RF26 The effectiveness of physical actvitiy interventions for people with osteoarthritis and comorbidity: a meta-analysis of obesity

Abstract

BackgroundOsteoarthritis (OA) is one of the diseases with the highest prevalence of comorbidity. Clinical guidelines recommend physical activity (PA) for people with OA irrespective of comorbidity. Research investigating the effectiveness of PA interventions in OA and comorbidity is needed. Objective: To synthesise existing evidence investigating the effectiveness of PA interventions in adults with OA and obesity.MethodsA systematic review with meta-analysis was conducted (PROSPERO Registration: CRD42017055582). Six electronic databases; MEDLINE, EMBASE, AMED, CINAHL, SportDiscus and CENTRAL were searched for studies from their inception to 29.03.17. Inclusion criteria were: randomised controlled trials (RCTs) comparing the effectiveness of any PA intervention to non-PA control group; including adults aged 45 years old and over with clinical or radiographic OA at any site; at least one of the comorbidities of interest (COPD, depression, diabetes, hypertension, obesity, T2DM); and measuring pain, physical function, quality of life, global health post intervention and adverse events. Included study risk of bias (ROB) was assessed using the Cochrane risk of bias tool. Two reviewers screened titles, abstracts and full text articles, checked data extraction, and carried out ROB assessment. Random-effects model meta-analysis pooled outcomes from sufficiently homogeneous studies to calculate effect sizes (Standardized Mean Difference (SMD) with 95% confidence interval (CI)). Meta-analysis findings of the OA and obesity subgroup are reported.ResultsThe literature search retrieved 8171 citations of which 14 studies (n=4224 participants) were included in the full review, with 9 (n=1382 participants) analysed in the OA and obesity subgroup. PA interventions included: aquatic, aerobic, strengthening and functional activity; of 1–18 months in duration.Four studies of OA and obesity measuring either Western Ontario Osteoarthritis Index (WOMAC) pain, WOMAC function or Six Minute Walking Test (6 MWT) and were included in three meta-analyses. Best estimates showed PA to improve WOMAC pain (n=3 studies; n=547 participants; SMD=-0.09 (95% CI) −0.65, 0.47), improve WOMAC function (n=3 studies, n=415 participants; SMD=-0.35 (95% CI) −0.89, 0.18) and the 6 MWT (n=4 studies, n=573 participants; SMD=-0.93 (95% CI) −0.49, 2.35). However, results were not statistically significant. There was substantial between-trial outcome heterogeneity (I²=89.4% (p=0.000); 77.5% (p=0.012); 97.8% (p=0.000); respectively); results should be interpreted with caution. ROB domain judgements were generally either low or unclear. A small minority of judgements were at high risk of bias.ConclusionBest estimates suggest small beneficial effects of physical activity on WOMAC pain, WOMAC function and the 6 MWT. Mixed effectiveness among individual RCTs was likely due to heterogeneous intervention types, intensity and duration.