Article Text

Download PDFPDF

Exercise for lower limb osteoarthritis: systematic review incorporating trial sequential analysis and network meta-analysis
  1. Olalekan A Uthman1,2,
  2. Danielle A van der Windt1,
  3. Joanne L Jordan1,
  4. Krysia S Dziedzic1,
  5. Emma L Healey1,
  6. George M Peat1,
  7. Nadine E Foster1
  1. 1Arthritis Research UK Primary Care Centre, Keele University, Keele, Staffordshire ST5 5BG, UK
  2. 2Warwick Centre for Applied Health Research and Delivery (WCAHRD), Division of Health Sciences, Warwick Medical School, University of Warwick, Coventry CV4 7AL UK
  1. Correspondence to : D van der Windt; d.van.der.windt{at}


STUDY QUESTION Which types of exercise intervention are most effective in relieving pain and improving function in people with lower limb osteoarthritis?

SUMMARY ANSWER As of 2002 sufficient evidence had accumulated to show significant benefit of exercise over no exercise. An approach combining exercises to increase strength, flexibility, and aerobic capacity is most likely to be effective for relieving pain and improving function.

WHAT IS KNOWN AND WHAT THIS PAPER ADDS Current international guidelines recommend therapeutic exercise (land or water based) as “core” and effective management of osteoarthritis. Evidence from this first network meta-analysis, largely based on studies in knee osteoarthritis, indicates that an intervention combining strengthening exercises with flexibility and aerobic exercise is most likely to improve outcomes of pain and function. Further trials of exercise versus no exercise are unlikely to overturn this positive result.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Selection criteria for studies

Nine electronic databases were searched from inception to March 2012 to identify randomised controlled trials comparing exercise interventions either with each other or with no exercise control for adults with knee or hip osteoarthritis. Trial sequential analysis was used to investigate reliability and conclusiveness of available evidence for exercise interventions. We conducted Bayesian network meta-analysis to combine both direct (within trial) and indirect (between trial) evidence and investigate the relative effectiveness of different types of exercise intervention. Exercise interventions were classified as either land or water based and as predominantly aimed at increasing strength, flexibility, and/or aerobic capacity (endurance).

Primary outcomes

The main outcomes for analysis, expressed as standardised mean difference (SMD) were changes in pain intensity reported by patients (often measured on a 0–10 numerical rating scale) and limitations in function (assessed with multi-item questionnaires).

Main results and role of chance

A total of 60 trials (44 knee osteoarthritis, 2 hip osteoarthritis, 14 mixed) covering 12 exercise interventions and 8218 patients met the inclusion criteria. Sequential analysis showed that as of 2002 sufficient evidence had been accrued to confirm significant benefit of exercise interventions over no exercise control. The network meta-analysis showed that for pain relief, five types of exercise intervention all including strengthening exercises were significantly more effective than no exercise control. For reduction in limitation in function only the combined intervention of strengthening, flexibility, and aerobic exercise (whether land or water based) was significantly more effective than no exercise control (standardised mean difference −0.63, 95% credible interval −1.16 to −0.10).

Summary of results of network meta-analysis for pain relief and improvement in function (unadjusted estimates based on data from 60 randomised controlled trials)

Bias, confounding, and other reasons for caution

Most (44/60) trials included in this network meta-analysis investigated knee osteoarthritis. In a sensitivity analysis of the 44 trials focusing only on knee osteoarthritis, effect estimates tended to be slightly larger than produced by the overall analysis, which also included trials focusing on hip osteoarthritis or including a mix of patients with lower limb osteoarthritis. Adjustment for potential confounding by differences between trials in the number of supervised sessions, duration of follow-up, or publication year did not greatly impact on the results of the analysis. A further limitation was the small number of studies contributing evidence to some of the comparisons in the network, which also limited the possibility to investigate specific types of exercise, such as t'ai chi.

Study funding/potential competing interests

This study was funded by the National Institute for Health Research (NIHR) under its programme grants for applied research programme (No RP-PG-0407-10386).


  • This is a summary of a paper that was published on as BMJ 2013;347:f5555