Skip navigation

Exercise therapy for OA of the hip or knee

Clinical bottom line: There is limited evidence that exercise provides modest reductions in pain and disability in patients with mild or moderate osteoarthritis of the hip or knee. The majority of trials were small, of low validity and had insufficient power.

Systematic review:

Van Baar ME, Assendelft WJJ, Dekker J, Oostendorp RAB, Bulsma JWJ. Effectiveness of exercise therapy in patients with osteoarthritis of the hip or knee. Arthritis and Rheumatism 1999; 42: 1361-1369.

Date review completed: September 1997

Number of trials included: 11

Number of patients: (mean number of patients per group in the two studies of high quality and sufficient power was 100-146)

Control groups: placebo, no treatment.

Main outcomes: pain; self-reported disability; observed disability; patient's global assessment of treatment effect. .

Inclusion criteria were randomised controlled trial which assessed exercise therapy in patients with osteoarthritis (OA) of the hip or knee and used the outcomes listed above. MEDLINE (1966-1997), EMBASE (1988-1997), CINAHL (1982-1997) and the Cochrane Controlled Trials Register were searched. Bibliographies of retrieved reports were checked for additional citations. No language restrictions were made.The validity of the trials and their power to detect a difference between treatments were assessed. Effect sizes, with 95% confidence intervals, were calculated.


Eleven studies were included. Only one study assessed OA of the hip. Two trials were described as being of adequate validity and sufficient power and are discussed here These assessed the effect of aerobic or resistance exercises in patients with mild-to-moderate OA of the hip or knee over 12 weeks. There were between 100-146 patients per treatment group.


In one study aerobic exercise was more effective than resistance exercise; the effect sizes were 0.31 (0.28 to 0.34) and 0.47 (0.44 to 0.5) respectively. The second trial assessed a combination of strengthening and range-of-motion exercises and functional training; the effect size for pain was higher, 0.58 (0.54 to 0.62).


Modest improvements were reported in the two high quality studies. The effect sizes were highest for patients randomised to aerobic exercise for both self-reported disability, 0.41 (0.38 to 0.44), and for observed disability in walking, 0.89 (0.85 to 0.93).

Adverse effects

The reviewers' stated that adverse effects of exercise therapy were not often mentioned in the trials. No further information was provided.



The reviewers' stated that different outcome measurement tools were used in the trials, most were small and no information about the long-term, beneficial or adverse effects, of exercise were provided. The reviewers' concluded that there is some evidence for small to modest beneficial effects of exercise in the treatment of OA of the hip or knee. These conclusions are based mainly on the results of the two trials of adequate validity and power. The studies of lower validity and inadequate power showed variable results and reported on few of the outcomes of interest.

Further reading

Related topics