mechanical block and needs urgent assessment[6].

The consensus recommends a criterion-based progression back to sport, and says that return should be both criterion based and time based, so that swelling, movement and recovered strength are considered alongside time for healing[16]. Most of the statements in that consensus rest on low-level evidence or expert opinion, which its authors acknowledge. Detailed, tear-by-tear guidance is beyond the scope of this article.

Knee, hip and ankle differences

What this means for runners. Most of the research is on the knee. Do not assume it applies to a hip or an ankle, where the common injuries, the warning signs and the treatments are different.

The evidence and its limits.

  • Hip. Groin or hip pain in a runner has many causes, including a stress fracture of the femoral neck, which must not be missed[7]. Worsening groin pain when you put weight on the leg needs urgent assessment[3]. Cartilage and labral damage in the hip are often associated with femoroacetabular impingement, where the shapes of the ball and socket cause them to pinch. In the UK FASHIoN trial of selected patients with symptomatic femoroacetabular impingement syndrome, hip-related quality of life improved with both physiotherapy-led care and keyhole surgery, and more so after surgery at one year[49]. The trial did not measure cartilage repair. Whether treating impingement prevents later osteoarthritis is unknown[50].
  • Ankle. Cartilage injuries of the ankle usually follow a sprain or other injury and commonly involve the bone of the talus as well. A review of non-operatively managed osteochondral lesions of the talus reported pooled clinical success in about 45% of patients, with widely varying definitions and treatments[51]. That figure applies to the ankle only, and is not a measure of healing. A period of restricted weight bearing is sometimes needed.

Does running wear out cartilage?

What this means for runners. The idea that every mile uses up a fixed supply of cartilage is not supported by the evidence. Recreational runners do not have more knee or hip osteoarthritis than non-runners. The picture is less clear for long careers of very high-volume or elite running, and a previous significant joint injury is a clearer risk factor than the running itself. If you already have knee osteoarthritis, running at a level you choose and tolerate has not been shown to speed it up.

The evidence and its limits.

Recreational and competitive running. A meta-analysis that combined 17 studies and about 115,000 people found hip or knee osteoarthritis in 3.5% of recreational runners, 13.3% of competitive runners and 10.2% of non-running controls[52]. These are percentages of the people studied, not anyone's lifetime risk, and three cautions apply. "Competitive" meant professional, elite or international-level athletes, not club runners who race. The comparisons of each running group with the controls were not individually conclusive. And the authors could not tell whether the associations were caused by running or by other factors such as previous injury. People whose joints hurt tend to stop running, which makes those who continue look healthier. A second review found the evidence too inconsistent to determine the role of running in knee osteoarthritis[53]. (Limited evidence: observational studies, and both comparisons with controls were statistically inconclusive.) For how activity in general relates to osteoarthritis, see activity, inactivity and osteoarthritis.

Running with existing osteoarthritis. In people aged 50 or older with knee osteoarthritis who chose whether and how much to run, runners showed no greater worsening on X-ray over four years than non-runners, and no more new frequent knee pain[19]. (Limited evidence.) Selection limits what we can infer: people with troublesome knees had probably cut back or stopped, so the study tells us little about starting running, high mileage or other diagnoses, and nothing about an unstable focal defect.

What happens to cartilage during a run. A review pooling 24 studies of 396 people (446 knees, mostly healthy) found that several knee MRI measures changed slightly after a run. Regional cartilage volume fell by about 3% to 5%, and composition measures also changed. Tibiofemoral cartilage T2 relaxation time, a composition measure, was back to baseline within 91 minutes, and existing cartilage defects were unchanged 48 hours later. The certainty of this evidence is very low[54]. A separate review concluded that cartilage recovers well after a single run and found moderate evidence that running does not produce new lesions[55]. These changes are usually interpreted as fluid being pressed out of the tissue and returning, not material being worn away. A change in volume or MRI signal after exercise is not a measure of tissue loss, regrowth or readiness for the next run.

Mileage, speed, hills, terrain and shoes. No weekly mileage, pace or lifetime distance has been identified at which cartilage begins to fail, and no training choice has been shown in a trial to preserve cartilage. Speed, gradient, surface and footwear all change the forces at the knee, but a higher force per step is not the same as more load per mile, because the number of steps changes too. Cushioned shoes, soft surfaces, a change of running form or cadence and avoiding downhills may each make running more comfortable for a particular joint. None is a proven way to protect cartilage.

Age and body weight. Osteoarthritis becomes more common with age, and excess body weight is an established risk factor; for people who are overweight, weight loss reduces pain and improves function[20]. Older runners are poorly represented in the research. See Training for Older Runners.

Activity, inactivity and osteoarthritis

What this means for runners. The idea that "motion is lotion" is half right. Joints do need regular loading: cartilage thins when a leg can no longer bear weight, as after a spinal cord injury. But the best population data do not show that active people develop less knee osteoarthritis than inactive people, or that recreational exercise protects against it. Among runners, osteoarthritis does not clearly rise with weekly mileage at recreational levels. The clearer risks are previous joint injuries, contact and pivoting sports played for many years, heavy kneeling or lifting at work, and excess body weight. A U-shaped relationship, with more osteoarthritis at both extremes of activity, is plausible but not established.

The evidence and its limits.

Are inactive people more likely to get osteoarthritis? Not clearly. The strongest study pooled individual data from six community cohorts: 5,065 people without knee osteoarthritis at the start, followed for 5 to 12 years. The amount of recreational activity, including sport, walking and cycling, was not associated with developing knee osteoarthritis or osteoarthritis-related knee pain[56]. (Limited evidence: activity was self-reported, once, at the start.) Studies comparing runners with inactive people often find less osteoarthritis in the runners: 3.5% of recreational runners against 10.2% of controls in one meta-analysis[52], and about half the expected rate in a survey of active marathon runners[57]. In cohorts of about 75,000 runners and 15,000 walkers, running was associated with less osteoarthritis and fewer hip replacements, partly because runners were leaner, while exercise other than running and walking was associated with more[58]. These studies cannot separate cause from selection. People whose joints hurt tend to stop running, and inactivity is often a consequence of osteoarthritis rather than its cause. For people who already have osteoarthritis, the case for activity is much stronger: exercise reduces pain and improves function[59].

Does this mean movement is needed to feed the joint? Not in any simple sense. Cartilage does need loading to maintain itself. In a small study of people with complete spinal cord injury, knee cartilage in some regions was 10% to 16% thinner six months after the injury than in healthy volunteers, with further thinning later[60], and animal studies show similar losses after immobilisation. But that does not show that movement is needed to deliver nutrients. Small molecules move into cartilage by diffusion, and in an experiment on human cartilage, cyclic loading did not significantly change the transport of two small tracers[37]. Loading also acts as a signal that keeps cartilage cells maintaining their matrix. Which of these matters more in a living joint is not known (mechanistic hypothesis). And osteoarthritis is a disease of the whole joint, so cartilage thinning from disuse and osteoarthritis are not the same thing.

Does osteoarthritis track training volume in non-contact sports? At recreational volumes, apparently not. A meta-analysis of nine case-control studies (12,273 people, 1,272 of them runners) found no higher prevalence of knee osteoarthritis in people running 8 to 32 km, 32 to 48 km or more than 48 km a week than in controls running less than 8 km a week; the estimate for the highest-volume group pointed, without reaching significance, towards less osteoarthritis (odds ratio 0.62, 95% confidence interval 0.35 to 1.10)[61]. Among 675 active marathon runners, arthritis was not associated with weekly mileage, years of running or the number of marathons completed; age, family history and previous surgery were the risk factors[57]. At the elite end the picture is less reassuring. A systematic review of sport participation found knee osteoarthritis no more common in sport participants overall than in people who had not taken part (7.7% against 7.3%), but former elite long-distance runners had higher odds of knee osteoarthritis than people who had not taken part (odds ratio 3.3), as did competitive weightlifters (6.9) and wrestlers (3.8)[62], and the running meta-analysis found 13.3% prevalence among elite or international competitors[52]. Those studies of elite athletes rarely separate the effect of training from the effect of injuries. (Limited evidence.)

Is there a U-shaped relationship? A curve with more osteoarthritis among both the inactive and the most heavily loaded is often proposed, and it fits the disuse and elite-athlete findings. But the best population data show no relationship across ordinary activity levels[56], and the running data do not show risk rising with mileage[61]. The upper arm of the curve rests mainly on studies of elite athletes and of people with physically demanding jobs, and the lower arm mainly on comparisons that are prone to reverse causation. (Mechanistic hypothesis; not established.)

What about contact and pivoting sports? The evidence of harm is stronger here. In the same systematic review, soccer players had about three and a half times the odds of knee osteoarthritis, and even non-elite soccer and American football players without a history of knee injury had more osteoarthritis than people who had not taken part[62]. Among 1,207 retired professional footballers (soccer players) in the UK, knee pain, knee osteoarthritis on X-ray and knee replacement were roughly two to three and a half times as common as among men in the general population; knee injury was the main attributable risk factor, but the excess remained after allowing for it[63]. Sports with twisting, cutting and collisions combine high joint loads with a high rate of ligament and meniscus injuries, and a significant knee injury raises the odds of later osteoarthritis roughly four- to six-fold[38].

Loading at work. Physically demanding work is associated with knee osteoarthritis. The odds rose by about a quarter for every 5,000 hours of cumulative kneeling or squatting[64], and physically demanding occupations overall were associated with about 50% higher odds than sedentary work[65]. These are observational associations; the second review was funded by a pharmaceutical company, Merck Serono, which the authors state had no role in the study. Prolonged kneeling and heavy lifting are different exposures from running.

Cycling: what it may help, and what remains unproven

What this means for runners. Easy cycling is a good way to keep a sore knee moving and to hold on to fitness, and in knee osteoarthritis it reduces pain. There is no evidence that a particular cycling prescription repairs cartilage. The numbers I used in my own rehabilitation were arbitrary. You do not need to hit a particular power, duration or number of sessions; choose what your joint tolerates and what fits your day.

The evidence and its limits.

Symptoms and function. A review of 11 trials of stationary cycling in knee osteoarthritis found that it reduced pain; the effects on stiffness, daily activity and quality of life were smaller and less convincing, and the programmes varied widely[29]. (Limited evidence.) In a small trial in older adults with knee osteoarthritis, ten weeks of cycling at 40% or 70% of heart-rate reserve both improved pain, function and fitness, with no significant difference between them, and day-to-day pain did not increase in either group[32]. A small trial without a significant difference cannot show that two intensities are equivalent, or establish a minimum or an optimum resistance. None of these trials measured cartilage repair. Cycling has also featured in some studied meniscus rehabilitation programmes, such as a five-minute stationary-bike warm-up in the DREAM trial[30], but those studies cannot isolate its contribution.

Forces in the knee. In nine people whose knee replacements contained force sensors, the forces between the thigh bone and shin bone during cycling at 25 to 120 watts were lower than during walking. Force rose in step with power, was lower at 60 than at 40 revolutions per minute, and a lower saddle increased the shearing component[31]. An earlier modelling study in healthy volunteers found no effect of pedalling rate[66]. These are measurements of load, in replaced joints, of one part of the knee. They support the idea that easy cycling is a low-load activity. They do not tell us about healing, about the kneecap joint, or about the hip, and they do not identify a best saddle height or cadence beyond avoiding a low saddle and a hard, slow grind. Watts measure the power produced at the pedals (the rate of doing work), not the stress on a particular patch of cartilage. Across studies, bike set-up clearly alters knee movement and forces, while links between set-up and actual injury remain weakly established[36].

The specific dose: 1 W/kg for 15 minutes, three times a day. No clinical trial was found showing that about 1 W/kg for 15 minutes three times a day heals a torn meniscus or a knee or hip cartilage defect. That is a search result, not proof that no such study exists, and it means the dose is untested, not that it has been shown to be ineffective. In healthy young adults, six weeks of cycling at 60 watts for 20 minutes three times a week produced no measurable change in cartilage thickness on ultrasound[67], which tells us little about injured joints.

Frequent short bouts. No study was located comparing several short sessions a day with one session of the same total length for cartilage or meniscus outcomes. In a single-visit study of 37 older adults with knee osteoarthritis, five two-minute bouts of moderate cycling, separated by two minutes of seated rest, led to less pain and fatigue immediately afterwards than ten continuous minutes[34]. That does not test rides spread across a working day. (Limited evidence, and about short-term comfort only.) Splitting exercise up is a reasonable way to make it tolerable or convenient. Whether it does anything more for the joint is unknown.

Passive motion. Machines that move the knee continuously without muscle effort are used after some cartilage operations, on the strength of animal experiments. A systematic review found only four clinical studies and concluded that evidence of benefit in patients was lacking[68]. One surgical team describes encouraging 10 to 15 minutes of cycling a day in a phase of recovery from cartilage-cell implantation[33]. That is a description of practice after surgery. It is not a trial, and it cannot be transferred to a different injury.

Does movement "feed" cartilage? This is the reasoning I used, and it is only partly supported. See how cartilage responds to loading for the detail. Two kinds of fluid movement are easily confused: the circulation of synovial fluid within the joint, and the movement of fluid and dissolved substances through the cartilage itself. Cartilage does depend on the joint fluid for its nutrition, and loading does move fluid through it. But in pieces of human femoral-head cartilage loaded to simulate walking, cyclic loading did not significantly change the transport of two small tracers, urea and iodide, while it increased the rate at which albumin, a large protein, left the tissue by 30% to 100%. The authors inferred that small nutrients such as glucose and oxygen might gain little from pumping under those conditions[37]. (Mechanistic hypothesis.) The experiment does not tell us how an injured meniscus responds, or whether loading helps healing in a living knee or hip. So "movement pumps nutrients into cartilage" is too simple, more pumping does not necessarily mean more healing, and nothing in this work points to a minimum power or a number of daily sessions. The well-supported reasons to keep a joint moving are different ones: prolonged unloading thins cartilage (see activity, inactivity and osteoarthritis), and movement maintains range, muscle and confidence.

Additional treatments

What this means for runners. Judge each treatment by what it has been shown to do. Most of what is on offer aims at symptoms, which can be worthwhile if it lets you rehabilitate. Very little has been shown to change the structure of cartilage. Reliable restoration of a focal defect with normal hyaline cartilage has not been demonstrated for these conservative treatments, and surgery produces repair tissue or transfers existing tissue rather than restoring the original. Every treatment also has a cost in money, time and attention that could have gone into rehabilitation.

Medication and injections
Treatment Mainly studied in Symptoms and function Cartilage structure Risks and costs Practical role
Anti-inflammatory gel (topical NSAID) Knee osteoarthritis Helps. Strong evidence; recommended by NICE[20] No effect shown Skin irritation; little reaches the bloodstream A reasonable first choice for a sore osteoarthritic knee
Anti-inflammatory tablets (oral NSAID) Osteoarthritis Help No effect shown Stomach and bowel, kidney, liver and heart risks; see NSAIDs and Running Lowest dose for the shortest time, with advice from a pharmacist or doctor
Paracetamol Osteoarthritis Little benefit on average None Liver damage in overdose Occasional short-term use; not recommended as routine treatment[20]
Steroid injection Osteoarthritis flare Short-term relief, typically weeks[20] Injections every 12 weeks for two years caused slightly more cartilage thinning than saline[69] Infection (rare), temporary flare Occasional use to settle a flare and allow exercise; avoid regular repeated injections
Hyaluronic acid injection Osteoarthritis Disputed; NICE advises against[20] No regeneration shown Cost; injection risks Not recommended
Platelet-rich plasma (PRP) Knee osteoarthritis Conflicting. No better than saline in a large trial[70]; a specialist consensus supports it for selected patients[71] No benefit to medial tibial cartilage volume at 12 months in that trial Cost; products differ Uncertain. If offered, ask what evidence supports that preparation
"Stem cell" injections Knee osteoarthritis May slightly improve pain and function; limited evidence of low certainty[72] Uncertain High cost; safety uncertain; products not standardised Best confined to proper clinical research
Surgery
Treatment Who it is for Symptoms and function Cartilage structure Risks and costs Practical role
Keyhole washout or "tidy-up" for osteoarthritis Not recommended[20] No worthwhile benefit None Surgical risk Avoid
Debridement of a focal lesion Selected unstable flaps and small defects Improvement reported, but not compared with rehabilitation alone[73] Does not replace lost cartilage Surgical risk Individual decision with a specialist
Microfracture Selected small, contained defects Variable; not shown to be superior to debridement in small defects[73] Fibrocartilage repair tissue; results may decline over 10 years or more[74] Long rehabilitation; durability Discuss alternatives and what happens if it fails
Osteochondral grafts (plugs of bone and cartilage) Selected defects, including those involving bone Supported by selected case series Transfers existing cartilage on its bone Donor-site problems (own tissue); availability and graft failure (donor tissue) Specialist selection
Cartilage-cell implantation (ACI, MACI) Larger defects in a knee with little osteoarthritis Better than microfracture at five years in one trial[75] Repair tissue fills the defect; its quality and durability vary Two operations; cost; long rehabilitation NICE recommends it as an option for eligible patients[76]; access varies, including in the US
Joint replacement Advanced osteoarthritis that is substantially affecting quality of life despite other treatment[20] Effective for pain and function Replaces the surfaces Major surgery; revision; limits on impact sport A separate discussion about returning to running is needed
Supplements
Treatment Mainly studied in Symptoms and function Cartilage structure Risks and costs Practical role
Collagen (hydrolysed, or undenatured type II) Knee osteoarthritis Possible benefit; limited evidence, with very inconsistent results and industry involvement[77][78] No regeneration shown Cost Optional. If you try it, decide in advance what improvement would justify continuing
Glucosamine and chondroitin Osteoarthritis Largely no better than placebo in the biggest trial[79]; NICE advises against glucosamine[20] No reliable effect Cost Not recommended
Vitamin D Knee osteoarthritis No benefit for pain in a two-year trial[80] No reduction in cartilage loss[80] Harm from excessive doses Correct a deficiency for general health, not for cartilage
Omega-3 and krill oil Knee osteoarthritis Mixed: one trial positive, a later one negative[81][82] Not shown Cost No clear role

Pain relief and anti-inflammatory medication

What this means for runners. Medication can reduce pain from an osteoarthritic joint. It does nothing for the cartilage, and it can hide the signs that you are overloading the joint. Use it to make daily life and rehabilitation possible, not to get through a run that the joint would otherwise refuse.

The evidence and its limits. The UK guideline for osteoarthritis recommends offering an anti-inflammatory gel for knee osteoarthritis, and considering tablets only if a gel is ineffective or unsuitable, taking account of the risks to the stomach and bowel, kidneys, liver and heart, at the lowest effective dose for the shortest time, and with a medicine to protect the stomach[20]. It advises against routine use of paracetamol and weak opioids, because the benefit is small, and against strong opioids and glucosamine altogether. This guidance is for osteoarthritis. For a new injury such as a suspected meniscus tear, the NHS suggests paracetamol or ibuprofen in the short term alongside rest and then gentle movement[1]. Anti-inflammatories and running are covered in NSAIDs and Running; the advice in the tendon article is specific to tendons.

Injections and "regenerative" treatments

What this means for runners. No injection has been shown to regrow joint cartilage. Steroid injections give short-term relief in osteoarthritis and should not become a regular habit. PRP and stem-cell injections are expensive, are marketed more confidently than the evidence justifies, and are best regarded as unproven.

The evidence and its limits.

  • Steroid. NICE suggests considering a steroid injection when other medication is ineffective or unsuitable, or to support exercise, and notes that relief lasts only 2 to 10 weeks[20]. In a two-year trial, 40 mg of triamcinolone injected every 12 weeks caused slightly more thinning of the cartilage than saline injections (a difference of about 0.11 mm), with no better pain control[69]. (Strong evidence for that schedule.) That is an argument against scheduled repeat injections; it does not show that a single injection carries the same risk.
  • Hyaluronic acid. NICE recommends against it for osteoarthritis[20].
  • PRP. In the RESTORE trial of 288 adults with knee osteoarthritis, three weekly injections of one PRP preparation were no better than saline for pain or for medial tibial cartilage volume at 12 months[70]. The trial was publicly funded; the manufacturer supplied the PRP kits free of charge. A consensus of European and international cartilage specialists nevertheless considers PRP appropriate for selected patients after other treatment has failed, though not as a first treatment or in severe osteoarthritis[71]. (Limited evidence; the trials and the experts disagree.) PRP preparations differ, so results for one may not apply to another. In the UK, NICE says the procedure should only be used with special arrangements for clinical governance, consent, and audit or research, because the evidence on efficacy is limited in quality[83].
  • Stem-cell and similar cell injections. A Cochrane review of 25 trials concluded that they may slightly improve pain and function, with low certainty, and that effects on joint structure and long-term safety are uncertain[72]. These injections are different from cartilage-cell implantation, which is an established operation.

When surgery may help

What this means for runners. Surgery has a clear place for mechanical problems: a locked knee, a loose fragment, some repairable meniscus tears, and unstable injuries involving bone. For a symptomatic focal defect that has not responded to rehabilitation, several operations can help selected people, but recovery is long, a return to your previous level of running is not guaranteed, and none of them restores a normal joint. Surgery to "clean out" an osteoarthritic knee does not help.

The evidence and its limits.

  • Small defects. In a randomised trial of 65 patients with small symptomatic defects, microfracture (making small holes in the bone so that marrow cells form repair tissue) was not superior to simply trimming the damaged cartilage at two years[73]. Both groups had rehabilitation, and there was no group that had rehabilitation alone, so the trial cannot say how much of the improvement was due to surgery at all. Not being superior does not make the two operations equivalent, and neither has been shown to be better than rehabilitation alone. (Limited evidence.)
  • Durability of microfracture. A review of long-term studies, mostly of defects of 2 to 4 cm², reported declining results and frequent progression of osteoarthritis at 10 years and beyond[74]. The studies had no comparison groups, so they cannot show that microfracture caused the decline.
  • Larger defects. In the SUMMIT trial of defects of at least 3 cm², cartilage-cell implantation on a membrane (MACI) gave better pain and function than microfracture at five years, although MRI appearance did not differ significantly between the groups[75]. (Limited evidence; a single trial, sponsored by the manufacturer.) Better symptoms without a clearly better scan is a reminder that the two do not always go together, and the pain and function advantage does not show that normal cartilage was regenerated.
  • Availability. NICE recommends cartilage-cell implantation as an option for a symptomatic knee defect in eligible patients: those who have not had previous surgery to repair cartilage, who have minimal osteoarthritis, and whose defect is larger than 2 cm², with the operation done at a specialist (tertiary) centre[76]. A recommendation does not guarantee access to every procedure or product everywhere in the UK. In the US, availability depends on the surgeon and on insurance cover.
  • Return to sport. Rates of return to sport after cartilage surgery vary widely with the procedure and the patient, and the published estimates have important weaknesses[84]. "Returned to sport" in these studies rarely means returned to the same running volume. An international expert panel agreed that pain, movement, strength and functional tests should guide the return, but could not agree on minimum time periods[21]. Rehabilitation after cartilage surgery is specific to the procedure and to the site of the repair, and typically involves a period of restricted weight bearing and movement. Follow the programme you are given.
  • Other operations. Realigning the leg, stabilising the knee or restoring meniscus function may be needed at the same time, because a cartilage repair is unlikely to last in a joint that continues to overload it.

Nutrition and supplements

What this means for runners. Eat enough, including enough protein, to support rehabilitation. No supplement has been shown to regrow cartilage. A few may modestly ease osteoarthritis symptoms in some people. Buying several at once is an expensive way to learn nothing about which, if any, is helping.

The evidence and its limits.

  • General nutrition. Adequate energy and protein support muscle rebuilding. No diet has been shown to repair a cartilage defect. See Nutrition.
  • Collagen. A meta-analysis of trials in knee osteoarthritis found improvements in pain and function with collagen supplements, but the results of individual trials were highly inconsistent[77]. Undenatured type II collagen is a different product, taken in milligram doses; a 180-day trial of 40 mg a day in 191 people reported improved symptoms[78]. That trial was funded by the manufacturer, and its authors included company employees. None of these studies measured cartilage repair, and swallowed collagen is digested, not delivered intact to a defect.
  • Glucosamine and chondroitin. In the GAIT trial of 1,583 people, neither supplement, alone or together, was clearly better than placebo overall[79]. A review covering many supplements for osteoarthritis found that the quality of the evidence was generally very low[85].
  • Vitamin D. In a two-year trial in knee osteoarthritis, vitamin D did not reduce pain or cartilage loss[80]. Correcting a genuine deficiency is worthwhile for other reasons.
  • Omega-3. One trial of krill oil, funded by the manufacturer, reported modest improvements in knee pain[81]; a later trial of 2 g a day found none[82].

If you want to try a supplement, try one at a time, decide beforehand what improvement would make it worth continuing and for how long you will give it, and check for interactions with any medicines you take. If you compete under anti-doping rules, use products that are batch-tested.

Long-term joint health

What this means for runners. You cannot guarantee that a joint will stay healthy, but the sensible things to do are reasonably clear: stay active, stay strong, avoid major joint injuries where you can, rehabilitate properly when they happen, and do not carry excess weight. None of these has been proven to prevent osteoarthritis in runners.

The evidence and its limits.

  • Stay active. Exercise is the core treatment for osteoarthritis[20], and stopping all activity because of an abnormal scan generally costs more than it protects. See activity, inactivity and osteoarthritis.
  • Keep your strength. Maintain the strength work from rehabilitation as a permanent part of training, not something that ends when you are back to running.
  • Take joint injuries seriously. A previous significant knee injury increases later osteoarthritis risk[38]. Preventing further injury and addressing persistent deficits, such as weakness or lost movement, are practical goals. Preserving meniscus tissue matters for the cartilage beneath it. Even excellent rehabilitation does not remove the extra risk.
  • Body weight. For people who are overweight and have osteoarthritis, losing weight reduces pain and improves function[20]. This is not advice for a lean runner: cutting food intake to unload a joint risks under-fuelling, with its own serious consequences. See Overtraining Syndrome.
  • Train to your joint's recent history. Increase load in line with what the joint has recently tolerated, not with what your fitness would allow. Deal with a flare-up by adjusting; do not train through repeated swelling.
  • Reassess if things change. Persistent swelling, mechanical symptoms or a steady decline in what you can do are reasons to be re-examined.

The supporting science

How cartilage responds to loading

What this means for runners. Cartilage is built to be squeezed and released, and it needs regular use. But it is not like muscle: its framework is laid down in youth and barely renews, so the adaptation you can expect is in how well the tissue holds water and resists load, not in thickness or in filling gaps.

The evidence and its limits.

Structure. Cartilage is a mesh of type II collagen that traps large, electrically charged proteoglycan molecules. These attract water and make the tissue swell against the collagen mesh, giving it stiffness. Under load, pressure in the trapped fluid carries most of the force, and fluid slowly seeps through the matrix[28].

Turnover. Radiocarbon dating shows that the collagen in adult cartilage is essentially the collagen laid down by the end of growth, even in osteoarthritic joints[41]. The proteoglycans turn over much faster. This is why changes in cartilage "quality" on specialised MRI are plausible, while regrowth of lost tissue is not.

Transport. With no blood supply, cartilage cells depend on substances moving through the matrix. Two processes are involved: diffusion, which is the random movement of molecules and needs no loading, and the bulk flow of fluid caused by compression. Their relative importance depends on the size of the molecule. In an experiment on pieces of human femoral-head cartilage loaded in a pattern simulating walking, fluid flow did not significantly change the transport of two small tracers, urea and iodide, but increased the rate at which albumin, a large protein, left the tissue by 30% to 100%. The authors inferred that pumping might add little to the supply of small nutrients such as glucose and oxygen under those conditions[37]. A later review of fifty years of transport research agrees that molecular size is the dominant factor and that physiological loading can increase the transport of large molecules[86]. Much of that later work concerns getting drugs into cartilage, which is a different question from healing. These are laboratory findings on healthy tissue. Damaged cartilage is more permeable and may behave differently, and no clinical trial has turned any of this into an exercise prescription.

Response to running. The small, temporary thinning of cartilage after a run[54][55] is usually interpreted as this fluid movement made visible.

Immobility and overload. Animal and human observations indicate that both prolonged unloading and injurious overload are bad for cartilage. Where the safe range lies for a given person has not been measured, and a graph of "optimal load" would be guesswork.

What rehabilitation studies actually demonstrate

What this means for runners. Exercise helps people with joint problems, but the average benefit in trials is modest, and almost all of the evidence concerns symptoms and function, mostly in older and less active people than a typical runner. Expect rehabilitation to make the joint work better, not to rebuild it.

The evidence and its limits.

  • Exercise for knee osteoarthritis. The 2024 Cochrane review pooled 139 trials with 12,468 participants. Compared with a sham or attention-only comparison, exercise improved pain by about 9 points and function by about 11 points on 100-point scales, with low to moderate certainty. Those averages fall slightly short of the review's own thresholds for an important difference (12 and 13 points), and because participants knew they were exercising, expectation may have contributed[59]. (For the comparison with a sham or attention-only control, certainty was low for pain, from 28 studies, and moderate for function, from 24 studies.) The sensible reading is realistic expectations, not that exercise is pointless: averages hide people who do much better, and exercise has many other benefits.
  • Type of exercise. The Cochrane review found no clear differences between types of exercise[59]. A 2025 network meta-analysis ranked aerobic exercise as probably the most beneficial type[87]. The best programme is likely to be one you will keep doing.
  • Heavier versus lighter strength training. In the START trial of 377 adults with knee osteoarthritis, 18 months of high-intensity strength training was no better than low-intensity training or an attention control for knee pain or knee joint forces[22]. This does not make heavy training wrong for a runner. It does mean there is no evidence that heavier loading is better for the joint.
  • Focal defects. The direct evidence is thin: one uncontrolled study of 48 patients[43], and the rehabilitation given alongside surgery in surgical trials[73].
  • Cartilage composition. One small trial[42].
  • Meniscus tears. Good trial evidence for degenerative tears[24], growing evidence in younger adults[25][26], and expert consensus for the details[16].

What remains uncertain

What this means for runners. Many of the questions that matter most to runners have not been studied. Where the evidence runs out, careful trial and error, guided by how your own joint responds, is the honest answer.

The evidence and its limits.

Question Current answer
Does gentle cyclic exercise speed the repair of human cartilage or meniscus? A mechanistic rationale exists; no direct clinical evidence
Is any particular cycling power, duration or frequency best? Untested
Are several short sessions better than one longer one? Not established for any structural outcome
What level of pain or discomfort is safe? No validated threshold for cartilage or meniscus problems
Which tests predict a successful return to running? Sensible criteria exist; none has been validated for predicting outcome
Does rehabilitation slow the progression of a defect or of osteoarthritis? Unknown; the evidence is about symptoms and function
Does surgery for a focal defect prevent later osteoarthritis? Not established
Is ultramarathon-level training safe for a joint with cartilage damage? Very little direct evidence in either direction
How should advice change for older runners? Poorly studied
Does regular activity protect against osteoarthritis, and is there a U-shaped relationship with volume? Recreational activity is not clearly protective or harmful; a U-shaped curve is plausible but not established

See also

References

  1. ↑ 1.0 1.1 1.2 1.3 1.4 NHS. Meniscus tear (knee cartilage damage). Page last reviewed 2 September 2026. Link
  2. ↑ 2.0 2.1 Cleveland Clinic. Septic arthritis. Link
  3. ↑ 3.0 3.1 3.2 Robertson GA, Wood AM. Femoral neck stress fractures in sport: a current concepts review. Sports Medicine International Open, 2017, 1(2):E58–E68. doi:10.1055/s-0043-103946
  4. ↑ Stress fractures: diagnosis and management in the primary care setting. British Journal of General Practice, 2019, 69(681):209. doi:10.3399/bjgp19X702137
  5. ↑ MedlinePlus (US National Library of Medicine). Knee pain. Reviewed 7 November 2024. Link
  6. ↑ 6.0 6.1 6.2 6.3 Academy of Medical Royal Colleges, Evidence-Based Interventions programme. Knee MRI for suspected meniscal tears. Published January 2020, reviewed September 2024. Link
  7. ↑ 7.0 7.1 Warden SJ, Davis IS, Fredericson M. Management and prevention of bone stress injuries in long-distance runners. Journal of Orthopaedic & Sports Physical Therapy, 2014, 44(10):749–765. doi:10.2519/jospt.2014.5334
  8. ↑ Hoenig T, Eissele J, Strahl A, Popp KL, Stürznickel J, Ackerman KE, Hollander K, Warden SJ, Frosch KH, Tenforde AS, Rolvien T. Return to sport following low-risk and high-risk bone stress injuries: a systematic review and meta-analysis. British Journal of Sports Medicine, 2023, 57(7):427–432. doi:10.1136/bjsports-2022-106328
  9. ↑ American Physical Therapy Association. State of direct access to physical therapist services in the United States. 2025. Link
  10. ↑ Delaney RK, Sisco-Taylor BL, Wang X, Scherr K, Ubel PA, Haaland B, Kahn VC, Hamstra D, Wei JT, Madanay F, Davis JK, Greeno TU, Fagerlin A. Two specialists, two recommendations: discordance between urologists' and radiation oncologists' prostate cancer treatment recommendations. Urology, 2022, 169:156–161. doi:10.1016/j.urology.2022.06.009. Funded by the US National Institutes of Health; the authors report no disclosures.
  11. ↑ Derrett S, Walley GD, Bridgman SA, Richards P, Maffulli N. Magnetic resonance imaging, knee arthroscopy, and clinical decision making: a descriptive study of five surgeons. International Journal of Technology Assessment in Health Care, 2009. Link
  12. ↑ Lyman S, Oh LS, Reinhardt KR, Mandl LA, Katz JN, Levy BA, Marx RG. Surgical decision making for arthroscopic partial meniscectomy in patients aged over 40 years. Arthroscopy, 2012, 28(4):492–501. Link
  13. ↑ Siemieniuk RAC, Harris IA, Agoritsas T, et al. Arthroscopic surgery for degenerative knee arthritis and meniscal tears: a clinical practice guideline. BMJ, 2017, 357:j1982. doi:10.1136/bmj.j1982
  14. ↑ Abram SGF, Judge A, Beard DJ, Wilson HA, Price AJ. Temporal trends and regional variation in the rate of arthroscopic knee surgery in England: analysis of over 1.7 million procedures between 1997 and 2017. Has practice changed in response to new evidence? British Journal of Sports Medicine, 2019, 53(24):1533–1538. doi:10.1136/bjsports-2018-099414
  15. ↑ Samsson KS, Grimmer K, Larsson MEH, Morris J, Bernhardsson S. Effects on health and process outcomes of physiotherapist-led orthopaedic triage for patients with musculoskeletal disorders: a systematic review of comparative studies. BMC Musculoskeletal Disorders, 2020. doi:10.1186/s12891-020-03673-9
  16. ↑ 16.00 16.01 16.02 16.03 16.04 16.05 16.06 16.07 16.08 16.09 16.10 16.11 16.12 16.13 Prill R, Ma CB, Wong SE, Beaufils P, Monllau JC, et al. The formal EU-US Meniscus Rehabilitation 2024 Consensus: an ESSKA-AOSSM-AASPT initiative. Part II: prevention, non-operative treatment and return to sport. Knee Surgery, Sports Traumatology, Arthroscopy, 2025, 33(8):3014–3024. doi:10.1002/ksa.12689. Unfunded; several authors declare consultancies with orthopaedic companies.
  17. ↑ Seeley MK, Park J, King D, Hopkins JT. A novel experimental knee-pain model affects perceived pain and movement biomechanics. Journal of Athletic Training, 2013, 48(3):337–345. doi:10.4085/1062-6050-48.2.02
  18. ↑ Noehren B, Sanchez Z, Cunningham T, McKeon PO. The effect of pain on hip and knee kinematics during running in females with chronic patellofemoral pain. Gait & Posture, 2012, 36(3):596–599. doi:10.1016/j.gaitpost.2012.05.023
  19. ↑ 19.0 19.1 Lo GH, Musa SM, Driban JB, Kriska AM, McAlindon TE, Souza RB, et al. Running does not increase symptoms or structural progression in people with knee osteoarthritis: data from the Osteoarthritis Initiative. Clinical Rheumatology, 2018, 37(9):2497–2504. doi:10.1007/s10067-018-4121-3
  20. ↑ 20.00 20.01 20.02 20.03 20.04 20.05 20.06 20.07 20.08 20.09 20.10 20.11 20.12 20.13 20.14 National Institute for Health and Care Excellence. Osteoarthritis in over 16s: diagnosis and management. NICE guideline NG226, 19 October 2022. Link
  21. ↑ 21.0 21.1 21.2 21.3 21.4 Lorentz SG, et al. Rehabilitation and return-to-play following knee cartilage injuries: an international Delphi consensus statement. Journal of Cartilage & Joint Preservation, 2024, 4:100193. doi:10.1016/j.jcjp.2024.100193
  22. ↑ 22.0 22.1 Messier SP, et al. 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, 2021, 325:646–657. doi:10.1001/jama.2021.0411
  23. ↑ 23.0 23.1 NHS Better Health. Couch to 5K running plan. Link
  24. ↑ 24.0 24.1 24.2 Noorduyn JCA, et al. Effect of physical therapy vs arthroscopic partial meniscectomy in people with degenerative meniscal tears: five-year follow-up of the ESCAPE randomized clinical trial. JAMA Network Open, 2022, 5:e2220394. Link
  25. ↑ 25.0 25.1 25.2 Skou ST, Hölmich P, Lind M, et al. Early surgery or exercise and education for meniscal tears in young adults. NEJM Evidence, 2022, 1(2):EVIDoa2100038. doi:10.1056/EVIDoa2100038
  26. ↑ 26.0 26.1 26.2 van der Graaff SJA, Eijgenraam SM, Meuffels DE, et al. Arthroscopic partial meniscectomy versus physical therapy for traumatic meniscal tears in a young study population: a randomised controlled trial. British Journal of Sports Medicine, 2022. doi:10.1136/bjsports-2021-105059
  27. ↑ 27.0 27.1 27.2 Fox AJS, Bedi A, Rodeo SA. The basic science of human knee menisci: structure, composition, and function. Sports Health, 2012. Link
  28. ↑ 28.0 28.1 28.2 Sophia Fox AJ, Bedi A, Rodeo SA. The basic science of articular cartilage: structure, composition, and function. Sports Health, 2009, 1:461–468. doi:10.1177/1941738109350438
  29. ↑ 29.0 29.1 Luan L, Bousie J, Pranata A, Adams R, Han J. Stationary cycling exercise for knee osteoarthritis: a systematic review and meta-analysis. Clinical Rehabilitation, 2021. doi:10.1177/0269215520971795
  30. ↑ 30.0 30.1 Skou ST, Lind M, Hölmich P, Jensen HP, Jensen C, Afzal M, Jørgensen U, Thorlund JB. Study protocol for a randomised controlled trial of meniscal surgery compared with exercise and patient education for treatment of meniscal tears in young adults. BMJ Open, 2017, 7(8):e017436. Link
  31. ↑ 31.0 31.1 31.2 31.3 Kutzner I, Heinlein B, Graichen F, Rohlmann A, Halder AM, Beier A, Bergmann G. Loading of the knee joint during ergometer cycling: telemetric in vivo data. Journal of Orthopaedic & Sports Physical Therapy, 2012, 42(12):1032–1038. doi:10.2519/jospt.2012.4001
  32. ↑ 32.0 32.1 Mangione KK, et al. The effects of high-intensity and low-intensity cycle ergometry in older adults with knee osteoarthritis. Journals of Gerontology Series A, 1999, 54(4):M184–M190. PubMed
  33. ↑ 33.0 33.1 Girgis MY, Carey JL. Autologous chondrocyte implantation: second-look arthroscopy. Video Journal of Sports Medicine, 2023. doi:10.1177/26350254231161057
  34. ↑ 34.0 34.1 Aiyegbusi AI, Kolawole FO, Oke IK, Fawole HO. Immediate effects of exercise snacking on physical fatigue, mood, and pain in elderly adults with knee osteoarthritis. Bulletin of Faculty of Physical Therapy, 2024, 29:92. doi:10.1186/s43161-024-00254-y
  35. ↑ Jin Y, Geng X, Wang Q, Xue F, Lu S, Huang L. Psychosocial and physiological health outcomes of outdoor green exercise versus indoor exercise in knee osteoarthritis patients coexisting with type 2 diabetes mellitus: a randomized controlled trial. Frontiers in Endocrinology, 2025, 16:1560536. doi:10.3389/fendo.2025.1560536
  36. ↑ 36.0 36.1 Johnston TE, et al. The influence of extrinsic factors on knee biomechanics during cycling: a systematic review of the literature. International Journal of Sports Physical Therapy, 2017, 12(7):1023–1033. PubMed
  37. ↑ 37.0 37.1 37.2 37.3 O'Hara BP, Urban JPG, Maroudas A. Influence of cyclic loading on the nutrition of articular cartilage. Annals of the Rheumatic Diseases, 1990, 49(7):536–539. doi:10.1136/ard.49.7.536
  38. ↑ 38.0 38.1 38.2 Poulsen E, et al. Knee osteoarthritis risk is increased 4–6 fold after knee injury: a systematic review and meta-analysis. British Journal of Sports Medicine, 2019, 53:1454–1463. doi:10.1136/bjsports-2018-100022
  39. ↑ 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, 53:1268–1278. doi:10.1136/bjsports-2018-099257
  40. ↑ International Cartilage Regeneration & Joint Preservation Society. ICRS clinical cartilage injury evaluation system. Link
  41. ↑ 41.0 41.1 Heinemeier KM, Schjerling P, Heinemeier J, Møller MB, Krogsgaard MR, Grum-Schwensen T, Petersen MM, Kjaer M. Radiocarbon dating reveals minimal collagen turnover in both healthy and osteoarthritic human cartilage. Science Translational Medicine, 2016, 8:346ra90. doi:10.1126/scitranslmed.aad8335
  42. ↑ 42.0 42.1 Roos EM, Dahlberg L. Positive effects of moderate exercise on glycosaminoglycan content in knee cartilage: a four-month, randomized, controlled trial in patients at risk of osteoarthritis. Arthritis & Rheumatism, 2005, 52(11):3507–3514. doi:10.1002/art.21415
  43. ↑ 43.0 43.1 Wondrasch B, Arøen A, Røtterud JH, Høysveen T, Bølstad K, Risberg MA. The feasibility of a 3-month active rehabilitation program for patients with knee full-thickness articular cartilage lesions: the Oslo Cartilage Active Rehabilitation and Education Study. Journal of Orthopaedic & Sports Physical Therapy, 2013. doi:10.2519/jospt.2013.4354
  44. ↑ Houck DA, et al. Do focal chondral defects of the knee increase the risk for progression to osteoarthritis? A review of the literature. Orthopaedic Journal of Sports Medicine, 2018, 6:2325967118801931. doi:10.1177/2325967118801931
  45. ↑ Damsted C, Thorlund JB, Hölmich P, Lind M, Varnum C, Villumsen MD, Hansen MS, Skou ST. Effect of exercise therapy versus surgery on mechanical symptoms in young patients with a meniscal tear: a secondary analysis of the DREAM trial. British Journal of Sports Medicine, 2023, 57(9):521–527. doi:10.1136/bjsports-2022-106207
  46. ↑ Allaire R, Muriuki M, Gilbertson L, Harner CD. Biomechanical consequences of a tear of the posterior root of the medial meniscus: similar to total meniscectomy. Journal of Bone and Joint Surgery (American), 2008, 90:1922–1931. doi:10.2106/JBJS.G.00748
  47. ↑ Bernard CD, Kennedy NI, Tagliero AJ, Camp CL, Saris DBF, Levy BA, Stuart MJ, Krych AJ. Medial meniscus posterior root tear treatment: a matched cohort comparison of nonoperative management, partial meniscectomy, and repair. American Journal of Sports Medicine, 2020, 48(1):128–132. PubMed
  48. ↑ Choi BS, Chung J, Kwak J, Han HS. Subchondral insufficiency fracture is a predictive factor of osteoarthritis progression and conversion to arthroplasty in non-surgically treated medial meniscus root tear. Knee Surgery, Sports Traumatology, Arthroscopy, 2023, 31(10):4492–4500. doi:10.1007/s00167-023-07444-6
  49. ↑ Griffin DR, et al. Hip arthroscopy versus best conservative care for the treatment of femoroacetabular impingement syndrome (UK FASHIoN): a multicentre randomised controlled trial. Lancet, 2018. Link
  50. ↑ Griffin DR, et al. The Warwick Agreement on femoroacetabular impingement syndrome (FAI syndrome): an international consensus statement. British Journal of Sports Medicine, 2016, 50:1169–1176. doi:10.1136/bjsports-2016-096743
  51. ↑ Buck TMF, Lauf K, Dahmen J, Altink JN, Stufkens SAS, Kerkhoffs GMMJ. Non-operative management for osteochondral lesions of the talus: a systematic review of treatment modalities, clinical- and radiological outcomes. Knee Surgery, Sports Traumatology, Arthroscopy, 2023, 31:3517–3527. Link
  52. ↑ 52.0 52.1 52.2 Alentorn-Geli E, Samuelsson K, Musahl V, Green CL, Bhandari M, Karlsson J. 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, 47(6):373–390. doi:10.2519/jospt.2017.7137
  53. ↑ Timmins KA, et al. Running and knee osteoarthritis: a systematic review and meta-analysis. American Journal of Sports Medicine, 2017. doi:10.1177/0363546516657531
  54. ↑ 54.0 54.1 Coburn SL, Crossley KM, Kemp JL, Warden SJ, West TJ, Bruder AM, Mentiplay BF, Culvenor AG. Is running good or bad for your knees? A systematic review and meta-analysis of cartilage morphology and composition changes in the tibiofemoral and patellofemoral joints. Osteoarthritis and Cartilage, 2023, 31(2):144–157. doi:10.1016/j.joca.2022.09.013
  55. ↑ 55.0 55.1 Khan MCM, O'Donovan J, Charlton JM, Roy JS, Hunt MA, Esculier JF. The influence of running on lower limb cartilage: a systematic review and meta-analysis. Sports Medicine, 2022, 52:55–74. doi:10.1007/s40279-021-01533-7
  56. ↑ 56.0 56.1 Gates LS, Perry TA, Golightly YM, et al. Recreational physical activity and risk of incident knee osteoarthritis: an international meta-analysis of individual participant-level data. Arthritis & Rheumatology, 2022, 74(4):612–622. doi:10.1002/art.42001
  57. ↑ 57.0 57.1 Ponzio DY, Syed UAM, Purcell K, Cooper AM, Maltenfort M, Shaner J, Chen AF. Low prevalence of hip and knee arthritis in active marathon runners. Journal of Bone and Joint Surgery (American), 2018, 100(2):131–137. Abstract
  58. ↑ Williams PT. Effects of running and walking on osteoarthritis and hip replacement risk. Medicine & Science in Sports & Exercise, 2013, 45(7):1292–1297. doi:10.1249/MSS.0b013e3182885f26
  59. ↑ 59.0 59.1 59.2 Lawford BJ, et al. Exercise for osteoarthritis of the knee. Cochrane Database of Systematic Reviews, 2024, CD004376. doi:10.1002/14651858.CD004376.pub4
  60. ↑ Vanwanseele B, et al. Knee cartilage of spinal cord-injured patients displays progressive thinning in the absence of normal joint loading and movement. Arthritis & Rheumatism, 2002. PubMed
  61. ↑ 61.0 61.1 Burfield M, Sayers M, Buhmann R. The association between running volume and knee osteoarthritis prevalence: a systematic review and meta-analysis. Physical Therapy in Sport, 2023, 61:1–10. doi:10.1016/j.ptsp.2023.02.003
  62. ↑ 62.0 62.1 Driban JB, Hootman JM, Sitler MR, Harris KP, Cattano NM. Is participation in certain sports associated with knee osteoarthritis? A systematic review. Journal of Athletic Training, 2017, 52(6):497–506. doi:10.4085/1062-6050-50.2.08
  63. ↑ Fernandes GS, Parekh SM, Moses J, Fuller C, Scammell B, Batt ME, Zhang W, Doherty M. Prevalence of knee pain, radiographic osteoarthritis and arthroplasty in retired professional footballers compared with men in the general population: a cross-sectional study. British Journal of Sports Medicine, 2018, 52(10):678–683. doi:10.1136/bjsports-2017-097503
  64. ↑ Verbeek J, Mischke C, Robinson R, et al. Occupational exposure to knee loading and the risk of osteoarthritis of the knee: a systematic review and a dose-response meta-analysis. Safety and Health at Work, 2017, 8(2):130–142. doi:10.1016/j.shaw.2017.02.001
  65. ↑ Wang X, Perry TA, Arden N, et al. Occupational risk in knee osteoarthritis: a systematic review and meta-analysis of observational studies. Arthritis Care & Research, 2020, 72(9):1213–1223. doi:10.1002/acr.24333. Funded by Merck Serono; the authors state that the funder was not involved in study design, data collection or interpretation.
  66. ↑ Ericson MO, Nisell R. Tibiofemoral joint forces during ergometer cycling. American Journal of Sports Medicine, 1986, 14(4):285–290. doi:10.1177/036354658601400407
  67. ↑ Güvener O, Dağ F, Şahin G, Özçakar L. Effects of recumbent and upright bicycle exercises on knee joint cartilage and lower extremity muscles. European Journal of Applied Physiology, 2026, 126:541–548. doi:10.1007/s00421-025-05916-w
  68. ↑ Fazalare JA, et al. The use of continuous passive motion following knee cartilage defect surgery: a systematic review. Orthopedics, 2010, 33:878. PubMed
  69. ↑ 69.0 69.1 McAlindon TE, et al. Effect of intra-articular triamcinolone vs saline on knee cartilage volume and pain in patients with knee osteoarthritis: a randomized clinical trial. JAMA, 2017. doi:10.1001/jama.2017.5283
  70. ↑ 70.0 70.1 Bennell KL, et al. Effect of intra-articular platelet-rich plasma vs placebo injection on pain and medial tibial cartilage volume in patients with knee osteoarthritis: the RESTORE randomized clinical trial. JAMA, 2021, 326:2021–2030. doi:10.1001/jama.2021.19415. Funded by the Australian National Health and Medical Research Council (grant 1106274); the manufacturer (Regen Lab SA) supplied the PRP kits free of charge.
  71. ↑ 71.0 71.1 ESSKA–ICRS. Platelet-rich plasma injections for the management of knee osteoarthritis: the ESSKA-ICRS consensus. Knee Surgery, Sports Traumatology, Arthroscopy, 2024. doi:10.1002/ksa.12320
  72. ↑ 72.0 72.1 Whittle SL, et al. Stem cell injections for osteoarthritis of the knee. Cochrane Database of Systematic Reviews, 2025, CD013342. doi:10.1002/14651858.CD013342.pub2
  73. ↑ 73.0 73.1 73.2 73.3 Randsborg PH, et al. Microfracture versus arthroscopic debridement for the treatment of symptomatic cartilage lesions of the knee: 2-year results from a multicenter double-blinded randomized controlled trial. American Journal of Sports Medicine, 2025, 53:2107–2117. doi:10.1177/03635465251346961
  74. ↑ 74.0 74.1 Gopinatth V, et al. Microfracture for medium size to large knee chondral defects has limited long-term efficacy: a systematic review. Journal of Experimental Orthopaedics, 2024, 11:e70060. doi:10.1002/jeo2.70060
  75. ↑ 75.0 75.1 Brittberg M, Recker D, Ilgenfritz J, Saris DBF; SUMMIT Extension Study Group. Matrix-applied characterized autologous cultured chondrocytes versus microfracture: five-year follow-up of a prospective randomized trial. American Journal of Sports Medicine, 2018, 46:1343–1351. doi:10.1177/0363546518756976. Industry-sponsored.
  76. ↑ 76.0 76.1 National Institute for Health and Care Excellence. Autologous chondrocyte implantation for treating symptomatic articular cartilage defects of the knee. Technology appraisal guidance TA477, 4 October 2017. Link
  77. ↑ 77.0 77.1 Simental-Mendía M, et al. Effect of collagen supplementation on knee osteoarthritis: an updated systematic review and meta-analysis of randomised controlled trials. Clinical and Experimental Rheumatology, 2025, 43:126–134. doi:10.55563/clinexprheumatol/kflfr5
  78. ↑ 78.0 78.1 Lugo JP, Saiyed ZM, Lane NE. Efficacy and tolerability of an undenatured type II collagen supplement in modulating knee osteoarthritis symptoms: a multicenter randomized, double-blind, placebo-controlled study. Nutrition Journal, 2016, 15:14. doi:10.1186/s12937-016-0130-8. Industry-funded; authors included employees of the manufacturer.
  79. ↑ 79.0 79.1 Clegg DO, et al. Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. New England Journal of Medicine, 2006. doi:10.1056/NEJMoa052771
  80. ↑ 80.0 80.1 80.2 McAlindon T, et al. Effect of vitamin D supplementation on progression of knee pain and cartilage volume loss in patients with symptomatic osteoarthritis: a randomized controlled trial. JAMA, 2013, 309:155–162. Link
  81. ↑ 81.0 81.1 Stonehouse W, et al. Krill oil improved osteoarthritic knee pain in adults with mild to moderate knee osteoarthritis: a 6-month multicenter, randomized, double-blind, placebo-controlled trial. American Journal of Clinical Nutrition, 2022, 116:672–685. doi:10.1093/ajcn/nqac125. Industry-funded (Swisse Wellness); the sponsor collaborated on study design and monitoring.
  82. ↑ 82.0 82.1 Laslett LL, et al. Krill oil for knee osteoarthritis: a randomized clinical trial. JAMA, 2024, 331:1997–2006. doi:10.1001/jama.2024.6063
  83. ↑ National Institute for Health and Care Excellence. Platelet-rich plasma injections for knee osteoarthritis. HealthTech guidance HTG497 (formerly interventional procedures guidance IPG637), 23 January 2019. Link
  84. ↑ Wagner KR, Kaiser JT, DeFroda SF, Meeker ZD, Cole BJ. Rehabilitation, restrictions, and return to sport after cartilage procedures. Arthroscopy, Sports Medicine, and Rehabilitation, 2022, 4:e115–e124. doi:10.1016/j.asmr.2021.09.029
  85. ↑ Liu X, et al. Dietary supplements for treating osteoarthritis: a systematic review and meta-analysis. British Journal of Sports Medicine, 2018, 52:167–175. Link
  86. ↑ DiDomenico CD, Lintz M, Bonassar LJ. Molecular transport in articular cartilage: what have we learned from the past 50 years? Nature Reviews Rheumatology, 2018, 14:393–403. doi:10.1038/s41584-018-0033-5
  87. ↑ Yan L, et al. Comparative efficacy and safety of exercise modalities in knee osteoarthritis: systematic review and network meta-analysis. BMJ, 2025, 391:e085242. doi:10.1136/bmj-2025-085242