Skip navigation
Acute Pain | Chronic Pain | General

Non-pharmacological interventions for osteoarthritis of the hip and knee

Clinical bottom line: There were too few trials for any intervention to draw firm conclusions. The best evidence exists for exercise, where all trials showed a benefit in pain and/or functional status. Evidence for TENS is unclear, but TENS may be associated with some modest, short-term pain relief. Based on single trials only, there is preliminary evidence that topical capsaicin and laser therapy may be associated with pain relief. There does not appear to be any benefit with diathermy interventions, and there is no consistent evidence for acupuncture.

There is a lack of evidence for any other type of intervention. For all of these interventions, more trials are needed.

Osteoarthritis causes pain or dysfunction in 20% of the elderly, and is often present in those aged 40 (US figures). Although NSAIDs are widely used, these are associated with adverse effects. Non-pharmacological interventions for pain relief and function include superficial and deep heat, cold, exercise, weight loss, acupuncture, transcutaneous electrical nerve stimulation (TENS), low energy laser, vibration, topically applied creams, pulsed electromagnetic fields and orthotic devices.

Systematic review

Puett DW, Griffin MR. Published trials of nonmedicinal and noninvasive therapies for hip and knee osteoarthritis. Ann-Intern-Med. 1994 Jul 15; 121(2): 133-40. ISSN: 0003-4819.

Inclusion criteria were controlled trials of hip or knee osteoarthritis; non-medicinal or non-invasive interventions; pain and/or lower extremity function outcomes.

Reviewers summarised trials and based their conclusions on original trial conclusions. Non-randomised trials were included.


All trials were randomised unless otherwise stated, but were of very variable quality.

Diathermy (deep heat) followed by exercise

One evaluator-blind trial compared short-wave diathermy and exercise with exercise alone in 42 patients over 4 weeks. There was no benefit of diathermy over control on any measure.

One double-blind trial compared ultrasound with sham ultrasound in 74 patients over 4-6 weeks. There was no benefit of ultrasound over placebo on any measure.


One pseudo-randomised trial compared quadriceps strengthening exercises plus diathermy with diathermy alone in 61 patients over variable follow-up periods. There was a statistically significant 10% therapeutic gain in the exercise group on a measure of functional incapacity (including pain), and a significant improvement in muscle strength.

One trial of 120 patients with hip, knee or ankle osteo- or rheumatoid arthritis compared two different types of aerobic exercise plus a stretching and strengthening exercise programme with a stretching and strengthening exercise programme alone. Aerobic exercise was either pool activity or walking, and the programme was 3 times weekly for 12 weeks. There were no group differences in pain relief. There were significant benefits of aerobic exercise (both types) on aerobic capacity and all subscales of the Arthritis Impact Measurement Scale (AIMS) (i.e. physical activity, anxiety and depression).

One non-blind trial of 102 patients compared an eight week programme of lectures, group discussions, supervised light stretching and strengthening followed by up to 30 minutes of walking with routine care and telephone follow-ups. There was a significant improvement with exercise compared with control in measures of pain, walking test, AIMS subscales for physical activity, and for arthritis impact.

Topical capsaicin

One double-blind trial of 70 patients showed positive benefit of capsaicin over placebo at 1, 2 and 4 weeks, using a physician rating of pain. Capsaicin for pain relief in osteoarthritis is reviewed in more detail in a separate review

Laser therapy

One trial of 50 patients showed benefits of infrared and red laser therapy compared with placebo on measures of pain relief, with twice daily treatment for ten days.


One double-blind trial of 40 patients with osteoarthritis of neck, hand or weight-bearing joints compared traditional acupuncture with sham acupuncture over an eight week period. No benefit of acupuncture was found on any measure.

One trial of 32 patients compared acupuncture with no treatment, and found significant improvement in patient rating of pain, and a physical activity test with acupuncture.

Transcutaneous Nerve Stimulation (TENS)

One trial of 30 patients compared TENS with sham TENS 3 times daily over 3 weeks, and found no significant benefit of TENS on any measure, with the exception of pain relief immediately after TENS, with TENS giving significantly longer pain relief.

One double-blind crossover trial of 12 patients compared TENS with sham TENS over two weeks, and found significantly more pain relief with TENS on one measure only.

One trial of 56 patients compared acupuncture-like TENS with sham TENS twice daily for 6 weeks, and found significantly better pain relief with TENS.

Pulsed electromagnetic fields

One double-blind trial of 27 patients compared pulsed electromagnetic field treatment with sham treatment, with 18 treatments per month. Group differences were not stated.

Adverse effects

Not stated by reviewers

Related topics