Holger Cramer, Romy Lauche, Petra Klose, Silke Lange, Jost Langhorst, Gustav J Dobos

Breast cancer is the cancer most frequently diagnosed in women worldwide. Even though survival rates are continually increasing, breast cancer is often associated with long-term psychological distress, chronic pain, fatigue and impaired quality of life. Yoga comprises advice for an ethical lifestyle, spiritual practice, physical activity, breathing exercises and meditation. It is a complementary therapy that is commonly recommended for breast cancer-related impairments and has been shown to improve physical and mental health in people with different cancer types. To assess effects of yoga on health-related quality of life, mental health and cancer-related symptoms among women with a diagnosis of breast cancer who are receiving active treatment or have completed treatment. We searched the Cochrane Breast Cancer Specialised Register, MEDLINE (via PubMed), Embase, the Cochrane Central Register of Controlled Trials (CENTRAL; 2016, Issue 1), Indexing of Indian Medical Journals (IndMED), the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) search portal and Clinicaltrials.gov on 29 January 2016. We also searched reference lists of identified relevant trials or reviews, as well as conference proceedings of the International Congress on Complementary Medicine Research (ICCMR), the European Congress for Integrative Medicine (ECIM) and the American Society of Clinical Oncology (ASCO). We applied no language restrictions. Randomised controlled trials were eligible when they (1) compared yoga interventions versus no therapy or versus any other active therapy in women with a diagnosis of non-metastatic or metastatic breast cancer, and (2) assessed at least one of the primary outcomes on patient-reported instruments, including health-related quality of life, depression, anxiety, fatigue or sleep disturbances. Two review authors independently collected data on methods and results. We expressed outcomes as standardised mean differences (SMDs) with 95% confidence intervals (CIs) and conducted random-effects model meta-analyses. We assessed potential risk of publication bias through visual analysis of funnel plot symmetry and heterogeneity between studies by using the Chi(2) test and the I(2) statistic. We conducted subgroup analyses for current treatment status, time since diagnosis, stage of cancer and type of yoga intervention. We included 24 studies with a total of 2166 participants, 23 of which provided data for meta-analysis. Thirteen studies had low risk of selection bias, five studies reported adequate blinding of outcome assessment and 15 studies had low risk of attrition bias.Seventeen studies that compared yoga versus no therapy provided moderate-quality evidence showing that yoga improved health-related quality of life (pooled SMD 0.22, 95% CI 0.04 to 0.40; 10 studies, 675 participants), reduced fatigue (pooled SMD -0.48, 95% CI -0.75 to -0.20; 11 studies, 883 participants) and reduced sleep disturbances in the short term (pooled SMD -0.25, 95% CI -0.40 to -0.09; six studies, 657 participants). The funnel plot for health-related quality of life was asymmetrical, favouring no therapy, and the funnel plot for fatigue was roughly symmetrical. This hints at overall low risk of publication bias. Yoga did not appear to reduce depression (pooled SMD -0.13, 95% CI -0.31 to 0.05; seven studies, 496 participants; low-quality evidence) or anxiety (pooled SMD -0.53, 95% CI -1.10 to 0.04; six studies, 346 participants; very low-quality evidence) in the short term and had no medium-term effects on health-related quality of life (pooled SMD 0.10, 95% CI -0.23 to 0.42; two studies, 146 participants; low-quality evidence) or fatigue (pooled SMD -0.04, 95% CI -0.36 to 0.29; two studies, 146 participants; low-quality evidence). Investigators reported no serious adverse events.Four studies that compared yoga versus psychosocial/educational interventions provided moderate-quality evidence indicating that yoga can reduce depression (pooled SMD -2.29, 95% CI -3.97 to -0.61; four studies, 226 participants), anxiety (pooled SMD -2.21, 95% CI -3.90 to -0.52; three studies, 195 participants) and fatigue (pooled SMD -0.90, 95% CI -1.31 to -0.50; two studies, 106 participants) in the short term. Very low-quality evidence showed no short-term effects on health-related quality of life (pooled SMD 0.81, 95% CI -0.50 to 2.12; two studies, 153 participants) or sleep disturbances (pooled SMD -0.21, 95% CI -0.76 to 0.34; two studies, 119 participants). No trial adequately reported safety-related data.Three studies that compared yoga versus exercise presented very low-quality evidence showing no short-term effects on health-related quality of life (pooled SMD -0.04, 95% CI -0.30 to 0.23; three studies, 233 participants) or fatigue (pooled SMD -0.21, 95% CI -0.66 to 0.25; three studies, 233 participants); no trial provided safety-related data. Moderate-quality evidence supports the recommendation of yoga as a supportive intervention for improving health-related quality of life and reducing fatigue and sleep disturbances when compared with no therapy, as well as for reducing depression, anxiety and fatigue, when compared with psychosocial/educational interventions. Very low-quality evidence suggests that yoga might be as effective as other exercise interventions and might be used as an alternative to other exercise programmes.