Why the conservative phase comes first

If you have arrived here researching injection options for knee osteoarthritis, the recommendation to 'try physio first' can feel like a delay rather than a plan. It is not. The OARSI 2019 guidelines — with over 4,000 citations, the most widely referenced international framework for knee OA — designate structured land-based exercise and arthritis education as 'Core Treatments': universally recommended for every patient, regardless of symptom severity or comorbidity profile. That classification places exercise and physiotherapy above injectables in the evidence hierarchy, not below them.

This matters practically as well as clinically. In NHS musculoskeletal pathways and most private referral frameworks, completing a documented conservative phase — supervised physiotherapy, self-management education, and weight management where relevant — is a required step before intra-articular injectables are considered appropriate. An insurer or referring clinician will typically expect evidence that these options have been genuinely attempted, not simply mentioned at a consultation.

The rationale runs deeper than pathway compliance. Mild-to-moderate knee OA often follows a self-limiting course, and addressing modifiable factors — muscle strength, joint load, body weight — during this window may genuinely alter the structural trajectory of the disease. There is a systemic dimension too: OA-related immobility is associated with a standardised mortality ratio of approximately 1.55, making the restoration of mobility a cardiovascular and mortality issue, not only a pain one.

The sections that follow examine what the evidence actually shows across the three pillars of conservative care: exercise, physiotherapy, and weight management.

What structured exercise does for knee OA

Exercise, in the knee OA context, means something more specific than 'keep active'. Current 2024–2025 guidelines coalesce around a structured programme combining three modalities: resistance training targeting the quadriceps, hamstrings, and hip abductors; low-impact aerobic activity such as walking, cycling, or aquatic exercise; and neuromuscular or balance work. The recommended delivery is two to three sessions per week, each around 45 minutes, blended with a consistent home programme — with patient education covering joint protection and pain-coping running alongside.

At these doses, structured exercise produces meaningful reductions in pain and improvements in function across knee OA severity grades. The evidence base underpinning this is robust; the challenge is not its quality but its application.

A 2024 review in the Journal of Rheumatology (Huffman et al.) found that most people with knee OA remain insufficiently physically active despite guideline endorsement. The gap is an implementation problem, not an evidence one — and it is familiar enough that guidelines now explicitly call on clinicians to refer patients to supervised exercise or structured self-management programmes rather than simply advising general activity. Referral, in this framing, is an active clinical decision, not a default suggestion.

How the three modalities should be weighted relative to one another is still being refined. The 2024 OARSI Year in Review (Lawford et al.) identifies optimal exercise prescription — specifically the balance between resistance, aerobic, and neuromuscular components — as an open research question. No single modality has established clear superiority; current best practice is a blend.

How physiotherapy compares to a corticosteroid injection

The clearest direct comparison comes from a 2020 trial published in the New England Journal of Medicine (Deyle et al., n=156). Researchers randomised patients with symptomatic knee OA to either a course of physical therapy or a single glucocorticoid injection — the most commonly used intra-articular treatment in routine practice. At one year, the physical therapy group scored meaningfully better on the WOMAC — a validated composite measure of pain, stiffness, and physical function — than those who had received the injection. The difference was not marginal; physical therapy produced superior medium-term outcomes.

Strengthening the quadriceps, hamstrings, and hip abductors reduces the compressive load transmitted through the knee with every step. Neuromuscular training recalibrates gait patterns and proprioception, reducing the abnormal joint-loading that drives pain sensitisation. Manual therapy addresses soft-tissue restriction and centrally mediated pain components. None of these mechanisms are available to a corticosteroid, which acts on synovial inflammation and whose effect is, by design, temporary.

There is also a durability argument. A supervised physiotherapy course builds a home-exercise habit that continues to deliver benefit after formal sessions end. Injection effects, even when pronounced initially, are time-limited — and repeated corticosteroid use carries a documented risk of accelerated cartilage loss, making it a counterproductive long-term strategy for most patients.

Injections are not wrong; they have a defined role that later sections address. But the clinical evidence does not support reaching for them before a genuine physiotherapy trial has been completed.

Weight management: the underused lever

Body weight bears on knee OA in a straightforward but easily underestimated way. For every kilogram of body mass, the compressive force transmitted across the knee joint during walking is roughly three to six times greater than the added weight alone — a multiplier that explains why the mechanical benefit of even modest weight loss is disproportionately large. The OARSI 2019 guidelines (Bannuru et al.) make this explicit: dietary weight management is listed as a co-core recommendation specifically for patients with a BMI above 25, placed on equal footing with structured exercise rather than treated as a secondary lifestyle suggestion.

A reduction of 5–10% of body weight is associated with measurable falls in joint load and reported symptom burden — not a transformation, but clinically meaningful gain within a realistic timeframe. The NHS similarly prioritises weight management alongside exercise as a primary intervention, reflecting the same biomechanical reasoning.

What remains less settled is how best to achieve that weight reduction in practice. The 2024 Journal of Rheumatology review (Huffman et al.) identifies adherence strategies for weight management as an active area of investigation: the goal is clear; the optimal programme design is not. Where structured weight-loss support — whether dietetic referral, behavioural intervention, or supervised programme — is available, guideline language supports referring patients to it alongside physiotherapy, not instead of it. The two interventions address different mechanisms and their effects are additive: exercise rebuilds load-bearing capacity; weight loss reduces the load itself.

The risk of moving to injections too early

The case for completing a conservative phase before moving to injections is not only about maximising what physiotherapy and weight management can deliver — it is also about avoiding a harm that is specific to one of the most commonly used intra-articular agents. Repeated corticosteroid injections are associated in published evidence with accelerated cartilage loss in the knee joint. They act on synovial inflammation to produce temporary symptomatic relief but do not modify the underlying disease process; at higher cumulative doses or frequency, the evidence points toward net structural deterioration rather than preservation.

This matters because knee OA is already characterised by progressive cartilage thinning. A treatment pathway that temporarily quietens the joint while simultaneously hastening that thinning is, for most patients, a counterproductive trade-off — particularly where a genuine conservative trial could have delivered durable improvement without that structural cost.

It is important to separate this concern from intra-articular treatment as a whole. Other agents — hyaluronic acid, polyacrylamide hydrogel, and biologics such as PRP — carry distinct profiles and are not implicated in the same cartilage-loss mechanism. The corticosteroid-specific risk is one of the reasons the conservative window exists as a formal treatment stage rather than a procedural formality: completing it before escalating to injectables is, in this sense, both a positive step and a protective one.

When to escalate beyond conservative care

No single evidence-based threshold defines how long a conservative trial should last before escalation becomes appropriate. The 2024 OARSI Year in Review (Lawford et al.) identifies optimal exercise duration and programme design as open research questions — so any fixed rule ('try physio for three months, then consider injection') reflects clinical convention rather than a defined evidence boundary.

In practice, the markers used are functional rather than calendrical. Persistent pain continuing to limit ordinary daily activity — walking, stair use, disturbed sleep — despite a structured, genuinely adhered-to exercise and physiotherapy course is the most common prompt for reconsideration. Progressive functional decline over time, or imaging findings suggesting structural change beyond what conservative measures can address, add further clinical weight to that judgement.

The reverse matters equally: patients who have not completed a proper structured programme — not because it failed but because it was not genuinely attempted — are typically counselled to do so before injection is arranged. This reflects both OARSI's designation of exercise as a core treatment applicable regardless of comorbidity profile and standard MSK pathway reasoning.

When escalation does appear warranted, a specialist MSK assessment is the appropriate next step — not injection directly. That review covers what has been tried, current symptoms, examination findings, and imaging, to establish which pathway fits: a revised conservative strategy, an intra-articular agent such as viscosupplementation or a newer option, or something else. The right answer depends on what that assessment actually shows — and an AMSK assessment can establish clearly where on the treatment ladder a patient currently stands.

1. [1] Osteoarthritis – Treatment and support – NHS. (2024). https://www.nhs.uk/conditions/osteoarthritis/treatment/ https://www.nhs.uk/conditions/osteoarthritis/treatment/