Is yoga a safe therapeutic intervention?
Date of publication of the systematic review: March 2015
Design
Systematic review of 94 randomized controlled trials (RCTs).
Participants
A total of 8430 healthy participants or patients with reported medical condition (mean age range: 10.1 to 84.5 years, overall male %: 29.2%) who had joined the included randomized controlled trial.
Intervention
Except 32 studies without specifying the style of yoga, others were yoga program with Hatha style (8/94), yogic breathing or pranayama (4/96) and other yoga style (50/94), ranged from 1 day to 18 months with a median duration of 10 weeks.
Comparator
Comparison 1: Yoga program versus active treatment, usual care or no treatment.
Major Outcomes
Outcome 1: Intervention-related adverse events;
Outcome 2: Non-serious adverse events;
Outcome 3: Serious adverse events.
Settings
The reviewers did not state whether the studies were conducted in in-patient or out-patient settings.
Comparison    Yoga program versus active treatment, usual care or no treatment
Main Results
Compared to active treatment, usual care or no treatment, yoga program is not significantly associated with higher chance of causing intervention-related adverse events (pooled odds ratio(OR): 1.17, 95% CI: 0.48 to 2.84), non-serious adverse events (pooled OR: 0.92, 95% CI: 0.57 to 1.49) and serious adverse events (pooled OR: 1.02, 95% CI: 0.37 to 2.81).
Comparison: Yoga program versus active treatment, usual care or no treatment among participants of yoga-related RCTs
Outcomes No. of studies (Total no. of participants) No. of events/ No. of participants Heterogeneity test (I2) Pooled OR (95% CI) Overall quality of evidence*
Intervention Comparator
1 94 (8430) Not reported Not reported 0% 1.17 (0.48 to 2.84) High
2 94 (8430) Not reported Not reported 0% 0.92 (0.57 to 1.49) High
3 94 (8430) Not reported Not reported 0% 1.02 (0.37 to 2.81) High
Keys: OR: odds ratio; CI: confidence interval.
Comparison    Yoga program versus active treatment, usual care or no treatment
Main Results
Compared to active treatment, usual care or no treatment, yoga program is not significantly associated with higher chance of causing intervention-related adverse events (pooled odds ratio(OR): 1.17, 95% CI: 0.48 to 2.84), non-serious adverse events (pooled OR: 0.92, 95% CI: 0.57 to 1.49) and serious adverse events (pooled OR: 1.02, 95% CI: 0.37 to 2.81).
Comparison: Yoga program versus active treatment, usual care or no treatment among participants of yoga-related RCTs
Outcomes 1 2 3
No. of studies (Total no. of participants) 94 (8430) 94 (8430) 94 (8430)
No. of events/ No. of participants Intervention Not reported Not reported Not reported
Comparator Not reported Not reported Not reported
Pooled OR (95% CI) 1.17 (0.48 to 2.84) 0.92 (0.57 to 1.49) 1.02 (0.37 to 2.81)
Overall quality of evidence* High High High
Keys: OR: odds ratio; CI: confidence interval.
Conclusion
Benefits
Not applicable.
Harms
Yoga practice is not significantly associated with a higher incidence of intervention-related adverse events, non-serious adverse events and serious adverse events, compared to active treatment, usual care or no treatment. For Outcome 1-3, the overall quality of evidence is high. Further research is unlikely to have an important impact on our confidence in this estimate of effect.
Link to Original Article
https://www.ncbi.nlm.nih.gov/pubmed/26116216
The synopsis is based on the following article:
Cramer H, Ward L, Saper R, Fishbein D, Dobos G, Lauche R. The safety of yoga: a systematic review and meta-analysis of randomized controlled trials. American Journal of Epidemiology. 2015 Jun 26:kwv071.


* Interpretation of quality assessment results:
• Very low: Further research is most likely to have an important impact on our confidence in this estimate of effect.
• Low: Further research is likely to have an important impact on our confidence in this estimate of effect.
• Moderate: Further research is fairly likely to have an important impact on our confidence in this estimate of effect.
• High: Further research is unlikely to have an important impact on our confidence in this estimate of effect.
• Very high: Further research is most unlikely to have an important impact on our confidence in this estimate of effect.

Details of assessment method can be found at Chung VC, Wu XY, Ziea ET, Ng BF, Wong SY, Wu JC. Assessing internal validity of clinical evidence on effectiveness of CHinese and integrative medicine: Proposed framework for a CHinese and Integrative Medicine Evidence RAting System (CHIMERAS). European Journal of Integrative Medicine. 2015 Aug 31;7(4):332-41.