What the evidence can and cannot promise

Meaningful slowing of knee osteoarthritis is genuinely achievable through the right combination of conservative measures — but reversal is not, and that distinction matters from the outset.

Cartilage has no direct blood supply and a very limited capacity for self-repair. Once lost, it does not regenerate, and no pharmacological or biological agent has yet been approved as a true disease-modifying drug capable of halting or reversing that structural loss. This is not a caveat to bury in small print; it is the biological reality that shapes every treatment decision that follows.

What 'slowing' means in practice is measurable and clinically significant: lower pain scores, better function on validated scales such as the WOMAC and KOOS, preserved mobility, and — for many patients — a substantial deferral of, or complete avoidance of, total knee replacement. The goal of conservative management is to keep the joint working as well as possible for as long as possible, within the honest limits of what the evidence supports.

Why exercise is the single most important first step

Structured exercise is the one intervention every major clinical guideline agrees on. Both OARSI (2019, cited over 4,200 times) and NICE NG226 designate it a core, universal treatment for knee osteoarthritis — not an optional lifestyle add-on, and not reserved for milder cases. The recommendation applies regardless of disease severity.

The breadth of supporting evidence is substantial. A 2023 network meta-analysis of 39 randomised controlled trials (N=2,646) found that all five major exercise types — stationary cycling, resistance training, aquatic exercise, yoga, and traditional exercise — outperformed control conditions on pain and function. Stationary cycling ranked highest for pain relief (SUCRA 80.8%), while yoga showed the strongest results for stiffness and physical function. A 2025 umbrella review synthesising 58 systematic reviews reinforced this picture: 63.7% of reviews reported exercise improved all predefined outcomes, with muscle-strengthening the most commonly prescribed modality across the literature.

What the evidence does not support is treating exercise as generic advice to 'stay active.' Supervised programmes consistently outperform unsupervised activity. This distinction is clinically meaningful: it points to physiotherapy referral, with a structured and progressive programme tailored to joint load and individual capacity, rather than simply suggesting daily walks.

Some caution is warranted on the specifics. Trial methodology across the exercise literature is heterogeneous, and the precise optimal dose, frequency, and modality for any individual remain debated. The overall direction of the evidence is clear; the fine-grained protocol is not.

The biomechanical case for weight management

The arithmetic of joint loading makes a compelling case on its own terms. Every 1 lb (454 g) of body weight removed takes approximately 4 lbs (1.8 kg) of mechanical force off the knee with each step — and given that an average person takes 6,000–8,000 steps a day, even a modest reduction compresses the accumulated daily load on articular cartilage considerably.

This is the biomechanical logic behind clinical weight targets, which are more precisely defined than the general advice to 'lose some weight.' A 5–10% reduction in body weight is the threshold consistently supported by randomised trial data (Messier et al., 2018) as producing clinically meaningful improvements in both pain and physical function in people with established knee OA. In real terms for an 85 kg individual, that is 4–8 kg — an achievable rather than aspirational goal.

The preventive evidence is equally striking. Bliddal and colleagues (2014) found that a 5.1 kg weight reduction sustained over 10 years halved the likelihood of women developing symptomatic knee OA — a disease-onset effect, not merely symptom relief.

Weight management and exercise also reinforce each other. Reduced joint load makes movement less painful, which makes structured exercise more tolerable and adherence more likely — a compounding benefit that neither intervention delivers as effectively in isolation.

Current pharmacological sequencing

Guidance on pain medication for knee OA has shifted noticeably since the mid-2010s, and many patients — and some clinicians — are still working from older frameworks.

Current 2024–2025 guidelines recommend topical NSAIDs, principally diclofenac gel, as the preferred pharmacological starting point, with topical capsaicin cream as an alternative. The rationale is straightforward: local application delivers anti-inflammatory effect to the joint while substantially reducing the gastrointestinal and cardiovascular risks associated with oral NSAID use. Oral NSAIDs remain an option, but are now reserved for cases where topical agents have proved inadequate — and only at the lowest effective dose for the shortest necessary duration.

The most counterintuitive change concerns paracetamol. Updated guidelines conditionally discourage its use for knee OA, reflecting evidence of limited efficacy and concerns about potential harm with prolonged use. This is a clear departure from earlier analgesic ladder approaches in which paracetamol was the automatic first step. Patients who have been prescribed it long-term should raise this with their GP before making any changes — stopping existing medication unilaterally is not advisable.

Across all of these options, the same boundary applies: pharmacological management is adjunctive. Analgesia can reduce pain enough to make movement possible and daily function manageable, but medication alone does not slow joint deterioration. It supports the exercise and weight management strategies covered above; it does not replace them.

What injections can and cannot do

Injections earn their place in knee OA management not as disease modifiers but as enablers — their role is to reduce pain enough that exercise and rehabilitation become tolerable, not to substitute for them.

Corticosteroid injections act quickly on intra-articular inflammation and can break a pain cycle that has prevented someone from engaging with physiotherapy. Hyaluronic acid (HA) injections aim to supplement depleted synovial fluid, providing lubrication and some symptom relief. Neither, however, has been shown to slow structural cartilage loss — a distinction that current evidence makes clearly.

Platelet-rich plasma (PRP) now sits above HA in the emerging evidence hierarchy. A 2025 systematic review (Glinkowski et al.) found that PRP produced clinically significant improvements in pain and function lasting up to 12 months and outperformed HA in most outcome comparisons — a meaningful distinction for patients weighing options beyond standard viscosupplementation.

Supplements such as glucosamine, chondroitin, and curcumin attract considerable patient interest, but the evidence base remains moderate at best. None has demonstrated structural disease-modification in well-powered trials, and current guidelines treat them cautiously. Unloader braces, which shift mechanical load away from the more damaged compartment, may relieve pain in selected patients with predominantly medial OA, though evidence of progression-slowing is similarly limited.

For patients who want a fuller account of injection options — including dosing, candidacy, and the distinction between biologics and conventional injectates — the clinic's dedicated injectable therapies article covers these in depth.

When conservative management needs specialist input

Three to six months of well-structured conservative care — physiotherapy, weight management, and appropriate analgesia — is the reasonable clinical test. If pain and function have not improved meaningfully despite genuine adherence, or if symptoms are worsening rapidly, specialist input becomes the sensible next move.

What a specialist assessment offers beyond GP-level review is considerably more targeted. A consultant can determine where in the joint the mechanical problem is concentrated, whether imaging findings actually correspond to the clinical picture (they frequently diverge), and whether the conservative programme was genuinely optimised or only partially attempted. A patient who has attended general exercise classes is not equivalent to one who has completed a supervised physiotherapy protocol — that distinction shapes the next step.

Critically, escalation to specialist care does not mean immediate surgery. Where the full conservative pathway has been properly exhausted, realistic options short of joint replacement include joint-preservation procedures: high tibial osteotomy to redistribute compartmental load, focal cartilage repair where a discrete defect is identified, or advanced injectables where intra-articular inflammation remains the dominant driver. These are not applicable in every case — candidacy depends on joint alignment, OA grade, age, and functional demand — but they represent a meaningful intermediate stage that is often underused when patients move prematurely to replacement.

For patients at this decision point, specialist assessment can be arranged at londoncartilage.com.