Osteoarthritis: exercise, weight, topical NSAIDs first — the AU approach

What osteoarthritis actually is

Osteoarthritis is a chronic, heterogeneous whole-joint condition. The older “wear and tear” framing has been superseded — the RACGP knee and hip OA guideline and the ACSQHC Osteoarthritis of the Knee Clinical Care Standard (2024) describe a process involving cartilage, subchondral bone, synovium, capsule, ligaments and periarticular muscle, with inflammatory and metabolic contributions. OA is a chronic, treatable condition — not an inevitable consequence of ageing.

OA is the most common joint disease in Australia, affecting around 1 in 7 adults (Arthritis Australia, AIHW). The dominant sites are knee (medial, lateral, patellofemoral), hip (groin pain, restricted internal rotation early) and hand (first CMC, Heberden’s DIP nodes, Bouchard’s PIP nodes). Spine facet OA, first MTP (hallux rigidus) and glenohumeral OA also occur. OA is the leading cause of joint replacement in Australia, with around 110,000 hip and knee arthroplasties per year (AOANJRR).

A. Core clinical — the AU general practice framework

Diagnostic features

RACGP and the ACSQHC Knee OA Clinical Care Standard both treat OA as a clinical diagnosis in the typical patient. Key features:

• Age 45 or over
• Joint pain on use, relieved by rest
• Morning stiffness under 30 minutes (the “gel” phenomenon after inactivity)
• Restricted range of movement, sometimes crepitus
• No systemic or inflammatory features
• For the hand: Heberden’s nodes (DIP), Bouchard’s nodes (PIP), squaring of the first CMC joint

Imaging — selective, not routine

A weight-bearing knee X-ray, AP pelvis or hand X-ray is indicated for diagnostic uncertainty, surgical planning, or suspected alternative pathology. The ACSQHC standard is explicit that X-ray is not required for diagnosis. MRI is rarely indicated in general practice — incidental cartilage and meniscal abnormalities are common, do not change management, and risk a nocebo effect. Choosing Wisely Australia flags routine MRI for OA workup as a low-value practice.

Clinical vs radiographic mismatch

Radiographic Kellgren-Lawrence grade correlates poorly with symptom severity. Pain experience is modulated by central sensitisation, sleep, mood and beliefs — not by joint-space narrowing alone. Patient education and a structured approach to function matter more than imaging-driven escalation.

Red flags requiring urgent attention

Some presentations are not OA:

• Septic arthritis — single hot, swollen, exquisitely tender joint with fever; ED, joint aspirate before antibiotics
• Inflammatory arthritis (rheumatoid, psoriatic, axial spondyloarthritis) — symmetric small-joint involvement, morning stiffness over 30 minutes, soft-tissue swelling, raised CRP/ESR; refer to rheumatology (see the Australian Rheumatology Association)
• Crystal arthritis — gout, pseudogout (knee/wrist, chondrocalcinosis)
• Avascular necrosis of the hip — younger patient, corticosteroid/alcohol/trauma history, deep groin pain at rest; MRI required
• Osteoporotic or stress fracture — sudden focal pain after minor trauma
• Malignancy or metastasis — age over 50, weight loss, night pain, cancer history
• Polymyalgia rheumatica — age over 50, shoulder/hip-girdle stiffness over 1 hour, markedly raised ESR
• Slipped capital femoral epiphysis or Perthes — paediatric orthopaedics

B. Non-pharmacological care — first-line in AU

The RACGP 2018 guideline and the ACSQHC Knee OA Clinical Care Standard describe a core trio of education, exercise and weight management as first-line for every OA patient, before pharmacotherapy. This is consistent with OARSI 2019 guidelines (Bannuru).

Structured exercise

Sustained exercise reduces pain and improves function across modalities. The flagship AU program is structured GLA:D Australia — a 6-week education plus neuromuscular exercise program for hip and knee OA, delivered by trained physiotherapists, with an outcomes registry. RACGP and ACSQHC both endorse this style of program.

Equally legitimate modalities:

• Land-based — walking, cycling, resistance and balance training
• Aquatic / hydrotherapy — useful in severe weight-bearing OA
• Tai chi, yoga, Pilates — functional benefit comparable to land-based

“The best exercise is the one the patient sustains.” Refer to physiotherapy or exercise physiology, often via a GP Chronic Disease Management Plan (GPCCMP).

Weight management

Weight is the largest single modifiable risk factor for knee and hip OA. The Messier 2005 trial and subsequent work showed that 5–10% weight loss in overweight or obese patients with knee OA produces clinically meaningful pain reduction. Each kg of weight loss reduces knee load by approximately 4 kg per step.

Practical pathway: dietitian via GPCCMP; behavioural support; GLP-1 receptor agonists for eligible patients (OA benefit is secondary to weight benefit; primary tier does not currently endorse GLP-1s as OA pharmacotherapy).

Education and psychological supports

Explaining that the joint is not “wearing out” — that OA is chronic and treatable — is itself a meaningful intervention. The Arthritis Australia My Joint Pain tool is the standard AU consumer-facing self-management resource. Around 20–30% of patients with chronic OA pain also experience depression or anxiety, and pain catastrophising worsens outcomes. CBT and mindfulness-based approaches have supportive trial data; AU access is via the Better Access Mental Health Treatment Plan (Department of Health).

Adjuncts

• Physiotherapy — assessment, exercise prescription, manual therapy
• Occupational therapy — joint protection, splinting for hand OA, activity aids
• Podiatry — first MTP OA, footwear, orthoses
• Cane in the opposite hand — reduces hip/knee load 20–30%
• Unloader brace — modest benefit in unicompartmental knee OA
• TENS, heat, cold — symptomatic adjuncts

C. Pharmacotherapy — what works, what doesn’t

Pharmacotherapy is adjunctive, not a substitute for the core trio. The RACGP guideline, eTG Rheumatology, and AMH describe a stepped approach.

Topical NSAIDs — first-line for knee and hand OA

Diclofenac 1% or 2% gel (TDS-QID) and ketoprofen are first-line pharmacotherapy for knee and hand OA. The Derry Cochrane 2016 review showed comparable analgesic effect to oral NSAIDs with substantially lower systemic exposure. Less useful in deep joints such as the hip.

Oral NSAIDs — short courses, with caution

Per eTG and AMH: ibuprofen 400 mg three times daily, naproxen 250–500 mg twice daily, diclofenac, meloxicam, or celecoxib. Lowest effective dose, shortest period. Co-prescribe a proton pump inhibitor where GI risk is elevated (age over 65, prior GI bleed, concurrent anticoagulant or antiplatelet). Cardiovascular and renal risk constrain use in older patients with comorbidity; NPS MedicineWise provides AU prescribing summaries.

Paracetamol — modest effect

The Machado BMJ 2015 meta-analysis showed minimal effect for knee and hip OA pain, and RACGP no longer treats paracetamol as first-line on efficacy grounds. It remains a reasonable adjunct, or an option where NSAIDs are contraindicated, at 0.5–1 g up to four times daily within standard hepatic safety limits.

Intra-articular corticosteroid — for flares, not maintenance

Methylprednisolone 40 mg or triamcinolone 40 mg with 1% lignocaine provides short-term (4–12 week) flare relief, primarily for the knee. The McAlindon JAMA 2017 trial showed quarterly intra-articular triamcinolone over two years was associated with cartilage volume loss without symptom benefit compared with saline — supporting a limit of 3–4 injections per joint per year.

Duloxetine

Modest effect in chronic OA pain; PBS Authority Required for major depression, generalised anxiety disorder, or diabetic neuropathic pain. Useful where chronic pain and depressive overlay coexist.

Opioids — avoid routinely

The SPACE trial (Krebs JAMA 2018) showed no functional benefit from opioids vs non-opioid analgesics in chronic back, hip and knee pain. Choosing Wisely Australia and the Faculty of Pain Medicine ANZCA advise against routine long-term opioid prescribing for OA. Codeine has been Schedule 8 since February 2018; long-term opioid prescribing requires Authority and regular review. SafeScript and equivalent real-time monitoring should be checked.

Glucosamine, chondroitin, curcumin

The GAIT trial (NEJM 2006) and Wandel BMJ 2010 meta-analysis found no clinically significant benefit for glucosamine/chondroitin; the OARSI 2019 guideline gives a conditional position. TGA-listed, patient-funded, low-harm — a personal trial is reasonable but not recommended primary-tier care. Curcumin has small supportive trial data (Daily PLOS One 2016) but is not in AU primary-tier guidance.

Hyaluronic acid, PRP, stem cell

Intra-articular hyaluronic acid showed clinically irrelevant effect in the Rutjes Ann Intern Med 2012 meta-analysis and is not PBS-subsidised. PRP injections have heterogeneous trial data and remain non-routine. RACGP and Choosing Wisely Australia caution against marketing these as routine. Stem cell therapy outside clinical trials is not supported by primary-tier evidence; predatory clinics are common.

D. Surgery and joint replacement

Total knee and total hip arthroplasty are durable, effective procedures for severe symptomatic OA after optimised conservative care. The AOANJRR is one of the world’s largest arthroplasty registries and documents implant survivorship and revision rates.

Timing matters. Referring too early subjects patients to surgical risk before conservative options are exhausted; referring too late means deconditioning, weight gain and comorbidity reduce post-operative outcomes. Operate the patient, not the X-ray.

Pre-operative optimisation (“pre-hab”) improves outcomes — BMI optimisation (many centres prefer under 35), smoking cessation, glycaemic control, cardiovascular optimisation, dental review (to reduce late prosthetic infection), and pre-operative strengthening.

Access — public-system waiting lists for elective arthroplasty in many AU states remain long, with significant interstate and rural-metropolitan variation. Private health insurance and out-of-pocket costs shape every referral. Arthroscopic lavage and debridement for OA — including for degenerative meniscal tears in the OA knee — are not effective (Sihvonen NEJM 2013) and should not be offered for that indication.

Where AU primary tier is silent on specific surgical thresholds, NICE NG226 is a reasonable international reference for shared decision-making frameworks.

E. Australian operations

GLA:D Australia. Flagship structured exercise-and-education program for hip and knee OA. Provider directory at gladaustralia.com.au; part-funded through Primary Health Networks in some regions.

GP Chronic Disease Management Plan (GPCCMP). OA qualifies. Up to 5 allied health visits per calendar year (10 for Aboriginal and Torres Strait Islander patients under specific items), shared across physiotherapy, exercise physiology, dietetics, podiatry and psychology.

Mental Health Treatment Plan (Better Access). Up to 10 subsidised psychology sessions per calendar year for depression, anxiety or pain catastrophising overlay.

Specialty referral. Rheumatology if inflammatory or crystal arthritis is suspected. Orthopaedic surgery for joint replacement assessment when conservative care is optimised but symptoms remain disabling. Pain medicine input via the Faculty of Pain Medicine ANZCA where chronic pain is refractory.

Private health insurance. Influences access to elective arthroplasty, choice of surgeon and timing.

Real-time prescription monitoring. SafeScript, QScript and equivalents for controlled-drug prescribing.

When to escalate

Seek timely GP or emergency department review for:

• A hot, swollen, exquisitely tender joint with fever (possible septic arthritis)
• Sudden inability to weight-bear, or new acute joint deformity
• New neurological symptoms — leg weakness, numbness, bladder or bowel disturbance
• Deep joint pain at rest, especially in a younger patient or after corticosteroid or alcohol exposure (possible avascular necrosis)
• Joint pain with unintended weight loss, fevers or night pain (red flag for malignancy)
• Symmetric small-joint pain with morning stiffness over 30 minutes (possible inflammatory arthritis)

What this article is and is not

This is general health information drawn from current Australian primary-tier guidelines — the RACGP knee and hip OA guideline, the ACSQHC Osteoarthritis of the Knee Clinical Care Standard, Therapeutic Guidelines, AMH, NPS MedicineWise, the Australian Rheumatology Association, GLA:D Australia, Choosing Wisely Australia, the Faculty of Pain Medicine ANZCA, and major OA trials. It is not personal medical advice and does not create a doctor–patient relationship. Decisions about specific treatment — including imaging, pharmacotherapy, injections and surgery — are made with your own GP and treating clinicians.

For Australian consumer-friendly resources: Arthritis Australia, HealthDirect — Osteoarthritis, Better Health Channel — Osteoarthritis, and GLA:D Australia.

---

Sources cited

1. RACGP — Guideline for the management of knee and hip osteoarthritis (2018)
2. ACSQHC — Osteoarthritis of the Knee Clinical Care Standard (2024)
3. Therapeutic Guidelines (eTG) — Rheumatology
4. Australian Medicines Handbook
5. NPS MedicineWise
6. Arthritis Australia — My Joint Pain
7. GLA:D Australia
8. Australian Rheumatology Association
9. Choosing Wisely Australia
10. Faculty of Pain Medicine ANZCA
11. AOANJRR — Australian Orthopaedic Association National Joint Replacement Registry
12. HealthDirect — Osteoarthritis
13. Better Health Channel — Osteoarthritis
14. AIHW — Osteoarthritis
15. Better Access Initiative
16. Bannuru RR et al. — OARSI 2019 (Osteoarthritis Cartilage)
17. Derry S et al. — Topical NSAIDs (Cochrane 2016)
18. Machado GC et al. — Paracetamol (BMJ 2015)
19. Messier SP et al. — Weight loss + exercise (Arthritis Rheum 2005)
20. McAlindon TE et al. — Intra-articular triamcinolone (JAMA 2017)
21. Krebs EE et al. — SPACE opioid trial (JAMA 2018)
22. Sihvonen R et al. — Arthroscopic meniscectomy vs sham (NEJM 2013)
23. GAIT — Glucosamine + chondroitin (NEJM 2006)
24. Wandel S et al. — Glucosamine + chondroitin meta-analysis (BMJ 2010)
25. Rutjes AW et al. — Hyaluronic acid (Ann Intern Med 2012)
26. Daily JW et al. — Curcumin in OA (PLOS One 2016)
27. NICE NG226 — Osteoarthritis in over 16s