Exercise for osteoarthritis of the hip

Hip osteoarthritis is a major public health issue. International treatment guidelines recommend exercise to manage symptoms. This is an update of a review first published in 2009 and last updated in 2014.

To determine the effects of land-based exercise on pain, physical function, quality of life, participant-reported treatment success, study withdrawals, and adverse events in people with hip osteoarthritis.

We searched CENTRAL, MEDLINE, Embase, and two trial registries from February 2013 (search date for last update) to 5 February 2025. We also checked reference lists for additional studies.

We included randomised controlled trials (RCTs) of adults with hip osteoarthritis. Eligible comparisons were as follows. • Exercise versus attention control/placebo. • Exercise versus no treatment/usual care/limited education. • Exercise plus a co-intervention (A) versus the co-intervention (A) alone. We excluded perioperative exercise programmes; interventions involving vibration therapy, gait aids, or retraining; and studies that added unequal co-interventions in both groups or that compared only one form of exercise to another.

Our critical outcomes were pain, physical function, and quality of life. Our important outcomes were participant-reported treatment success, study withdrawals, and adverse events.

As this was a review update, we used the original Cochrane risk of bias tool (RoB 1).

We used standard methodological procedures expected by Cochrane. We synthesised results for each outcome using random-effects meta-analysis where appropriate. We used the GRADE approach to assess certainty of evidence.

Eighteen studies (1368 participants) met our inclusion criteria. Two studies (123 participants) evaluated exercise versus attention control or placebo, 10 studies (494 participants) evaluated exercise versus no treatment/usual care/limited education, and seven studies (751 participants) evaluated exercise plus a co-intervention (A) versus the co-intervention (A) alone. One trial (210 participants) included two comparator categories. Most studies were small and unblinded.

We converted all continuous effect estimates from standardised mean differences to mean differences (MDs) on a scale of 0 to 100. For pain and physical function, a negative MD indicates an improvement, while for quality of life, a positive MD indicates an improvement. Pain, physical function, quality of life, and participant-reported treatment success were measured immediately after treatment, while study withdrawals and adverse events were recorded at the end of follow-up. Exercise versus attention control/placebo Exercise, compared to attention control/placebo, may have little to no effect on pain (MD -6.31 points, 95% confidence interval (CI) -12.98 to 0.35; 2 studies, 123 participants; low certainty). Exercise may improve physical function slightly (MD -7.44 points, 95% CI -13.86 to -1.01; 2 studies, 123 participants; low certainty). No studies reported quality of life or participant-reported treatment success. Exercise may have little to no effect on study withdrawals (relative risk (RR) 0.83, 95% CI 0.23 to 3.03; 1 study, 106 participants; low certainty). Exercise may increase the risk of adverse events, but the evidence is very uncertain (RR 8.00, 95% CI 1.13 to 56.79; 1 study, 18 participants; very low certainty). Exercise versus no treatment/usual care/limited education Exercise, compared to no treatment/usual care/limited education, probably reduces pain slightly (MD -7.19 points, 95% CI -10.70 to -3.68; 9 studies, 449 participants; moderate certainty) and probably improves physical function slightly (MD -8.79 points, 95% CI -12.00 to -5.41; 9 studies, 447 participants; moderate certainty); however, these improvements are unlikely to be clinically meaningful. Exercise probably has little to no effect on quality of life (MD 2.31 points, 95% CI -1.15 to 5.91; 6 studies, 279 participants; moderate certainty). Exercise may have little to no effect on participant-reported treatment success (RR 1.57, 95% CI 0.62 to 3.99; 2 studies, 69 participants; low certainty) or study withdrawals (RR 1.51, 95% CI 0.80 to 2.86; 7 studies, 404 participants; low certainty). Exercise may have little to no effect on adverse events, but the evidence is very uncertain (RR 2.95, 95% CI 0.62 to 13.96; 6 studies, 257 participants; very low certainty). Exercise plus a co-intervention (A) versus the co-intervention (A) alone Exercise plus a co-intervention (A), compared to the co-intervention (A) alone, probably has little to no effect on pain (MD -3.86 points, 95% CI -8.07 to 0.35; 7 studies, 751 participants; moderate certainty), physical function (MD -2.37 points, 95% CI -6.59 to 1.86; 7 studies, 751 participants; moderate certainty), or quality of life (MD 3.60 points, 95% CI -1.30 to 8.36; 4 studies, 456 participants; moderate certainty). Exercise may have little to no effect on participant-reported treatment success (RR 1.25, 95% CI 0.62 to 2.54; 2 studies, 399 participants; low certainty), and probably has little to no effect on study withdrawals (RR 0.85, 95% CI 0.55 to 1.32; 6 studies, 683 participants; moderate certainty). Exercise probably reduces the risk of adverse events slightly (RR 0.75, 95% CI 0.58 to 0.97; 6 studies, 731 participants; moderate certainty).

Compared with attention control or placebo, exercise may have little to no effect on pain and may improve physical function slightly. There was no evidence for quality of life or participant-reported treatment success. Exercise may have little to no effect on withdrawals. Evidence regarding adverse events is very uncertain. Compared with no treatment, usual care or limited education, exercise probably improves pain and physical function slightly, although these effects are unlikely to be clinically meaningful. Exercise probably has little to no effect on quality of life and may have little to no effect on participant-reported treatment success or withdrawals. Evidence regarding adverse events is very uncertain. Exercise plus a co-intervention (A), compared to the co-intervention (A) alone, probably has little to no effect on pain, physical function, quality of life, and study withdrawals, and may have little to no effect on participant-reported treatment success. Exercise probably reduces the risk of adverse events slightly.

No funding.

The original protocol was for a review on exercise for osteoarthritis of the hip or knee (https://doi.org/10.1002/14651858.CD004376). The review on hip osteoarthritis alone was first published in 2009 (https://doi.org/10.1002/14651858.CD007912), and last updated in 2014 (https://doi.org/10.1002/14651858.CD007912.pub2).