Methods

Design: parallel randomised controlled trial

Country: USA

Setting: study run through university but location of intervention delivery (e.g. clinic, community yoga studio) not specified

Participants

30 women with urge‐predominant urinary incontinence

Types of incontinence: all women had urge‐predominant UI

Diagnostic criteria for incontinence: not specified

Severity and duration of incontinence: at baseline, the 3‐day average number of UI episodes for yoga participants was median 2.67 (range 1.67 to 6.33) and for MBSR participants was median 3.00 (1.67 to 9.33)

Age: yoga median age 58 (range 22 to 79) years; MBSR median age 59 (range 34 to 74) years

Pregnant/postpartum/menopausal status: "most were postmenopausal"

Weight: BMI range 20.34 to 36.15 kg/m²; yoga median 25.82 (range 20.90 to 36.15) kg/m²; MBSR median 26.15 (range 20.34 to 33.22) kg/m²
Recruitment: radio and newspaper within the community and poster advertisements at the university hospital, outpatient clinics and campus sites
Inclusion criteria: "[W]omen aged 18 years or older who had 5 or more UUI episodes, with urge predominance (urge >50% of total incontinent episodes), on a 3‐day voiding diary."
Exclusion criteria: "...anticholinergic medication use within 2 weeks of baseline assessments and past nonpharmacologic treatment of UUI such as supervised behavioral therapy, supervised or unsupervised physical therapy, supervised biofeedback transvaginal electrical stimulation, tibial nerve stimulation, sacral neuromodulation, and botulinum toxin bladder injections. Women were also excluded if they had a past diagnosis of painful bladder, interstitial cystitis, and/or neurological disorder.”

Interventions

Yoga group: 8 weekly yoga classes following "basic principles of yoga" and focusing on physical poses and relaxation. "Information was given on restorative yoga that emphasized the use of gravity, props, and floor poses for muscle relaxation. No education was given on breathing techniques in this yoga control program. Class time was also spent on education about the 7 chakras, self‐massage, and essential oils and how they can also assist in muscle relaxation."
Teacher(s): the classes were taught by a certified massage therapist/yoga instructor

Home practice: not described

Mindfulness‐based stress reduction (MBSR) group: 8 weekly group meetings following the traditional protocol of the Center for Mindfulness at the University of Massachusetts Medical Center. "The 8‐week MBSR program is a structured program that teaches participants a variety of meditation practices, mindful‐yoga, walking meditation, and discussions on the relationship between stress, illness, and health."

Teacher(s): the classes were taught by a licensed occupational therapist with teacher training in mindfulness‐based therapy

Home practice: women were given 'A Mindfulness‐Based Stress Reduction Workbook' and 15‐ to 30‐minute tapes to listen to. Women were asked to record instances of formal and informal techniques used in home practice sessions. "On average, the participants did a structured practice 5 times/wk (range, <1‐15.4 times/wk) and unstructured practice 9.7 times/wk (range, <1‐22 times/wk)."

Adherence to yoga and comparison interventions: 11/15 (73%) yoga participants and 13/15 (87%) MBSR participants completed at least 5 of 8 sessions

Common interventions: women were asked not to seek any other incontinence treatment while participating in the study. No women received any of the usual treatments of UUI, including bladder education, fluid management, or pelvic floor muscle exercises. None of the group sessions focused on bladder control.
Co‐interventions: no mention of included or excluded co‐interventions
Duration and follow‐up: interventions were provided for 8 weeks with follow‐up visits at 6 and 12 months

Number of withdrawals, with reasons: loss to follow‐up in the yoga group was 4/15 (27%) at 8 weeks, 5/15 (33.3%) at 6 months and 6/15 (40%) at 1 year. Loss to follow‐up in the MBSR group was 2/15 (13%) at 8 weeks, 5/15 (33.3%) at 6 months, and 3/15 (20%) at 1 year. It was not reported whether the women who did not complete follow‐up included withdrawals from treatment. Reasons for loss to follow‐up were not reported.

Outcomes

Participant report of continence or improvement. Patient Global Impression of Improvement (PGI‐I). Women self reported their overall impression of improvement, and the outcome was dichotomised as much better or very much better versus other responses. Information reported at 8 weeks, 6 months and 1 year.

Condition‐ or symptom‐specific quality of life. Overactive Bladder Health‐Related Quality of Life (OAB‐HRQL). The OAB‐HRQL score is obtained from summing the 13 items on the HRQL subscales of the Overactive Bladder Symptom and Health‐Related Quality of Life Questionnaire (OAB‐q). This score is transformed to a 0‐to‐100 scale in which higher scores reflect better quality of life. Median per cent change and IQR are reported at 8 weeks, 6 months and 1 year.

Quantification of symptoms: number of incontinence episodes. The 3‐day average of incontinence episodes was collected from a diary. Median per cent change and IQR are reported at 8 weeks, 6 months and 1 year.

Other measures of social or emotional impact of incontinence. Bothersomeness, as measured on the Overactive Bladder Symptom and Quality of Life‐Short Form (OABq‐SF). The score from this 6‐item scale is transformed to a 0‐to‐100 scale in which higher scores reflect greater bother and impact. Median per cent change and IQR are reported at 8 weeks, 6 months and 1 year.

Other outcomes collected: urge incontinence episodes, daytime voids, nighttime voids and pad use

Notes

Adverse events: no discussion of safety or adverse events
Measurement of expectations or treatment preferences at baseline: none
Unpublished data: Dr Jan Baker emailed LSW the statistical analysis report and details of the calculation and interpretation of the OAB‐HRQL and the OABq‐SF measures, including the questionnaire and scoring manual, on 27 September 2017. No unpublished baseline or outcome data were used in the review.
Funding: university and government. “This study was funded by an unrestricted grant from the University of Utah and was also supported by the University of Utah Study Design and Biostatistics Center, with funding in part from the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through grant 8UL1TR000105 (formerly UL1RR025764).”

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

Randomisation was carried out by shuffling 'security tint' envelopes; the Cochrane Handbook for Systematic Reviews of Interventions indicates that shuffling envelopes corresponds to a low risk of bias for sequence generation. "Randomization was completed by block randomization at each of the 3 periods to maintain an equal number in each group. When an accurate count was obtained of women who signed the consent for participation in a group session, group assignments were placed in security envelopes and shuffled. Subjects were then allowed to pick an envelope. The envelope was opened in front of the research personnel and was noted on the randomization key along with the subjects’ study number. This randomization key was then placed in a sealed envelope. The randomization key was given to the data manager and was noted as either group 1 or group 2 on the data spread sheet. The statistician was given the randomization key."

Allocation concealment (selection bias)

Low risk

"When an accurate count was obtained of women who signed the consent for participation in a group session, group assignments were placed in security envelopes and shuffled. Subjects were then allowed to pick an envelope."

Blinding of participants and personnel (performance bias)
All outcomes

High risk

All participants and personnel were aware of the interventions being delivered. Study personnel may have been aware that MBSR was the intervention of interest and yoga considered an active control condition.

Blinding of outcome assessment (detection bias)
Symptom improvement and quality of life

Unclear risk

All outcomes were assessed by the participants themselves, who were not blinded to the interventions. However, since both interventions were active, it is unclear to what extent this may have biased outcome assessment.

Blinding of outcome assessment (detection bias)
Micturitions and incontinence episodes

Unclear risk

All outcomes were assessed by the participants themselves through diary entries, and the participants were not blinded to the interventions. However, since both interventions were active, it is unclear to what extent this may have biased outcome assessment.

Incomplete outcome data (attrition bias)
All outcomes

High risk

There were high rates of loss to follow‐up, and rates differed between groups. Loss to follow‐up in the yoga group was 4/15 (27%) at 8 weeks, 5/15 (33.3%) at 6 months, and 6/15 (40%) at 1 year. Loss to follow‐up in the MBSR group was 2/15 (13%) at 8 weeks, 5/15 (33.3%) at 6 months, and 3/15 (20%) at 1 year. Reasons for loss to follow‐up are not described.

Selective reporting (reporting bias)

Unclear risk

Study was registered and the primary and secondary outcomes from the trial registration are reported. However, only 2 of the secondary outcomes reported in the article are listed in the protocol.

Other bias

Low risk

Groups were comparable on baseline characteristics and the study appeared to be free of conflicts of interest related to funding.

Methods

Design: parallel randomised controlled trial

Country: USA

Setting: study run through university but location of intervention delivery (e.g. clinic, community yoga studio) not specified

Participants

19 women with either stress or urgency urinary incontinence

Types of incontinence: urgency‐predominant: all 12/19 (63%); yoga 6/10 (60%); control 6/9 (67%). Stress‐predominant: all 7/19 (37%); yoga 4/10 (40%); control 3/9 (33%).

Diagnostic criteria for incontinence: not specified

Severity and duration of incontinence: total incontinence episodes/day (based on 7‐day voiding diary) at baseline: yoga mean (SD) 2.77 (1.3); control mean (SD) 2.16 (1.2). Duration of incontinence not specified, but duration of 3 months or longer was an inclusion criterion.

Age: yoga mean (SD) age 60.5 (8.4) years; control mean (SD) age 62.4 (8.3) years.

Pregnant/postpartum/menopausal status: pregnancy in the last year was an exclusion criterion; menopausal status was not reported

Weight: yoga BMI mean (SD) 24.7 (2.7) kg/m²; control BMI mean (SD) 25.8 (3.8) kg/m²
Recruitment: "combination of newspaper advertisements, flyers posted in local community centers and businesses, and direct recruitment from clinician offices."
Inclusion criteria: "[W]omen had to be at least aged 40 years, report experiencing incontinence for at least 3 months, and document at least 7 episodes of incontinence on a screening 7‐day voiding diary, with at least half of those episodes being stress‐type or urgency‐type incontinence."
Exclusion criteria: "Women were excluded if they had severe mobility limitations that would prevent them from participating in a yoga therapy program (inability to walk up a single flight of stairs or at least 2 city blocks on level ground or an inability to stand up from a supine position unaided within 10 seconds) or if they reported formal yoga instruction within the past year or any prior use of yoga specifically to treat incontinence. Other exclusion criteria included pregnancy within the past 6 months; current urinary tract infection or hematuria (assessed by urine dipstick testing) or history of 3 or more urinary tract infections in the past year; major neurologic condition such as stroke, multiple sclerosis, or Parkinson disease; history of congenital defect leading to incontinence, fistula in the bladder or rectum, pelvic cancer or radiation, or interstitial cystitis or chronic pelvic pain; current symptomatic pelvic organ prolapse; body mass index greater than 35 kg/m2; or prior surgery to the urinary tract. Participants also could not have used practitioner‐supervised behavioral, pharmacological, or other clinical treatments (e.g. pessary) for incontinence within the past 3 months or be planning to initiate new clinical incontinence treatments during the study.”

Interventions

Yoga group: 90‐minute Iyengar yoga classes were provided twice per week for 6 weeks. The class focus was on physical poses and relaxation. "Tadasana (mountain
pose), Utkatasana (chair pose), Trikonasana (triangle pose), Malasana (squat pose), Viparita Karani Variation (legs up the wall pose), Salamba Set Bandhasana (supported bridge pose), Supta Baddha Konasana (reclined cobbler’s pose), and Savasana (corpse pose). While teaching these postures, instructors emphasized specific ways of practicing each posture to foster awareness of the pelvic floor structures and increase control over the pelvic floor muscles, in addition to improving general fitness and conditioning and promoting mindfulness, deep breathing, and relaxation."
Teacher(s): classes were taught by an experienced certified instructor and an assistant

Home practice: "Participants were also instructed to practice yoga at home for at least 1 additional hour per week and to record the dates and duration of practice in a home yoga diary. Participants were given a limited set of yoga props (mat, belt, and block) to take home and a manual with written descriptions and pictures depicting each of the key yoga postures featured in the classes. Tips on how to practice each posture safely and comfortably and how to adapt each posture to improve incontinence and pelvic floor function were also provided in the manual."

Wait‐list control group: "Women randomized to the control group did not attend group yoga therapy classes and were instructed to avoid outside yoga instruction for 6 weeks. At the end of the 6‐week study, control group participants were given a $180 gift certificate for yoga classes at a local yoga studio and a limited set of home yoga props (block, mat, and strap) to take home."

Adherence to yoga and comparison interventions: 9 (100%) women attended at least 1 group class, and 6/9 (67%) attended all group classes. 9 women (100%) completed at least 1 hour/week of home practice (self reported).

Common interventions: women were given a pamphlet with standard education and behavioural management strategies for incontinence
Co‐interventions: no mention of included or excluded co‐interventions
Duration and follow‐up: interventions were provided for 6 weeks, and there was a telephone call at 3 weeks and a clinic visit at 6 weeks

Number of withdrawals, with reasons: "After the randomization, but before the start of the yoga therapy program, 1 yoga therapy participant dropped out of the study citing worsened health." No other women in either group withdrew or were lost to follow‐up.

Outcomes

Participant report of continence or improvement. Overall satisfaction with change in urine leakage was measured on a 5‐point Likert scale, ranging from ‘‘very unsatisfied’’ to ‘‘very satisfied’’, dichotomised as "moderately satisfied" or more versus less than "moderately satisfied". Assessed at 6 weeks.

Condition‐ or symptom‐specific quality of life. “Incontinence Impact Questionnaire Short Form (IIQ‐7), a 7‐item measure of the impact of incontinence on physical activities, emotional health, travel, and social relationships.” The IIQ‐7 is measured on a scale from 0 to 100, with higher scores indicating more negative impact of incontinence upon activities, relationships and feelings. Mean (SD) of change is reported at 6 weeks.

Quantification of symptoms: number of micturitions. The mean episodes/day of voiding in the toilet, as measured on a 7‐day voiding diary. Mean (SD) of change is reported at 6 weeks.

Quantification of symptoms: number of incontinence episodes. The mean episodes/day of urinary incontinence, as measured on a 7‐day voiding diary. Mean (SD) of change is reported at 6 weeks.

Other measures of social or emotional impact of incontinence. Bothersomeness, as assessed on the Urogenital Distress Inventory 6 (UDI‐6) (a 6‐item questionnaire that assesses subjective bother associated with incontinence‐related symptoms). Scale is 0 to 100, with higher scores indicating more negative impact of incontinence upon activities, relationships and feelings. Mean (SD) of change is reported at 6 weeks.

Other outcomes collected: yoga adherence, yoga self efficacy, the Patient Perception for Bladder Condition (PPBC) ("a single‐item questionnaire assessing the degree to which participants consider their bladder condition to be a problem"), stress incontinence episodes/day, urgency incontinence episodes/day, daytime incontinence, nighttime incontinence, daytime voids and nighttime voids

Notes

Adverse events: “During a 3‐week telephone call and a 6‐week clinic visit, coordinators asked participants about any negative changes in their health and recorded any reported negative changes as adverse events on standardized forms. Adverse events were considered 'serious adverse events' if they met the standard definition of resulting in death, disability, or hospitalization. Participants were also encouraged to call study staff to report any negative changes in their health between scheduled calls or visits.”
Measurement of expectations or treatment preferences at baseline: none
Unpublished data: none
Funding: university/non‐profit. "This research was supported by a University of California San Francisco Osher Center for Integrative Medicine Pilot Award from the Mt Zion Health Fund." Investigators also report receiving support from the National Institutes of Health and the American Federation for Aging Research.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“Eligible participants were randomly assigned by a computer in a 1:1 ratio to participate in a 6‐week group‐based yoga therapy program (yoga therapy group) or receive no yoga instruction for 6 weeks followed by a gift certificate for local yoga studio classes (control group). Randomization was stratified by incontinence type (stress or stress‐predominant vs urgency or urgency‐predominant).”

Allocation concealment (selection bias)

Unclear risk

Although randomisation was carried out by computer, the exact procedures surrounding allocation were unclear.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

Participants and providers were aware of treatment assignment.

Blinding of outcome assessment (detection bias)
Symptom improvement and quality of life

High risk

Participants self reported symptom improvement, bothersomeness and quality of life. They were not masked to their treatment assignments and the control condition was an inactive, waiting‐list control.

Blinding of outcome assessment (detection bias)
Micturitions and incontinence episodes

Unclear risk

Participants recorded micturitions and incontinence episodes on 7‐day diaries, and the data were abstracted by masked assessors. However, since the participants were aware of treatment assignment, it is possible that this affected their assessment and recording of micturition and incontinence episodes.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

Attrition from yoga group at follow‐up was 10% (1/10), and there was no attrition from the control group.

Selective reporting (reporting bias)

Unclear risk

Study appears to be free of selective reporting. The primary outcome was prespecified in the trial registration, however none of the reported secondary outcomes are mentioned in the trial registration.

Other bias

Low risk

Although groups were different on the IIQ‐7 and in frequency of voiding in toilets at baseline, this is consistent with simple randomisation of a small sample. Funding is university, non‐profit and government, and there is no indication of researcher bias. First author was previously funded by pharma, which is disclosed.