Galantino 2004.

Study characteristics
Methods	Randomized controlled parallel‐group trial.
Participants	22 participants with chronic LBP. Settings: setting for trial and location of yoga intervention delivery not described. Country: USA. Recruitment: self‐referral through newspaper advertisements and referral through healthcare practitioners. Inclusion criteria: men and women aged 30–65 years with back pain for > 6 months and > 2 previous conservative medical interventions (physical therapy and chiropractic) without prolonged relief. Exclusion criteria: current history of chronic systemic disease, previous yoga experience, changes in pain medication during past 14 days.
Interventions	Yoga group: (n = 11) formal 1‐hour hatha yoga class twice per week for 6 weeks. Home practice: suggested 1 hour per day but not mandated or monitored. Control group: (n = 11) no treatment during observation period. Common interventions: usual care continued for both groups. Co‐interventions: changes in pain medication were not allowed during study. Duration and follow‐up: interventions provided for 6 weeks and there was an additional follow‐up at 3 months for yoga participants only.
Outcomes	Back‐specific function (Oswestry Disability Index) at 6 weeks. Depression (Beck Depression Inventory) at 6 weeks. Other outcomes collected: Sit and Reach Test and Functional Reach Test.
Notes	Adverse events: no discussion of safety or adverse events. Measurement of expectations or treatment preferences at baseline: none. Funding: not reported.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used random numbers to generate sequence.
Allocation concealment (selection bias)	Unclear risk	No details reported in publication.
Blinding of participants	High risk	No blinding as control participants received no treatment; outcomes based on self‐assessment.
Blinding of personnel/providers	High risk	No blinding.
Blinding of outcome assessors	High risk	Participants were not blinded, and outcomes were self‐reported. No mention of blinding of those who collected the information on outcomes from the participants.
Incomplete outcome data (attrition bias) All outcomes	High risk	Attrition > 30% in control group. 6/11 control group participants had missing outcomes and were treated as 'failures' in a dichotomous analysis.
Selective reporting (reporting bias)	Unclear risk	No protocol available.
Group similarity at baseline	High risk	Beck Depression Inventory substantially higher in control group.
Co‐interventions	Unclear risk	Not possible to assess; recorded by participants but not reported.
Compliance	Unclear risk	No information on class attendance.
ITT analysis	Low risk	ITT analysis stated.
Timing of outcome assessments	Low risk	Outcome assessments at a set point.
Other bias	Low risk	No other biases identified.