Relative efficacy of different types of exercises in the treatment of knee and hip osteoarthritis: A network meta-analysis of randomised controlled trials

Goh, Siew Li (2019) Relative efficacy of different types of exercises in the treatment of knee and hip osteoarthritis: A network meta-analysis of randomised controlled trials. PhD thesis, University of Nottingham.

[img] PDF (Thesis - as examined) - Repository staff only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (11MB)


Background: The need for exercise in osteoarthritis (OA) has been long-recognised, leading international guidelines and experts to universally endorse exercise as a core treatment for OA. However, currently there is no consensus on which type of exercise is most beneficial, and there is clear differentiation of exercise effects across different OA outcomes.

Objective: The primary objective was to estimate the relative efficacy of different exercises for pain, function, physical performance and quality of life (QoL) outcomes in people with knee and hip OA. This research aim was also accompanied by an update of the evidence in overall exercise versus usual care in conservatively managed OA and in those who had undergone joint replacement surgery. The primary time point was at, or nearest to, eight weeks. However, for analysis of post-joint replacements, the primary time point was three months.

Method: A literature search was performed in nine electronic databases and hand searches without limitation in publication date or language. Exercises were classified into aerobic, strengthening, flexibility/skills (flex/skills), mind-body and mixed. Inclusion criteria to be fulfilled were: i) randomised controlled trials (RCTs) of knee OA or hip OA or mixed knee and hip OA, including postoperative participants; ii) RCTs comparing one type of exercise with another type, or other non-exercise interventions; and iii) provision of outcomes for pain, function, performance or QoL. The first search was performed in December 2015 and a final update was conducted in December 2017. Upon completion of screening, the most widely used non-exercise control group was identified to perform NMA. Effect size (ES) of exercise benefits was measured as standardised mean difference between groups and was reported along with 95% confidence intervals (CIs) for conventional meta-analysis (performed using frequentist approach), and with 95% credible interval (CrI) for network meta-analysis (performed using Bayesian approach).

Results: In the first search, 13672 citations were identified from electronic databases and hand searches. After removal of duplicates and full texts screening, 199 citations (185 trials) met the inclusion criteria. After the update in December 2017 40 more eligible citations (32 trials) were added, giving a total of 239 citations (217 trials). However, only 106 trials were deemed eligible for analysis in this thesis because the focus was on the single most common non-exercise comparator used – usual care. In the conventional meta-analysis which was performed to examine the overall effect of exercise versus usual care, 77 trials (n=6472) contributed to the analysis. In the network meta-analysis that was performed to compare the relative efficacy of different exercises, 103 trials (n=9134) were included. Finally, in the meta-analysis of exercise versus usual care on postoperative outcomes, only nine trials (n=467) were eligible.

Compared to usual care, exercise conferred higher ES for pain (ES 0.56; 95%CI 0.44, 0.68), function (ES 0.51; 95%CI 0.38-0.63) and performance (ES 0.46; 95%CI 0.35, 0.57) compared to QoL (ES 0.21; 95%CI 0.11, 0.31) at eight weeks. Single component exercise, particularly aerobic exercise, was most likely to be the most effective for improving OA outcomes. For many outcomes, the differences between different exercises were not statistically significant, except for pain where aerobic or mind–body exercise were significantly better than mixed exercise. These results were generally driven by studies on knee OA, and by strengthening and mixed exercises. Overall evidence indicates that exercise therapy had better benefits in people with knee OA, in people not on the waiting-list for surgery, and in trials with younger participants (mean age <60-years old). Peak benefits were demonstrated at 2-3 months after commencement of exercise and were likely to remain significantly better than usual care beyond this point for another three months.

There were insufficient trials on subsets of hip OA and mixed OA, and OA awaiting surgery (mainly total joint replacement) to permit extensive analyses. The number of trials investigating postoperative outcomes was also small but were sufficient to be pooled for exploratory analysis. The results at three months after surgery suggest that there were no differences between groups given pre-operative exercise and groups given usual care. Although differences between groups were in favour of exercise for pain (ES 0.13), function (ES 0.25) and performance (ES 0.03), none were statistically significant.

Conclusion: Exercise significantly improves pain, function and performance for knee and hip OA, with single component exercises (particularly aerobic exercise) being consistently better than mixed exercise. However, due to the small number of studies, the results for particular OA subsets (i.e. hip and mixed OA, OA awaiting total joint replacement and postoperative OA) were uncertain. This indirectly reflects the gap in the evidence and the need for future confirmatory studies.

Item Type: Thesis (University of Nottingham only) (PhD)
Supervisors: Zhang, Weiya
Doherty, Michael
Hall, Michelle
Keywords: Osteoarthritis; Exercise therapy; Quality of life outcomes
Subjects: W Medicine and related subjects (NLM Classification) > WE Muscoskeletal system
Faculties/Schools: UK Campuses > Faculty of Medicine and Health Sciences > School of Medicine
Item ID: 57278
Depositing User: Goh, Siew
Date Deposited: 28 Sep 2023 14:23
Last Modified: 28 Sep 2023 14:23

Actions (Archive Staff Only)

Edit View Edit View