Knee Osteoarthritis: Exercise, Weight Loss and Treatment Without Surgery

What follows from this? The effect of exercise is real but moderate: an improvement of the order of a dozen points out of a hundred, not a miracle cure. At the same time, no widely available method has such consistent, repeatedly confirmed evidence at such low risk of harm. That is why exercise is the first recommendation in guidelines, including the 2019 American guideline from the ACR and Arthritis Foundation, in which a strong recommendation applies to exercise and to weight loss in people who are overweight or obese.

It is also important to be honest about what the review does not settle. Trials differed in duration (from 2 to 104 weeks) and type of exercise, so on the basis of the cited review you cannot say that one specific training, for example strength versus aerobic, is best. Guidelines usually hold that what counts is regularity and adjusting the effort so that it is sustainable, not choosing one fashionable method. Strengthening the quadriceps, low-joint-load aerobic exercise (walking, cycling, swimming) and balance exercises are the most commonly used set.

In practice you start with a small amount and increase gradually. A temporary, slight rise in pain in the first weeks is common but does not mean you are 'destroying' the joint, as long as the pain is not sharp, does not build for many hours after training and is not accompanied by swelling. Programmes led by a physiotherapist help choose technique and pace; many people begin with a few visits and then exercise on their own or in a group. You will find more on the role of strength training for the musculoskeletal system in the entry on strength training.

This means that the more weight lost, the greater the improvement, and that a sensible threshold from which participants felt a clear improvement in function was about 8% of body weight. For a person weighing 95 kg that is roughly 7–8 kilograms. It is therefore not 'lose weight until you reach your ideal weight', but a relatively achievable intermediate goal; the confidence interval is wide, however (from 5.2% to 13.3%), a reminder that this is an average, not a guarantee for one person.

The conclusion also needs a caveat: this is a single-group programme, without random allocation to a no-intervention group, and an analysis of people who completed the programme. The dose-response relationship does not by itself prove that weight loss is what causes the improvement, although weight reduction in people who are overweight received a strong recommendation in the 2019 ACR and Arthritis Foundation guideline. It is also worth knowing that exercise alone brings improvement even without major weight loss, so the two components complement rather than exclude each other. The wider context of excess body fat is described in the entry on obesity.

This supports a practice visible in guidelines: for people with knee pain, an NSAID gel or ointment is a sensible first choice, because with comparable functional effectiveness it has a better safety profile than oral preparations. In the 2019 ACR and Arthritis Foundation guideline, strong recommendations apply to both topical and oral NSAIDs in knee osteoarthritis, while paracetamol received a conditional recommendation, a weaker one that depends on the patient's preferences and situation.

Oral NSAIDs work, but they have known risks: stomach, kidneys, blood pressure and heart, especially in older people and those with other conditions. If pain requires such treatment, it is sensible to use the smallest effective dose for as short a time as possible, after consulting a doctor or pharmacist; we write about the risk of NSAIDs for blood pressure and kidneys in a separate article, and about ibuprofen and alcohol in another. We do not give doses here: they depend on the preparation, age and coexisting diseases.

The picture for steroids is therefore nuanced. A single injection may bring short-term relief, according to the Cochrane review mainly for a few weeks, with the effect fading after about half a year, and the quality of evidence is low. In the cited McAlindon trial, regular injections every 3 months for 2 years were linked to greater cartilage loss without a benefit in pain; it is one trial in 140 people, but it deserves attention. That is why a steroid injection is treated as an ad hoc solution for a flare, not a chronic treatment. In the ACR and Arthritis Foundation guideline, however, intra-articular glucocorticoids received a strong recommendation in knee osteoarthritis, which shows the difference of perspective: short-term relief versus the long-term fate of the joint.

For hyaluronic acid injections the picture is even less favourable, at least in this 2012 meta-analysis: the better the trial (larger, blinded, published and unpublished), the smaller the effect. This does not mean that no patient feels relief, as individual trials and preparations differ, but that there is no reliable basis for considering this procedure routinely effective. We discuss it more fully in the article on hyaluronic acid and joint pain. The decision to inject should be discussed with a doctor once this uncertainty and the cost are known.

The conclusion is unambiguous: washing out and cleaning the joint by arthroscopy brings no noticeable, lasting benefit in osteoarthritis, which is why it is not recommended as a treatment for knee pain in most people. The 2023 clinical review in JAMA likewise states that for degenerative meniscal tears the first-line treatment is exercise and that surgery is not indicated even with mechanical symptoms such as 'catching' of the joint. An exception is situations in which the cause of locking is an acute, clear meniscal injury, in which case the orthopaedic surgeon decides after examination.

This trial honestly shows both sides: a knee replacement gives greater improvement in pain and function than conservative treatment alone in patients with advanced disease, but you pay for it with more serious adverse events, while three quarters of patients in the conservative group avoided surgery within the year. That is why the decision about a replacement is based on whether, after exhausting conservative methods, pain and limitation of function are still large, not on the X-ray picture alone.

A conversation with the orthopaedic surgeon should cover: the intensity of night and resting pain, limits on walking and daily activities, the effectiveness of earlier treatment, coexisting diseases and expectations. The JAMA review indicates that joint replacement can be considered when the joint space is minimal or absent and the patient cannot cope with the pain after all appropriate conservative methods have been used.

In knee osteoarthritis the most sensible approach is 'from least to most invasive': movement, muscle strength, body weight, topical drugs, and only then oral drugs, injections and surgery. This is not fashion but a conclusion from evidence in which conservative methods have an incomparably better benefit-to-risk ratio than arthroscopy or repeated injections. At the same time, it must be admitted honestly that the effects are moderate and require consistency, which is a reason to seek the support of a physiotherapist, not to give up.