Secondary Logo

Journal Logo

Systematic Review

Effects of Yoga on Quality of Life and Pain in Women With Chronic Pelvic Pain: Systematic Review and Meta-Analysis

Russell, Natalie DPT1; Daniels, Bevin PT, DPT2; Smoot, Betty PT, DPTSc1; Allen, Diane D. PT, PhD1

Author Information
Journal of Women's Health Physical Therapy: July/September 2019 - Volume 43 - Issue 3 - p 144-154
doi: 10.1097/JWH.0000000000000135

Abstract

Erratum

In the article entitled “Effects of Yoga on Quality of Life and Pain in Women With Chronic Pelvic Pain: Systematic Review and Meta-Analysis” by Natalie Russell et al published in the July/September 2019 issue, the authors discovered errors in the calculations they performed for the published article. In this erratum, published and corrected numbers and wording (in bold print ) are specified within section headings for the abstract and body of the article as well as Tables 3 to 4 and Figures 2 to 5.

The published version reads: “Statistically significant improvements were seen following the yoga intervention for within-group analysis of QOL (ES = -1.4, CI: -1.8 to -1.1) and pain (ES: -2.2, CI: -2.7 to -1.6). Between-group analysis found statistically significant differences in QOL (ES= -1.5, CI: -2.0 to -1.0) and pain (ES = -1.4, CI: -1.7 to -1.0) favoring the yoga group.”

The corrected version should say: “Statistically significant improvements were seen following the yoga intervention for within-group analysis of QOL (ES = -1.3, CI: -2.0 to -0.7 ) and pain (ES = -2.2, CI: -3.8 to -0.7 ). Between-group analysis tended to favor yoga in QOL (ES = -1.6 , CI: -4.3 to 1.0 ) and pain (ES = -2.4 , CI: -5.2 to 0.3 ); sizes of differences reveal potential for clinical importance .”

Published: “The results support the use of yoga to improve pain and QOL...”

Corrected: “The results support the potential for yoga to improve pain and QOL...”

In the first paragraph, “74 participants” should be corrected to specify 74 yoga participants.

In the last sentence, the published wording is: “individual study data and ESs are presented in Table 3.”

The corrected wording is: “individual study data and ESs are presented in Table s 3 and 4 .”

In the second paragraph, the published version reads, “Statistically significant improvements in QOL were seen for both within group (Table 3; ES = -1.4, CI: -1.8 to -1.1) and between group (Table 4; ES = -1.5, CI: -2.0 to -1.0) meta-analyses.”

The corrected version should read, “Statistically significant improvements in QOL were seen for the within-group meta-analysis (Table 3; ES = − 1.3, CI: − 2.0 to -0.7 ), but the between-group meta-analysis was not significant (Table 4; ES = − 1.6, CI: − 4.3 to 1.0 ).” The same paragraph should end with the following corrected sentence: “For within-group analysis, the grand ES corresponded to an expected increase of 15.9 points in the psychological domain of the WHOQOL-BREF, and between-group analysis corresponded to an expected difference of 15.5 points between the control and intervention group.”

The first published sentence reads, “Statistically significant improvements in pain were seen for within group (Table 3; ES = -2.2, CI: -2.7 to -1.6) and between-group (Table 4; ES = -1.4, CI: -1.7 to -1.0) meta-analysis.”

The corrected sentence should read, “Statistically significant improvements in pain were seen for within-group meta-analysis (Table 3; ES = -2.2, CI: − 3.8 to – 0.7 ), but between-group (Table 4; ES = − 2.4, CI: − 5.2 to 0.3 ) meta-analysis was not significant .” Later in the same paragraph, these corrections translate “to an expected decrease of 3.5 points on the Numeric Pain Rating Scale following intervention,” and “The ES for between-group analysis translated to a difference of 20.4 points on the visual analog scale favoring the yoga group.”

The Discussion should be modified to reflect these corrections. In the first paragraph, the published final sentence reads, “Within-group and between-group results were statistically significant, and, likely, clinically meaningful in favor of the yoga intervention.”

The corrected version should read, “Within-group results were statistically significant. Between-group results from two studies tended to favor yoga but did not reach statistical significance.

Statistical significance should be removed from the first sentence, and the point values of the clinical units on the WHOQOL-BREF are corrected to 15.9 for the within-group effect and 15.5 for the between-group effect. The published final sentence in this section reads, “...the within-group analysis translates to a 50% improvement and the between group analysis translates to a 66% difference between groups favoring the yoga group.”

The corrected sentence should read, “...the within-group analysis translates to a 33% improvement and the between group analysis translates to a 36% difference between groups favoring the yoga group. These percentages are likely to be clinically important.

The published first sentence reads, “Regarding pain, the ESs for within group and between group were also both large (ES > 0.8) and statistically significant.”

The corrected sentence should read, “...the ESs within group and between group were also large (ES > 0.8).” The corrected average within-group improvement on the NPRS is 3.5 points which “equates to a 73 % improvement in pain.” Further corrections read, “A difference of 20.4 points was seen favoring the yoga intervention, with an expected 30% difference in pain scores between groups. This 20.4 -point difference exceeds the MCID of 18 to 19 points as described by Hagg et al.” 25

The last sentence of the third published paragraph reads, “While only 2 studies were included in the between-group analyses, both ESs indicated reductions in pain and improvement in QOL.”

The corrected sentence should read, “While only 2 studies were included in the between-group analyses, both ESs suggested clinically important differences in pain and QOL.”

In the sixth paragraph of this section, the published final sentences read, “The magnitudes for the pooled ESs were large for within-group and between-group comparisons for QOL and pain. Furthermore, the CIs were narrow in this review, thus the ES estimate can be considered to be fairly precise. The precision around the ES estimates and the large magnitude of the ESs for QOL and pain were such that conclusions could be drawn about the benefit of treatment despite the small number of studies.”

The corrected sentences should read as follows, “The magnitudes for the pooled ESs were large enough to suggest clinical importance in within-group and between-group comparisons for QOL and pain. Precision regarding improvements within groups was sufficient for statistical significance. However, the CIs were wide for between-group analyses; additional data would be needed to increase precision regarding differences between yoga and alternative treatments.

In the seventh (final) paragraph of this section, the published sentences about sample sizes read, “Finally, all of the studies had small sample sizes (n = 16-30), and the combined sample in this review was only 74, which may limit the generalizability of the results. However, the small samples yielded statistically significant results and thus did not limit power or increase the risk of a type 2 error.”

The corrected sentences should read, “Finally, all of the studies had small sample sizes, and the combined sample for the yoga group was only 74, which may limit the generalizability of the results. The small samples yielded statistically significant within-group results; between-group results had limit ed power and increase d risk of a type 2 error.

The first two published sentences read, “...provide moderate evidence to support the use of yoga to improve and .... Despite low to moderate quality of evidence, and a limited number of studies, the effects of the yoga interventions were large....”

The corrected sentences should read, “...provide moderate evidence to support the potential for yoga to improve and .... Despite low to moderate quality of evidence, and a limited number of studies, within-group effects of the yoga interventions were large....”

Third column heading is published as “Pretest N.” Corrected, the heading omits Pretest.

Fourth column: Pretest Mean (SD) for Saxena pain reads “33.3 (4).” Corrected, it reads, “ 66.9 (8.4) .”

Fifth column: posttest Mean (SD) for Huang Quality of Life reads “0.9 (0.8).” Corrected, it reads “0.9 (0. 7 ).” Goncalves Quality of Life reads “41.5 (15.7).” Corrected, it reads “41. 7 (15.7).” Saxena pain reads “28 (4.6).” Corrected, it reads “ 34.1 (10.1) .”

For ES and 95% CI: Goncalves Quality of Life the numbers read “-1.1 (-1.8, -0.4).” Corrected numbers are “-1.1 (-1. 7 , -0. 5 ).” Grand total for Quality of life reads “-1.4 (-1.8, -1.1).” Corrected, it reads “-1. 3 (-2.0 , - 0.7 ).” Huang pain reads “-1.4 (-2.2, -0,6).” Corrected, it reads “-1. 0 (- 1.7 , -0. 3 ).” Goncalves pain reads “-1.8 (-2.6, -1.1).” Corrected it, reads “-1.8 (-2. 5 , -1. 2 ).” Saxena pain reads “-1.3 (-1.9, -0.8).” Corrected, it reads “ -3.9 (-4.8 , -3.1 ).” Grand total pain reads “-1.4 (-1.7, -1.0).” Corrected, it reads “ -2.2 (-3.8, -0.7 ).”

The footnote “a” reads: “The random effects model was used for within-group QOL, between-group pain, and between-group QOL analysis because the P value for the Q heterogeneity statistics was less than .05.” The corrected footnote reads “ a The random effects model was used for within-group pain and QOL analyses because the P value s for the Q heterogeneity statistics were less than 0.05.”

Fifth column: Goncalves pain, mean (SD) reads “55.1 (22.0).” Corrected, it reads “55.1 ( 21.5 ).” Saxena pain reads “67.1 (6.6).” Corrected, it reads “67. 0 (6.6).”

For ES and 95% CI: Goncalves quality of life, the numbers read “-0.3 (-1.0, 0.3).” Corrected numbers read “-0.3 ( -1.0, 0. 4 ).” Grand total for quality of life reads “-1.5 (-2.0, -1.0).” Corrected, it reads “-1. 6 (-4.3, 1.0 ).” Grand total for pain reads “-2.2 (-2.7, -1.6).” Corrected, it reads “-2. 4 (-5.2, 0.3 ).”

The footnote “a” reads: “ a The random effects model was used for within-group QOL, between-group pain, and between-group QOL analysis because the P value for the Q heterogeneity statistics was less than .05.” The corrected footnote reads “ a The random effects model was used for between-group pain and QOL analyses because the P value s for the Q heterogeneity statistics were less than 0.05.”

Corrected Tables 3 to 4 ( http://links.lww.com/JWHPT/A34 ) and Figures 2 to 5 ( http://links.lww.com/JWHPT/A35 ) are available online.

Journal of Women’s Health Physical Therapy. 44(4):E19-E21, October/December 2020.

INTRODUCTION

The American College of Obstetricians and Gynecologists defines chronic pelvic pain (CPP) as noncyclical pain lasting 6 months or more affecting the pelvis, anterior abdominal wall, lower back, or buttocks that is serious enough to cause disability or lead to medical care.1 Up to 20% of women in the United States are affected by CPP.1 Direct and indirect costs of CPP reach $2 to $3 billion annually.1,2 Twenty-five percent of women with CPP spend 2 to 3 days in bed each month3 and experience decreased quality of life (QOL)4 or feelings of sadness, depression, and anxiety.3,4

The cause of CPP is often multifactorial and heterogeneous5,6 and may be linked to gynecological, urological, gastrointestinal, or other medical conditions6 such as endometriosis, interstitial cystitis, musculoskeletal pain, and irritable bowel syndrome.2 In addition, poor general health, catastrophizing, and self-reported stress are associated with CPP.7

Treatment of CPP is often multidisciplinary5 and may include pharmacological interventions,2,4,6,8 pelvic floor physical therapy,2,6,8 psychotherapy,6 and lifestyle modification.2 Physical therapy may include biofeedback, soft tissue mobilization,8 and exercise to stretch and/or strengthen pelvic floor muscles and their synergists in the hip and the back. Common pharmacological treatments range from opioid analgesics,6 tricyclic antidepressants,4,6,8 progesterone-like medications,9 and anticonvulsant medications.8 However, pharmacological treatments often have side effects and do not address the maladaptive effects of pain on the nervous system.

When pain becomes chronic, maladaptive neuroplastic changes can occur resulting in increased sensitivity and decreased precision of the nervous system to detect tissue damage.10 The International Association for the Study of Pain describes this sensitization of tissues to lead to nociplasticity, in which the sensation of pain occurs in uninjured tissues.11 Thus, treatment of chronic pain conditions should include strategies that address the sensitization of the nervous system.5 However, many treatment options for CPP do not adequately address these changes in the nervous system, which may explain why some women seek alternative or complementary approaches, such as yoga, to manage their pain.6

Yoga aims to unite mind and body in a way that promotes health.12 There are many different styles of yoga, each with unique components, but most include physical postures (asana), breathing exercises (pranayama), and meditation (dhyana).12 Because yoga can be practiced in the community or at home, it offers women an option for self-management.6

Yoga postures may help individuals identify and stretch the pelvic floor muscles6 as well as stretch and strengthen the pelvic floor synergists of the hip and the back. It has been theorized that hip muscles, such as the obturator internus, may influence the pelvic floor muscles via the obturator fascia.13 Thus, a wide legged forward fold in which the lower extremities are internally rotated, such as prasarita padottanasana, may stretch the obturator internus and put tension on the obturator fascia, allowing the pelvic floor muscles to stretch symmetrically. Breathing exercises may also influence the pelvic floor as a result of the respiratory diaphragm synergy in which the respiratory and pelvic diaphragms move in synchrony: caudally with inspiration and cranially with exhalation.14 Finally, yoga may help address the sensitization of the nervous system.15 Previous studies have reported that individuals who practice yoga experience a decrease in sympathetic nervous system activity, inflammatory markers, and cortisol levels, in addition to an increase in strength, flexibility, and cardiorespiratory capacity.12,16

Yoga has been used to manage conditions similar to CPP. In a recent systematic review, Chang et al17 found that yoga is effective at decreasing pain among patients with chronic low back pain. Kinser et al18 examined 15 studies that used physical activity in managing pregnancy-related low back and pelvic pain. Four of the 15 studies specifically used yoga as the physical activity, including 2 randomized controlled trials, 1 small nonrandomized controlled trial, and 1 small uncontrolled trial. Results showed that yoga-based interventions led to statistically significant improvements in pain.18

Currently, no systematic review has evaluated the use of yoga for patients with CPP. Physical therapists already incorporate yoga in clinical practice, and continuing education courses target the use of yoga for pelvic floor dysfunction and other orthopedic conditions. However, the research is not sufficient to inform clinicians regarding the benefit of yoga in patients with CPP. In addition, some yoga practices may have the potential to strain pelvic floor muscles and thereby precipitate or worsen pelvic pain.6 Therefore, some clinicians may hesitate to try yoga in this population because of fear of harm.

The purpose of this review was to determine the effect of yoga on QOL and pain among women with CPP. The hypotheses were that yoga would decrease pain and improve QOL within groups and result in better outcomes than comparison groups. The results of this review may provide clinicians with knowledge to make informed decisions about whether incorporating yoga into treatment will ultimately benefit patients.

METHODS

Search Strategy and Inclusion and Exclusion Criteria

A comprehensive review was conducted using the electronic databases PubMed, Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Physiotherapy Evidence Database (PEDro). The term yoga was searched in combination with the following words: pelvic pain, sexual pain, vulvodynia, cystitis, pelvic girdle, pelvic floor, dyspareunia, endometriosis, bladder pain, and coccydynia. The last search took place on December 29, 2017.

Studies were included if they were experimental in design, published in an academic journal, written in English, used yoga as the primary intervention, and assessed outcomes of pain and QOL. Studies that evaluated a population of women with endometriosis were not excluded because of the potential for women with endometriosis to experience associated noncyclical CPP.4 However, studies that looked at dysmenorrhea without a diagnosis of endometriosis were excluded. Studies were also excluded if the interventions of meditation, relaxation, stretching, exercise, or pranayama were examined in isolation.

Study Selection

Study selection began with removal of duplicates followed by removal of any article that did not contain “yoga” or “pelvic pain” in the title or the abstract. Remaining records were screened against inclusion and exclusion criteria. A recursive search of the remaining articles was performed. A second reviewer confirmed that the results of the search met the eligibility criteria.

Data Processing

The means and standard deviations (SDs) for outcomes of pain and QOL were extracted from individual studies for meta-analysis. If the data were reported as a subscale, then the subscale was used.

Within-group analyses were performed to determine the effect of yoga on pain and QOL for the intervention group. Effect sizes (ESs) using Cohen d and 95% confidence intervals (CIs) were calculated for each study using the pre- and postintervention means and SDs for each outcome. After weighting each ES by its inverse variance, within-group grand ESs and 95% CIs were calculated.

Between-group analyses were performed to compare the outcomes of pain and QOL in the intervention group compared with the control group. Effect sizes and 95% CIs were calculated for each study using the postintervention means and pooled SDs from both groups for each outcome. After weighting each ES by its inverse variance, between-group grand ESs and 95% CIs were calculated.

The heterogeneity statistic Q was calculated to determine the best model for pooling data. Fixed-effect analyses were performed when heterogeneity was not statistically significant (P > .05), and random-effects analyses were used when statistical heterogeneity reached statistical significance (P < .05). Effect sizes were converted back into clinical units by multiplying the grand ES by the preintervention SD from 1 study for within-group analyses and the pooled SD for between-group analyses. Data were synthesized and analyzed using Microsoft Excel for Mac Version 15.40.

Risk of Bias in and Across Studies

Quality of evidence was evaluated using the hierarchy of evidence described by Jewell19 in which Level 1a evidence is the highest and includes systematic reviews of homogeneous validated studies, and level 5 is the lowest and includes expert opinion on the basis of bench research. For randomized controlled trials or randomized case-controlled trials, the PEDro scale (with a documented reliability of intraclass correlation coefficient= 0.68; 95% CI, 0.57-0.76)20 was used to rate the potential for bias. On the PEDro scale, studies with a 6 of 10 or better are considered high quality.21 In the case of a single-arm trial, the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist was used to evaluate risk of bias. One point was given for each of the 22 criteria.

RESULTS

Study Selection

Two independent searches yielded 100 relevant articles with 37 duplicates. The remaining 63 titles and abstracts were screened for relevance, 54 of which were excluded because yoga or pelvic pain did not appear in the title or abstract, or pain was pregnancy-related. A recursive search of the remaining 9 articles was performed but did not result in any new articles. Six of the remaining articles were excluded because they did not meet the eligibility criteria. Reasons for exclusion included yoga as a secondary intervention, nonexperimental study design, or lack of reporting of quantitative data. Three articles remained for the systematic review, which included 1 single-arm trial, 1 randomized controlled trial, and 1 randomized case-controlled trial. The study selection process is shown in Figure 1.

Figure 1.
Figure 1.:
PRISMA diagram of study selection process. CINAHL indicates Cumulative Index to Nursing and Allied Health Literature; CPP, chronic pelvic pain; PEDro, Physiotherapy Evidence Database.

Summary of the Studies

The PEDro and STROBE scores are presented in Table 1. A brief summary of the studies can be found in Table 2, including participants, treatment groups, interventions, and outcomes. Tables 3 and 4 provide the data extracted from the studies, and Table 5 provides a list of the yoga postures (asanas) used in each study.

Table 1. - Results of the PEDRO and STROBE Checklists (Continued)
PEDro Criteria (0 = No or Not Described, 1 = Yes)
Goncalves et al4 Saxena et al22
Eligibility criteria were specified 1 1
Random allocation 1 1
Allocation was concealed 0 0
Blinding of subjects 0 0
Blinding of therapists 0 0
Blinding of assessors 0 0
Follow-up from 85% of participants 0 1
Intention-to-treat analysis 1 1
Between-group comparison 1 1
Point estimates and variability 1 1
Total PEDro scores 5 6
STROBE Criteria (0 = No or Not Described, 1 = Yes)
Huang et al6
Title and abstract
 Study design and outcomes described 1
Introduction
 Background and rationale 1
 Objectives and hypothesis stated 1
Methods
 Study design 1
 Setting, recruitment, data collection 1
 Participants: eligibility criteria, selection, follow-up 1
 Variables: outcomes, predictors, confounders 1
 Data sources/measurement 1
 Addresses sources of bias 0
 Explains how study size was determined 0
 How quantitative variables were handled 0
Results
 Participants: n reported at each stage, reasons for nonparticipation 1
 Descriptive data: characteristics of participants, follow-up time 1
 Outcome data 1
 Main results: unadjusted estimates, category boundaries 1
 Other analyses: subgroup analyses 0
Discussion
 Key results 1
 Limitations 1
 Interpretation 1
 Generalizability 0
Other information
 Funding 1
Total STROBE score 16
Abbreviations: PEDro, Physiotherapy Evidence Database;STROBE, Strengthening the Reporting of Observational Studies in Epidemiology.

Table 2. - Summary of Studies
Study Level of Evidence Subjects Study Design Description Outcomes
Huang et al6 (2017) 4 n = 16 Single-arm trial Iyengar yoga 90 min 2×/wk for 6 wk. Pain = NPRS, QOL = IPP
Goncalves et al4 (2017) 2b Yoga n = 28, control n =12 Randomized controlled trial Control group: medication and/or physical therapy 1×/wk.
Yoga group: 2-h sessions 2×/wk for 8 wk.
Pain = EHP-30 pain domain, QOL = EHP-30 emotional well-being domain
Saxena et al22 (2017) 3b Yoga n = 30, control n =30 Randomized case-controlled trial Control group: NSAIDs as needed. Yoga group: 1 h of yoga 5×/wk for 8 and NSAIDs as needed. Pain = 100 mm VAS, QOL = WHOQOL-BREF psychological domain
Abbreviations: EHP-30, Endometriosis Health Profile; IPP, Impact of Pelvic Pain Questionnaire; N, number of participants; NPRS, Numeric Pain Rating Scale; NSAIDs, nonsteroidal anti-inflammatory drug; QOL, quality of life; VAS, visual analog scale; WHOQOL-BREF, World Health Organization Quality of Life Brief Version.

Table 3. - Individual Study Data and Effect Sizes Using Cohen d for Within-Group Meta-Analysisa
Outcome Study Pretest N Pretest Mean (SD) Posttest Mean (SD) ES 95% CI
Quality of life Huang et al6 16 1.7 (0.9) 0.9 (0.8) −0.9 −1.6 to −0.2
Goncalves et al4 28 65.0 (21.4) 41.5 (15.7) −1.1 −1.8 to −0.4
Saxena et al22 30 −47.6 (12.2) −71.9 (10.0) −2 −2.6 to −1.4
Grand 74 1.4 1.8 to1.1
Pain Huang et al6 16 4.8 (1.6) 3.2 (2.3) −1.4 −2.2 to −0.6
Goncalves et al4 28 60.8 (15.6) 32.4 (22.0) −1.8 −2.6 to −1.1
Saxena et al22 30 33.3 (4) 28 (4.6) −1.3 −1.9 to −0.8
Grand 74 1.4 1.7 to1.0
Abbreviations: CI, confidence interval; ES, effect size; N, number of participants.
aThe fixed-effects model was used for within-group pain analysis as the P value for the Q heterogeneity statistics was greater than .05. The random-effects model was used for within-group quality of life. Quality of life scores from the study by Saxena et al22 were given a negative sign for consistency of data analysis.

Table 4. - Individual Study Data and Effect Sizes Using Cohen d for Between-Group Meta-Analysisa
Outcome Study Control N Yoga N Control Mean (SD) Yoga Mean (SD) ES 95% CI
Quality of life Goncalves et al4 12 28 48.2 (29.5) 41.7 (15.7) −0.3 −1.0 to 0.3
Saxena et al22 30 30 −43.2 (9.3) −71.9 (10.0) −3.0 −3.7 to −2.2
Grand 42 58 1.5 2.0 to1.0
Pain Goncalves et al4 12 28 55.1 (22.0) 32.4 (22.0) −1.0 −1.8 to −0.3
Saxena et al22 30 30 67.1 (6.6) 34.1 (10.1) −3.9 −4.7 to −3.0
Grand 42 58 2.2 2.7 to1.6
Abbreviations: CI, confidence interval; ES, effect size; N, number of participants.
aThe random-effects model was used for within-group QOL, between-group pain, and between-group QOL analysis because the P value for the Q heterogeneity statistics was less than .05. QOL scores from the study by Saxena et al22 were given a negative sign for consistency of data analysis.

Table 5. - List of Poses Included in Each Studya
Huang et al6 Goncalves et al4 Saxena et al22
Supta Padagusthasana (reclining big toe pose) Parighasana Matsyasana
Ardha Ananda Balasana (half happy baby pose) Parigrasana with a twist Vrikshasana
Adho Mukha Svanasana (downward facing dog pose) Upavista Konasana in Dandasana Trikonasana
Viparita Karani (legs up the wall pose) Dandasana Veerabhadrasana
Salamba Setu Bandhasana (supported bridge pose) Purvottanasana with legs crossed Utannapadmasana
Supta Baddha Konasana (reclining bound angle pose) Setu bandha sarvangasana Pavanmuktasana
Balasana (child's pose) Pavanamuktasana Shalabhasana
Baddha Konasana (butterfly or bound angle pose) Jatara Parivartanasana Dhanurasana
Prasarita Padottanasana (wide-legged forward bend pose) Bhujangasana Makrasana
Anjaneyasana (low lunge pose; 2 variations) Marjaryasana Suptavajrasana/Ustrasana
Savasana (corpse pose) Baddha konasana on the wall Padmasana
Viparita Karani (Legs up the Wall) Viparita Karani (Legs up the Wall)
Baddha Konasana (Butterfly or Bound Angle) Baddha Konasana (Butterfly or Bound Angle)
Savasana (corpse pose) Savasana (corpse pose)
Balasana (child's pose) Upavista Konasana in Dandasana
Adho Mukha Svanasana (downward facing dog) Dandasana
Bhujangasana
aPoses described in boldface are shared across studies.

Quality of Evidence and Risk of Bias

The PEDro scale was used to evaluate the risk of bias in the 2-group studies by Goncalves et al4 and Saxena et al,22 which received a score of 5 and 6, respectively (Table 1). The STROBE checklist was used to evaluate the risk of bias in the single-arm study by Huang et al,6 which met 16 of the 22 criteria (Table 1). Levels of evidence varied from 2b to 4 (Table 2).

Synthesis of Results

Sample sizes ranged from 16 to 30, representing a total of 74 participants. All studies aimed to evaluate the effects of yoga on CPP. The study by Goncalves et al4 was the only study with a population exclusively of women with endometriosis. All of the studies reported that participants may have used another intervention in addition to yoga. However, only Huang et al6 specified that participants had not recently started an intervention that could influence the results.

All of the studies included an intervention supervised by a yoga teacher ranging from 6 to 8 weeks in duration, but only Huang et al6 specified the style of yoga (Iyengar). Frequency of the sessions ranged from 2 to 5 times a week and lasted between 1 and 2 hours. All studies included breathing (pranayama), relaxation, and physical postures (asana). The studies by Goncalves et al4 and Saxena et al22 also reported the use of chanting. The sequence of yoga postures (asanas) varied, and only 3 postures were shared across all studies, including viparita karani (legs up the wall), baddha konasana (butterfly or bound angle), and savasana (corpse pose). All studies reported statistically significant improvements in pain and QOL measures following the yoga intervention.

Data Analysis

Within-group analysis included the intervention group from all 3 studies, while the between-group analysis included only the studies by Goncalves et al4 and the Saxena et al.22 Forest plots for all analyses are depicted in Figures 2 to 5, and the individual study data and ESs are presented in Table 3.

Figure 2.
Figure 2.:
Forest plot of effect sizes and 95% confidence intervals for within-group analysis of quality of life.
Figure 3.
Figure 3.:
Forest plot of effect sizes and 95% confidence intervals for between-group analysis of quality of life.
Figure 4.
Figure 4.:
Forest plot of effect sizes and 95% confidence intervals for within-group analysis of pain.
Figure 5.
Figure 5.:
Forest plot of effect sizes and 95% confidence intervals for between-group analysis of pain.

Quality of Life

The QOL data were extracted either from the psychological or emotional well-being domain of the measure utilized by each study. Two studies4,6 used negatively scaled outcome measures, meaning a lower score indicates a better outcome, while 1 study22 used a positively scaled measure, meaning a higher score indicates a better outcome. During data analysis, the means extracted from that study were given a negative sign for consistency.

Statistically significant improvements in QOL were seen for both within-group (Table 3; ES =−1.4, CI: −1.8 to −1.1) and between-group (Table 4; ES =−1.5, CI: −2.0 to −1.0) meta-analyses. The SD from the World Health Organization Quality of Life Brief Version (WHOQOL-BREF) in the study by Saxena et al22 was used to convert back to clinical units, as the alternative outcome measure was specific to endometriosis. For within-group analysis, the grand ES corresponded to an expected increase of 17.3 points in the psychological domain of the WHOQOL-BREF, and between-group analysis corresponded to an expected difference of 14.9 points between the control group and the intervention group.

Pain

Statistically significant improvements in pain were seen for within-group (Table 3; ES: −2.2, CI: −2.7 to −1.6) and between-group (Table 4; ES =−1.4, CI: −1.7 to −1.0) meta-analyses. For within-group meta-analysis, the SD from the study by Huang et al6 was used to convert the ES into clinically meaningful units for the Numeric Pain Rating Scale because of the prevalence of this measure in clinical practice. The ES translated to an expected decrease of 2.1 points on the Numeric Pain Rating Scale following intervention. The SD from the study by Saxena et al22 was used to calculate the difference between groups on the 100-mm visual analog scale. The ES for between-group analysis translated to a difference of 18.7 points on the visual analog scale favoring the yoga group.

Harm, Risk, and Cost

The study by Huang et al6 was the only study to report adverse events. Nine events were reported, all of which were determined to be unrelated to the yoga intervention. The study by Goncalves et al4 reported that 1 participant stopped the intervention because of medical reasons but did not state whether this was related to yoga. The study by Saxena et al22 reported no information regarding adverse events. Although 1 study6 reported that yoga postures have the potential to strain the pelvic floor musculature, none of the studies reported that this was observed. The other 2 studies4,22 did not discuss risk, and none of the studies discussed cost.

DISCUSSION

The purpose of this study was to systematically review the literature and synthesize the effect of yoga on pain and QOL in women with CPP. A literature search yielded 3 studies that met the eligibility criteria, ranging from levels of evidence 2b to 4. Meta-analyses were performed for the outcomes of pain and QOL. Within-group and between-group results were statistically significant, and, likely, clinically meaningful in favor of the yoga intervention.

Quality of Life

The ESs for within-group and between-group QOL were both large (ES > 0.8)23 and statistically significant. Effect sizes were converted back into clinical units from the psychological domain of the WHOQOL-BREF to get an expected improvement of 17.3 points for within-group effect and an expected difference of 14.9 favoring the yoga intervention for between-group effect. No minimal clinically important difference (MCID) has been established for the WHOQOL-BREF, but the within-group analysis translates to a 50% improvement and the between group analysis translates to a 66% difference between groups favoring the yoga group.

Pain

Regarding pain, the ESs for within group and between group were also both large (ES > 0.8) and statistically significant. The within-group ES was converted to clinical units for the Numeric Pain Rating Scale. The MCID for this outcome measure is 2 points,24 and the average change in women with CPP after the yoga interventions in these studies is 2.1 points, which also equates to a 33% improvement in pain. Between-group effects were evaluated using the 100-unit visual analog scale, as in the study by Saxena et al.22 A difference of 18.7 points was seen favoring the yoga intervention with an expected 49% difference in pain scores between groups. This 18.7-point difference also meets the MCID of 18 to 19 points as described by Hägg et al.25 Thus, the results of this study suggest that clinically meaningful changes in pain can be observed through the use of yoga.

Limitations and Strengths

In determining the strength of evidence, considerations include risk of study bias, consistency in ESs among studies, directness of comparisons, and the precision and magnitude of the ESs for the reported outcomes.

With regard to bias, only 2 of the studies had random assignment to groups and 1 had significant attrition in both groups. The results of the PEDro and STROBE checklists indicate that the included studies had a moderate risk of bias. According to Owens et al,26 risk of bias affects the strength of a body of evidence. Risk of bias within the included studies related primarily to lack of blinding. While the nature of the intervention means that a true double-blinded study would be difficult to achieve, at least the assessors could be blinded in future studies.

A notable strength of this review is the consistency of the effects across studies. Consistency, or similarity in direction and range of ESs, was high in this meta-analysis. Within-group ESs were statistically significant, consistently in the same direction, and were considered large (>0.8) for both pain and QOL. While only 2 studies were included in the between group analyses, both ESs indicated reductions in pain and improvements in QOL.

The findings regarding pain are consistent with 2 other reviews assessing yoga in low back pain populations by Chang et al17 and pregnancy-related low back and pelvic pain populations by Kinser et al.18 Both of these reviews found statistically significant improvements in pain, although neither review further investigated whether or not the findings were also clinically meaningful as has been done in the current review. The reviews by Kinser et al18 and Chang et al17 do not demonstrate the same consistency of results regarding QOL, which may be due to lack of consistency among outcome measures for QOL utilized in various studies as well as a lack of established MCIDs for the outcome measures utilized in previous studies.

Another important consideration in considering strength of evidence is the directness of the comparisons in the included studies. All 3 studies utilized yoga interventions in a group setting with other women with CPP, and the studies that had control groups did not utilize a group intervention. The group setting may have led to increased feelings of social support, which could have influenced patient reports of QOL and pain. The study by Goncalves et al4 did collect data on social support and did not find a statistically significant change after the yoga intervention. However, 30 minutes of conversation and interaction among participants was reported as part of the intervention. Although they did not describe the type of conversation involved, social interaction among participants may have influenced outcomes in this study. The other 2 studies did not examine the effects of social support. The findings of this meta-analysis may have been confounded by the effects of group interventions; future studies should use the same mode of intervention delivery for experimental and control groups.

Finally, the precision and magnitude of the ES should be considered when evaluating the strength of the evidence. Effect sizes of 0.2 are considered small, 0.5 are considered medium, and 0.8 are considered large. The magnitudes for the pooled ESs were large for within-group and between-group comparisons for QOL and pain. Furthermore, the CIs were narrow in this review, thus the ES estimate can be considered to be fairly precise. The precision around the ES estimates and the large magnitude of the ESs for QOL and pain were such that conclusions could be drawn about the benefit of treatment despite the small number of studies.

Two additional limitations warrant mention. First, some of the studies lacked reporting of important information that could influence clinical application of the findings. For example, Saxena et al22 made no mention of adverse events and failed to report reasons for attrition. Finally, all of the studies had small sample sizes (n = 16-30), and the combined sample in this review was only 74, which may limit the generalizability of the results. However, the small samples yielded statistically significant results and thus did not limit power or increase the risk of a type 2 error.

Implications

Based on this assessment of the strength of evidence, the results of this review provide moderate support for a yoga intervention for women with CPP, which is consistent with the results of yoga interventions with populations of low back pain and pregnancy-related low back and pelvic pain. However, the studies were heterogeneous without clear recommendations regarding the style of yoga, frequency, or duration, and do not provide evidence for the mechanism of the effects of yoga on CPP. Previous research has indicated that yoga helps modulate the sympathetic nervous system and the hypothalamic-pituitary-adrenal system by decreasing blood pressure, heart rate, cortisol, and cytokine levels.15 Future studies that collect appropriate pain, QOL, and biomarker data may help determine whether yoga actually addresses the sensitization of the nervous system in women with CPP as prior studies suggest.27

When planning a yoga intervention for their patients with CPP, physical therapists should consider the time required for the intervention, cost to the patient, and the style of yoga. The study by Saxena et al22 was the highest quality of the studies and recommended 1-hour sessions, 5 times a week, for 8 weeks. The lowest quality study was the study by Huang et al,6 which used 90-minute sessions 2 times a week. This evidence suggests that the practice of yoga for 1- to 2-hour sessions totaling 3 to 5 hours per week for 6 to 8 weeks may lead to improvements in pain and QOL. While it is possible that a shorter amount of time could also lead to benefit, there is no evidence supporting a shorter intervention at this time. Patients should be open to incorporating 3 to 5 hours of yoga practice per week.

Regarding cost, a physical therapist may instruct in yoga practices during therapy sessions if he or she has the knowledge to do so, recommend that a patient attend public yoga classes, which can range from $10 to $25 per class,3 or recommend that a patient could practice on her own. The cost of individual therapy sessions or even attending a public yoga class may be prohibitive for some women. However, a solitary practice removes specific feedback from an instructor as well as the social aspect of a group class, both of which could influence the outcomes. In addition, the yoga interventions in the included studies were all designed specifically for the target population, and outcomes may differ if women attend public classes.

Finally, the style of yoga must be considered. Huang et al6 specified the participants practiced Iyengar yoga, but the other 2 studies did not specify style. Iyengar lies between the 2 physical extremes of yoga, which range from passively holding poses for several minutes to quickly transitioning between poses. The various styles of yoga may have different effects on the musculoskeletal and nervous systems. For example, a passive practice may calm the nervous system but may not be biomechanically appropriate for a patient with CPP and hypermobility. Many styles of yoga also include a practice of mula bandha, which is often taught as a pelvic floor contraction,28 which may not always be appropriate for women with CPP. There may also be information that is left out even when the style of yoga is disclosed, such as the use of props, the emphasis on breathing exercises, and the use of hands-on adjustments or verbal cues. Much of this information was missing from the studies reviewed and may be difficult to reproduce. At the current stage of research, the evidence does not support the use of one style of yoga over another for women with CPP.

Three postures were shared across all 3 studies: including viparita karani (legs up the wall), baddha konasana (butterfly or bound angle), and savasana (corpse pose). Legs up the wall and corpse pose may be considered restorative poses that promote relaxation. Butterfly or bound angle pose has a number of variations that can be restorative or may be more active. In this pose, the practitioner is seated on the floor in spinal flexion with the lower extremities in external rotation and the soles of the feet together. While not possible to make a conclusive statement about the effect of these postures on the nervous system or the pelvic floor, it is possible to speculate. If these postures are intended for full-body relaxation they may influence the central nervous system in such a way to reduce nervous system sensitization, as well as increase parasympathetic activity, both of which could influence the perception of pain. Likewise, if the practitioner truly experiences full-body relaxation, this would include the pelvic floor musculature, which is often overactive in women with CPP. Finally, the lower extremity positioning may be important. Passive external rotation while in hip flexion may put the obturator internus muscle on slack, which may also influence the pelvic floor musculature. While there may have been other reasons these yoga interventions led to positive outcomes, physical therapists should consider including these 3 postures if they plan to incorporate a yoga intervention for their patients with CPP.

While this review suggests that yoga may decrease pain and improve psychological and emotional well-being domains of QOL in women with CPP, it does not provide information about the effects of yoga when combined with physical therapy. Most of the participants in the studies were not receiving a physical therapy intervention simultaneously. Further research should address the effects of a yoga practice in conjunction with physical therapy.

In addition, future studies should evaluate the effects of practice duration and frequency, practice setting (group/individual, clinic/home-based/studio-based), style of yoga, and the selection and sequence of yoga poses. A study that aims to determine the minimal amount of time needed for intervention success could help physical therapists determine whether to incorporate yoga in clinical practice or recommend it as a self-management tool. Studies that evaluate practice setting, such as home or yoga studio, as well as the selection and sequence of poses could further inform physical therapists about yoga as a treatment strategy.

CONCLUSION

The results of this systematic review provide moderate evidence to support the use of yoga to improve pain and psychological and emotional QOL domains in women with CPP. Despite low to moderate quality of evidence, and a limited number of studies, the effects of the yoga interventions were large and the findings are consistent with previous studies that have examined the effects of yoga in other populations. While yoga interventions may not resolve CPP, given the results of this review, physical therapists may want to utilize yoga as a tool to help their patients decrease their pain and improve their QOL.

ACKNOWLEDGMENTS

The authors thank John Schott, DPT; Andrew Kim, DPT; and Grace Wang, DPT, for their thoughtful insights when reviewing multiple drafts of this manuscript and for helping to make this review possible.

REFERENCES

1. Anghari A. Prevalence of chronic pelvic pain among women: an updated review. Pain Physician. 2014;17(2):E141–E147.
2. Bishop LA. Management of chronic pelvic pain. Clin Obstet Gynecol. 2017;60(3):524–530. doi:10.1097/GRF.0000000000000299.
3. International Pelvic Pain Society. Patient education brochure. November 2013. Available at: https://www.pelvicpain.org/.
4. Goncalves AV, Barros NF, Bahamondes L. The practice of Hatha Yoga for the treatment of pain associated with endometriosis. J Altern Complement Med. 2017;23(1):45–52. doi:10.1089/acm.2015.0343.
5. Vandyken C, Hilton S. The Puzzle of pelvic pain: a rehabilitation framework for balancing tissue dysfunction and central sensitization II: a review of treatment considerations. J Womens Health Phys Therap. 2012;36(1):44–54. doi:10.1097/JWH.0b013e31824e0ab4.
6. Huang AJ, Rowen TS, Abercrombie P, Subak LL, Schembri M. Development and feasibility of a group-based therapeutic yoga program for women with chronic pelvic pain. Pain Med. 2017;18(10):1–9. doi:10.1093/pm/pnw306.
7. Naliboff BD, Stephens AJ, Lai HH, et al. Clinical and psychosocial predictors of urological chronic pelvic pain symptom change in 1 year: a prospective study from the MAPP research network. J Urol. 2017;198(4):848–857. doi:10.1016/j.juro.2017.05.065.
8. Prather H, Spitznagle TM, Dugan SA. Recognizing and treating pelvic pain and pelvic floor dysfunction. Phys Med Rehabil Clin N Am. 2007;18(3):477–496. doi:10.1016/j.pmr.2007.06.004.
9. Cheong YC, Smotra G, Williams AC. Non-surgical interventions for the management of chronic pelvic pain. Cochrane Database Syst Rev. 2014;(3):CD008797. doi:10.1002/14651858.CD008797.pub2.
10. Lotze M, Moseley G. Theoretical considerations for chronic pain rehabilitation. Phys Ther. 2015;95(9):1316–1320. doi:10.2522/ptj.20140581.
11. Kosek E, Cohen M, Baron R, et al. Do we need a third mechanistic descriptor for chronic pain states? Pain. 2016;157(7):1382–1386. doi:10.1097/j.pain.0000000000000507.
12. Wren AA, Wright MA, Carson JW, Keefe FJ. Yoga for persistent pain: new findings and directions for an ancient practice yoga-based protocols for pain management. Pain. 2011;152(3):477–480. doi:10.1016/j.pain.2010.11.017.
13. Tuttle LJ, DeLozier ER, Harter KA, Johnson SA, Plotts CN, Swartz JL. The role of the obturator internus muscle in pelvic floor function. J Womens Heal Phys Ther. 2016;40(1):15–19. doi:10.1097/JWH.0000000000000043.
14. Talasz H, Kremser C, Kofler M, Kalchschmid E, Lechleitner M, Rudisch A. Phase-locked parallel movement of diaphragm and pelvic floor during breathing and coughing-a dynamic MRI investigation in healthy females. Int Urogynecol J. 2011;22(1):61–68. doi:10.1007/s00192-010-1240-z.
15. Pascoe MC, Thompson DR, Ski CF. Yoga, mindfulness-based stress reduction and stress-related physiological measures: a meta-analysis. Psychoneuroendocrinology. 2017;86:152–168. doi:10.1016/j.psyneuen.2017.08.008.
16. Pascoe MC, Bauer IE. A systematic review of randomised control trials on the effects of yoga on stress measures and mood. J Psychiatr Res. 2015;68:270–282. doi:10.1016/j.jpsychires.2015.07.013.
17. Chang DG, Holt JA, Sklar M, Groessl EJ. Yoga as a treatment for chronic low back pain: a systematic review of the literature. J Orthop Rheumatol. 2016;3(1):1–8. doi:http://dx.doi.org/10.13188/2334-2846.1000018.
18. Kinser PA, Pauli J, Jallo N, et al. Physical activity and yoga-based approaches for pregnancy-related low back and pelvic pain. J Obstet Gynecol Neonatal Nurs. 2017;46(3):334–346. doi:10.1016/j.jogn.2016.12.006.
19. Jewell D. Guide Evidence-Based to Physical Therapy Practice. Sudbury, MA: 2003.
20. Maher CG, Sherrington C, Herbert RD, Moseley AM, Elkins M. Reliability of the PEDro Scale for rating quality of randomized controlled trials. Phys Ther. 2003;83(8):713–721. doi:10.1093/ptj/83.8.713.
21. de Morton NA. The PEDro scale is a valid measure of the methodological quality of clinical trials: a demographic study. Aust J Physiother. 2009;55(2):129–133. doi:10.1016/S0004-9514(09)70043-1.
22. Saxena R, Gupta M, Shankar N, Jain S, Saxena A. Effects of yogic intervention on pain scores and quality of life in females with chronic pelvic pain. Int J Yoga. 2017;10(1):9–15. doi:10.4103/0973.
23. Sullivan GM, Feinn R. Using effect size—or why the P value is not enough. J Grad Med Educ. 2012;4(3):279–282. doi:10.4300/JGME-D-12-00156.1.
24. Salaffi F, Stancati A, Silvestri CA, Ciapetti A, Grassi W. Minimal clinically important changes in chronic musculoskeletal pain intensity measured on a numerical rating scale. Eur J Pain. 2004;8(4):283–291. doi:10.1016/j.ejpain.2003.09.004.
25. Hägg O, Fritzell P, Nordwall A. The clinical importance of changes in outcome scores after treatment for chronic low back pain. Eur spine J. 2003;12(1):12–20. doi:10.1007/s00586-002-0464-0.
26. Owens DK, Lohr KN, Atkins D, et al. Grading the strength of a body of evidence when comparing medical interventions 2009 Aug 5. In: Methods Guide for Effectiveness and Comparative Effectiveness Reviews [Internet]. Rockville, MD: Agency for Healthcare Research and Quality (US); 2008-. Available from: https://www-ncbi-nlm-nih-gov.ezproxy.cul.columbia.edu/books/NBK47091/.
27. Vandyken C, Mdt C, Acup C, Hilton S. Physical therapy in the treatment of central pain mechanisms for female sexual pain. Sex Med Rev. 2017;5(1):20–30. doi:10.1016/j.sxmr.2016.06.004.
28. Kiel D. Functional Anatomy of Yoga; A Guide for Practitioners and Teachers. Chichester, England: Lotus Publishing; 2014.
Keywords:

alternative medicine; endometriosis; rehabilitation; vulvodynia

© 2019 Section on Women's Health, American Physical Therapy Association.