Abstract
Objective:
To support the research agenda in yoga for health by comprehensively identifying systematic reviews of yoga for health outcomes and conducting a bibliometric analysis to describe their publication characteristics and topic coverage.
Methods:
We searched 7 databases (MEDLINE/PubMed, Embase, PsycINFO, CINAHL, AMED, the Cochrane Database of Systematic Reviews, and PROSPERO) from their inception to November 2019 and 1 database (INDMED) from inception to January 2017. Two authors independently screened each record for inclusion and one author extracted publication characteristics and topics of included reviews.
Results:
We retrieved 2710 records and included 322 systematic reviews. 157 reviews were exclusively on yoga, and 165 were on yoga as one of a larger class of interventions (e.g., exercise). Most reviews were published in 2012 or later (260/322; 81%). First/corresponding authors were from 32 different countries; three-quarters were from the USA, Germany, China, Australia, the UK or Canada (240/322; 75%). Reviews were most frequently published in speciality journals (161/322; 50%) complementary medicine journals (66/322; 20%) or systematic review journals (59/322; 18%). Almost all were present in MEDLINE (296/322; 92%). Reviews were most often funded by government or non-profits (134/322; 42%), unfunded (74/322; 23%), or not explicit about funding (111/322; 34%). Common health topics were psychiatric/cognitive (n=56), cancer (n=39) and musculoskeletal conditions (n=36). Multiple reviews covered similar topics, particularly depression/anxiety (n=18), breast cancer (n=21), and low back pain (n=16).
Conclusions:
Further research should explore the overall quality of reporting and conduct of systematic reviews of yoga, the direction and certainty of specific conclusions, and duplication or gaps in review coverage of topics.
Keywords: Yoga, Systematic Reviews, Bibliometrics
Introduction
Yoga is a set of physical, mental and spiritual practices which arose in India thousands of years ago, has evolved over time, and is currently practiced in a range of forms around the world. Yoga traditionally consists of eight components, called the limbs of yoga: rules of conduct (yama) morality (niyama), physical postures (asanas), breath control (pranayama), withdrawal of the senses (pratyahara), concentration (dharana), meditation or withdrawal of the mind (dhyana) and oneness of meditative awareness (samadhi).[1] Many current forms of yoga focus primarily on postures (asanas), usually with the addition of breath control (pranayama), and sometimes also incorporating elements of concentration or meditation, however this varies with different styles or schools of yoga that are practiced in different areas of the world and for different purposes.[2, 3]
Yoga has become increasingly popular outside India during the past fifty years. According to the 2017 National Health Interview Survey (NHIS), yoga was used in the past 12 months by 14.3% (35.2 million) of US adults.[4] Likewise, in Australia, rates of yoga practice during the past 12 months have been estimated to be as high as 12% [5], and in Germany the lifetime prevalence of yoga practice was recently estimated at 15.1%.[6] Although reported prevalence rates vary by country as well as the reporting time frame and the criteria for duration or frequency of yoga practice, reports that compare use over time consistently find increases.[7, 8] In accordance with the increase in yoga use, there has been a worldwide increase in the number of journal publications on yoga, with the annual yoga research output growing by 7.79% between 2008 and 2016.[9]
In the US, most people using yoga do so for general health and wellness; however, approximately 18% of those who practice yoga use yoga to treat a specific health condition.[10] Yoga is believed to have many therapeutic benefits, and may be used to relieve depression or anxiety, reduce pain and improve physical function, and enhance quality of life in a range of health conditions.[11] Many randomized controlled trials (RCTs) have been conducted to evaluate the potential benefits and harms of yoga for health conditions.[12] Systematic reviews of RCTs are considered to be the most complete and unbiased sources of evidence on the effectiveness of an intervention, and an overview of the systematic review evidence base for yoga as a therapeutic option was published in 2013.[13] However, given the increasing popularity of yoga practice and interest in utilizing yoga as a therapeutic treatment strategy, it is likely that the number of systematic reviews has increased in recent years. An updated identification and description of these reviews may be used to highlight the sources and characteristics of evidence for the potential role of yoga in health. One method to characterize a field of research is through bibliometric analysis, which collects data on numbers of scientific articles and publications, classified by date, countries, and other characteristics, to map the development and current state of an area of research.[14] Bibliometric analyses may also serve as a foundation for further exploration of the more detailed characteristics and findings of research in a field (e.g., reporting adequacy, research quality, or specific results). Our aim is therefore to use bibliometric analysis methods to comprehensively identify systematic reviews of randomized controlled trials of yoga for health outcomes, describe their publication characteristics, and explore their coverage of health topics.
Methods
We have prepared a Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [15] checklist to assist in navigating through the reported elements of this study (Appendix A). There is no published or registered protocol for this study.
Eligibility criteria
Types of studies
For the purposes of this study, we considered systematic reviews of RCTs of yoga to be studies that 1) sought to comprehensively identify relevant RCTs on a yoga-focused research question, 2) stated inclusion criteria for the studies, 3) searched at least one electronic database, and 4) reported the critical appraisal of the included studies. We did not require reviews to restrict the evidence to RCTs, and we did not require them to perform, or attempt to perform, a meta-analysis. We included both completed systematic reviews, and protocols for systematic reviews in which the inclusion criteria, plans for searching, and plans for critical appraisal were reported. We excluded overviews, umbrella reviews, scoping reviews, narrative reviews, and integrative reviews unless they clearly stated in the title or abstract that they were systematic reviews and they met the four criteria above. We did not restrict review eligibility by language, date, or publication type with the exception of reviews available only as abstracts, due to the limited information typically available from abstracts. When reviews had multiple associated records, we considered the record of the most recent journal publication to be the primary record.
Types of participants
Systematic reviews of studies of all types of participants were eligible. We did not restrict eligibility by participant age, sex, or health status.
Types of interventions and comparisons
Any review that stated the intervention of interest was ‘yoga’ was included, without restriction by style or specific component of yoga. We wished to take a broad approach to the use of yoga for health, and did not restrict our inclusion to reviews that used specific terms such as yoga ‘therapy’.[16] There was no restriction on the type of comparisons within the systematic reviews, and comparisons between yoga and any inactive or active intervention, including other types of yoga, were eligible for inclusion.
Systematic reviews of a class of interventions which could include yoga, such as exercise or mind-body therapies, were only eligible if either a) yoga was explicitly stated in the research question to be one of the interventions of interest, or b) the term ‘yoga’ was included in a reported search strategy. We included reviews based on the research question and search strategy, not on the retrieval of yoga trials, as the identification of trials does not necessarily mean that identification was systematic or intended to be comprehensive.
Types of outcomes
Systematic reviews were eligible if they stated that their goal was to assess the effects of yoga interventions upon a health condition or health-related outcome, including cognitive, emotional or physical factors that are associated with the risk, prevention, or treatment of any health state.
Searching and Screening
We searched MEDLINE through PubMed, Embase, PsycINFO, CINAHL, the Allied and Complementary Medicine Database (AMED), IndMED, the Cochrane Database of Systematic Reviews, and PROSPERO for records of ongoing or completed systematic reviews. The search strategy for MEDLINE/PubMed (see Table 1) was adapted for other databases. No date or language restrictions were applied. Databases were searched in October 2016, and searches were updated in January 2017 and November 2019. The final date for all searches was November 19, 2019, with the exception of IndMED which was last searched in January 2017 and unavailable in November 2019.
Table 1.
Search strategy for MEDLINE/PubMed.
| #1 Yoga[mh] #2 Yoga* OR Yogic OR Pranayam* OR Asana* #3 #1 OR #2 #4 systematic review* OR systematic overview OR metaanalysis OR meta-analysis OR systematic* review OR systematic[sb] OR meta-analysis[pt] #5 #3 AND #4 |
After deduplication, we uploaded records to Abstrackr, a freely available online screening tool (http://abstrackr.cebm.brown.edu)[17] and two authors (LSW and AV or KP) independently screened titles and abstracts in random order. The full text of all records considered potentially relevant by one or both authors was then obtained. Two authors (LSW and AV or EAP) independently assessed each full text for inclusion. Disagreement was resolved by consensus or consultation with a third author (KP or RL).
Data extraction and analysis
One author (LSW) extracted bibliographic data (date of publication, country of origin, journal of publication, MEDLINE inclusion, statement of funding) and data on the review question. This included the participants (age, sex, health status), intervention (type of yoga if the review was focused only on yoga, class of intervention if the review was of a broader set of interventions), and topic (treatment or prevention, type of medical condition). Data for all protocols and reviews were collected in a standardized Excel spreadsheet. All data were downloaded into Stata 10 and descriptive analyses of frequencies and means were carried out, as well as chi-square tests when appropriate (StataCorp. 2007. Stata Statistical Soft-ware: Release 10. College Station, TX: StataCorp LP.).
Results
A total of 2710 records were retrieved through databases searching. After excluding duplicate records, two authors independently screened the titles and abstracts of 1319 records and selected 981 for full text assessment. After full text screening, we excluded 646 records, primarily because they were not focused on yoga or were not systematic reviews (e.g., did not carry out a risk of bias or other quality assessment). We included 335 records for a total of 322 reviews and protocols. See Figure 1 for the search and selection process and Appendix B for citations of all included reviews and protocols.
Figure 1.
Flowchart of records through screening and selection process.
Yoga focus
Of the 322 studies, half carried out a review focused solely on yoga (n=157/322 (49%); 152 completed systematic reviews and 5 protocols), and half carried out a review in which yoga was considered as one of a broader class of therapies (n= 165/322 (51%); 152 completed systematic reviews and 13 protocols).
Almost all reviews focused solely on yoga did not specify a particular type or school of yoga in their research question (146/157; 93%). More than half of the reviews focused solely on yoga did not state any eligibility criteria for the characteristics of the yoga intervention (87/157; 55%). The most frequently mentioned criterion was that the yoga must have a physical component such as asanas (28/157; 18%). The remaining 42 reviews most often required that postures be one of two or more components of the yoga intervention (e.g., both asanas and meditation must be present), or that postures could be one of a number of components of the yoga intervention (e.g., asanas or meditation or breathwork could be present). Of the reviews focused solely on yoga, between 1 and 25 databases (mean (sd) 6.05 (0.29) and median (IQR) 5 (4, 7)) were searched and the number of yoga RCTs retrieved per review ranged from none to 44 (mean (sd) 10.3 (8.2) and median (IQR) 8 (5, 13)). About half of reviews conducted a meta-analysis of at least two trials (80/157; 51%), and a quarter of these proposed a subgroup analysis based on the duration, components, or style of the yoga intervention (18/80; 23%); most of these reviews were able to carry out some subgroup analysis (13/18; 72%).
The broader class of therapies including yoga was most often characterized as exercise (48/165; 29%) or mind-body (40/165; 24%) interventions. Other categories commonly found were complementary or integrative medicine, meditation or mindfulness, and non-pharmacological interventions, but there were more than 50 permutations of article title terms for intervention types, from ‘alternative’ and ‘balance training’ to ‘relaxation’ and ‘stretch’. See Figure 2 for a word cloud of the terms used in the titles of reviews and protocols not focused solely on yoga. The following results will consider both yoga-focused and broader reviews, unless otherwise stated.
Figure 2.
Words in titles of articles including yoga as part of a larger group of interventions
Characteristics of reviews (bibliometrics)
The earliest systematic review was published in 1997, but the vast majority (297/322 ; 92%) were published in 2011 or later. For systematic reviews focused solely on yoga, the earliest review was published in 2005. These dates do not reflect the earliest Cochrane reviews, as we included only the latest version of Cochrane reviews, which are regularly updated. For example, the Cochrane review of yoga for epilepsy published in 2017 is an update of the review originally published in protocol form in 1998 and as a completed systematic review in 1999. Reviews focused solely on yoga and reviews including yoga as one of a broader class of interventions appeared to increase at similar rates over time (Figure 3).
Figure 3.
Publication of yoga systematic reviews on health over time.
The first or corresponding authors of the 322 publications were based in 32 different countries. Seventy-five percent of publications originated from one of six countries: the USA, Germany, China, the UK, Australia, and Canada (see Figure 4 for a chart displaying all countries producing at least five reviews). Authors from the same six countries also accounted for 120/157 (76%) reviews that were focused solely on yoga. Approximately 2% of all reviews and 3% of reviews focused on yoga originated from India (n=6/322 and n=4/157 respectively).
Fig 4.
Countries producing yoga systematic reviews on health.
Many reviews (111/322; 34%) did not explicitly state the source of funding for the review. Among those that did, funding was most often from government sources (76/211; 36%) or there was an explicit statement that no funding had been received (72/211; 34%). Some reviews were funded by non-profit organizations (43/211; 20%) or by a combination of government and non-profits (16/211; 8%). The remainder were funded by an academic organization (not clearly non-profit), a for-profit organization, or a non-governmental organization (1 review each). When yoga-only and broader reviews were compared with regard to funding source (grouped as government funding, non-profit funding alone, no funding, or not stated) yoga-only reviews were more likely to receive non-profit funding alone while broader reviews were slightly more likely to receive some government funding (overall chi-square statistic = 10.63; p-value is .031, df=4) (Figure 5).
Fig 5.
Stated funding sources for yoga systematic reviews on health
The 322 protocols and reviews were published in 170 different journals. The five journals with the most publications were the Cochrane Database of Systematic Reviews (n = 46; 14%), Journal of Alternative and Complementary Medicine (n = 15; 5%), Evidence-based Complementary and Alternative Medicine (n = 12; 4%), Complementary Therapies in Medicine (n = 11; 3%), and Complementary Therapies in Clinical Practice (n = 8; 2%). Of the remaining journals, 1 published 7 articles, 2 published 5 articles, 2 published 4 articles, 10 published 3 articles, 25 published 2 articles, and 125 published 1 article. This means that more than half of the identified reviews (175/322; 54%) were found in journals publishing only one or two reviews.
Overall, 161 articles (50%) were published in medical specialty journals, 70 articles (22%) in complementary medicine journals (of which 4 were published in yoga journals), 59 articles (18%) in systematic review journals, and 19 articles (6%) in general medical journals. The remaining 13 articles were published in multidisciplinary journals (n = 8), health services journals (n = 2), an evidence-based medicine (EBM) journal (n = 1), an economics journal (n = 1), or as a PhD dissertation (n = 1). Protocols were almost exclusively published in systematic review journals (16/18; 89%). Reviews focused solely on yoga were more likely to be published in complementary medicine journals, and less likely to be published in systematic review journals (chi-square statistic 22.78, df = 4, p<.001)(See Figure 6).
Figure 6.
Types of journals publishing yoga systematic reviews on health
Although we identified reviews and protocols from multiple databases, checking of MEDLINE revealed that most publications (296/322; 92%) were available in MEDLINE. Completed reviews were more likely to be included in MEDLINE than were review protocols (286/304; 94% vs 10/18; 56%).
Population and health topics of reviews
Most protocols and reviews (230/322; 71%) either simply specified that included populations should be adults or did not specify any inclusion criteria other than by medical condition or health status. The remaining studies focused on yoga in women (48/322; 15%), older people (18/322; 6%), children or adolescents (18/322; 6%), or specified men, veterans, or workers (2 reviews each) or focused on healthcare workers or prisoners (1 review each). In women, the conditions examined were breast cancer (18/48; 38%), pregnancy (13/48; 275%), menstrual disorders (6/48; 13%), menopause (6/48; 13%), urinary incontinence (2/48; 4%) or multiple health outcomes pertaining to older women, metabolic syndrome, or sleep (1 review each). In older people, the conditions of interest were dementia or other cognitive outcomes (6/18), balance or falls (4/18), sleep (2/18), urinary incontinence or cardiovascular disease (1 review each) or looked broadly at multiple health outcomes (4/18). In children, the conditions examined were primarily psychological or cognitive (9/18), respiratory (3/18), or digestive (2/18), or looked broadly at multiple health outcomes (3/18). There was one review on yoga for cerebral palsy in children.
Most reviews (254/322; 79%) specified that they focused on yoga for the treatment of a health-related condition or symptom. Only a few reviews (7/322; 2%) stated that they were oriented towards prevention, while all remaining reviews (61/322; 19%) either had a mixed focus on both treatment and prevention or were unclear.
Among all protocols and reviews, 62% (200/322) focused on a single condition (e.g, breast cancer), 16% (52/322) focused on a group of conditions (e.g, any cancers), and 18% (59/322) focused on one or more symptoms, risk factors, or other processes not restricted to particular conditions (e.g, pain). The remaining reviews (13/322; 4%) focused broadly on a special population (e.g, yoga in pediatric populations).
Among the 252 reviews focused on a single condition or group of related conditions, the most common conditions were psychiatric or cognitive (n = 56), cancer (n = 39), and musculoskeletal (n = 36) (see Figure 7 for topics with 5 or more reviews). The most common psychiatric/cognitive conditions were depression/anxiety (n = 18), psychotic disorders (n = 11), and post-traumatic stress disorder (PTSD) (n = 8). The most common cancer condition was breast cancer (n = 21), and the most common musculoskeletal condition was back or neck pain (n = 23), of which low back pain (n = 16) was the most common topic.
Fig 7.
Most common topics in yoga systematic reviews on health
Discussion
We found 322 systematic reviews assessing yoga as a health intervention, far more than the 26 reviews on yoga for acute and chronic conditions found in 2013 by McCall et al.[13] The difference in numbers may partially be due to different inclusion criteria between this project and the earlier overview. For example, McCall et al. included only reviews of yoga to treat single conditions in unhealthy people. However, it is likely that our larger number of systematic reviews is primarily due to increases in publication rates over the last decade, since 92% of the reviews we included were published in 2011 or later.
Half of the reviews we found focused solely on yoga and half included yoga with other interventions or as one of a larger class of interventions. This is consistent with the McCall overview, in which 14/26 reviews included therapeutic yoga interventions in combination with other interventions.[13] When we looked at classes of interventions, we observed that about a quarter of reviews with this broader focus characterized yoga as a physical exercise intervention (a solely body-based orientation to yoga) and about a quarter of reviews characterized yoga as a mind-body intervention. This inconsistent categorization of yoga poses difficulties in finding systematic review evidence on yoga. For example, yoga is included in some reviews on exercise therapies, but not all exercise reviews explicitly include or exclude yoga. Overall, not only those searching for review evidence on yoga but also those producing review evidence on related interventions (eg, exercise) face uncertainty about how to identify and classify yoga. This unclear conceptualization is also reflected in the current approach to yoga as an exercise, as a meditation practice, or both, depending on the teacher and setting.
Some of this difficulty in categorizing yoga in research and policy may be ameliorated with standardized terminology across fields to differentiate therapeutic yoga interventions and yoga therapy. We observed that most included reviews were explicitly about yoga as a treatment of a health condition, most commonly mental health or musculoskeletal. This reflects the research agenda on yoga as a treatment, in contrast with the population use of yoga for general wellness.[10, 18] The growing view of yoga used for health reasons as ‘therapeutic yoga’ rather than simply ‘yoga’, and the refinement of distinctions between therapeutic yoga and yoga therapy, reflects the increasing focus in the clinical and research community on yoga as a management strategy for chronic diseases or conditions, and the development of different requirements for the different approaches to the use of yoga in clinical settings.[19-29] In the future, while yoga therapy may still be classed as a type of exercise therapy, relaxation therapy, or other type of therapy, reviews on these classes of intervention may be more likely to highlight the inclusion or exclusion of yoga, making these broader reviews covering yoga easier to identify in the literature.
The overall heterogeneity of yoga practice poses a challenge for the conduct and applicability of yoga research, and we identify several areas that could be improved. Because the term yoga covers a range of interventions, clarifying the scope of the yoga intervention is particularly important, yet we observed that over half of the reviews focused solely on yoga did not state any eligibility criteria for the characteristics of the yoga intervention. Furthermore, the appropriateness of combining different yoga interventions in meta-analyses relies upon decisions that the yoga interventions are sufficiently clinically similar. While almost one-fifth of yoga-focused reviews conducting meta-analyses mentioned a planned subgroup analysis based on some characteristic of the yoga intervention, the rationale for combining evidence overall and by subgroup needs to be made explicit and based upon a theoretical and clinical rationale.
Overall, half of all yoga-related reviews were published in medical specialty journals, which each published only one or two articles. An additional 22% were published in complementary medicine journals and 18% were published in systematic review journals. The five journals with the greatest number of reviews were the Cochrane Database of Systematic Reviews and four complementary medicine journals. We observed that reviews focused only on yoga, compared to reviews covering broader interventions, were more frequently found in complementary journals and less frequently found in systematic review journals. Nevertheless, almost half of yoga-focused reviews were published in specialty journals (74/157; 47%). Given the dispersion of yoga systematic reviews, a central repository of yoga research might be useful to researchers and others seeking access to evidence on yoga. A central repository of yoga research could be a database of yoga trials or yoga reviews, or both, that could be held by a research organization (such as the Cochrane Complementary Medicine Field) or a yoga organization (such as a journal or a practice organization). In the meantime, the list of included protocols and reviews reported as Appendix B is a step in this direction.
Authors of reviews were based in 32 different countries, with 75% of reviews originating from authors in the USA, Germany, China, the UK, Australia, or Canada. In contrast with the estimated 46% of yoga RCTs originating from India [12] we found less than 5% of yoga systematic reviews originated from India, and this was true both for reviews of broader topics and for reviews restricted to yoga. This also contrasts with the observation that India is the first [30] or second [9] most productive source of published research on yoga. It is not clear why there is this discordance between the percentage of systematic reviews and the percentage of other research literature on yoga originating from India. Although all of the reviews from India were published in 2015 or later, we did not find enough reviews overall to be able to estimate whether numbers might be increasing in more recent years.
Only 65% of reviews explicitly stated the source of any funding for the systematic review, and one-third of these stated that there was no funding. The most common source of funding was government, followed by non-profits. Overall, the yoga-only reviews were more likely to be funded by non-profits while reviews on broader topics were more likely to be funded by governments. Difference in source of funding for yoga-only and broader focus reviews may be a chance finding or may be due to government funders wishing to address broad questions that have a scope beyond yoga, while non-profits have a more focused interest on yoga. We note that most countries do not have funding schemes for research into yoga or other complementary therapies.
When looking more closely at the topics of individual reviews, we observed that the most common medical problems addressed were psychiatric (primarily depression and anxiety), musculoskeletal (most often low back pain) and cancer (primarily breast cancer) followed closely by metabolic problems (e.g., diabetes) and women’s health (e.g., pregnancy or menopause). Multiple reviews on the same or similar conditions require further examination. While the number of yoga reviews is not as great as the number of yoga RCTs (e.g., over 2,400 controlled trials including yoga in the title are listed in the Cochrane Central Register of Controlled Trials (CENTRAL) as of April 2021), for some topics the number of reviews may approach or exceed the number of RCTs. For example, we found a total of 16 reviews on low back pain. We extracted the number of RCTs found among the six reviews that were focused solely on yoga for low back pain and found that they contained between 4 and 14 RCTs each. The number of RCTs per review is thus exceeded by the overall number of reviews on low back pain. When there are multiple reviews on a topic, researchers should clarify to what extent the research questions in these reviews overlap, or they simply duplicate earlier reviews, indicating research waste. Assessing whether more recent reviews incorporate new evidence that is of higher quality would also be important. It has been shown that although risk of bias is often high in yoga RCTs, newer studies have lower risk of bias and are better reported[31], perhaps due to the presence of reporting guidelines such as the CONSORT statement.
Further research into systematic reviews of yoga should consider several additional areas that impact the usefulness and applicability of the research. For example, there is a need to explore how reviews address the reporting of yoga interventions. In addition to the specific components, sequencing, and modifications of the yoga intervention, elements such as the training of the person delivering the yoga intervention are likely important in understanding the effects of the intervention and how it might be transferred to practice. Systematic reviews should explore how the included randomized controlled trials report the yoga interventions, incorporate this detail, when available, into review conclusions, and promote better reporting in future trials. Similarly, there is a need to explore how reviews address the reporting of adverse events. Many randomized controlled trials do not report on adverse events or they do so inconsistently. Systematic reviews should consider potential harms, as well as potential benefits, in summarizing the effects of yoga interventions, and should highlight when trials fail to provide the necessary information. The characteristics of the yoga intervention and the potential for adverse events are key characteristics that may determine effectiveness for particular health conditions. Inclusion of these considerations in the methods, results, and conclusions of reviews would make systematic reviews more useful in supporting clinical guidance on the use of yoga. Finally, further examination of the reporting (e.g., PRISMA compliance), quality (e.g., AMSTAR ratings), or risk of bias (e.g., ROBIS assessment) of systematic reviews is required, both for yoga systematic reviews overall and within specific topic areas, as better reporting and quality might correspond to a more trustworthy and relevant estimation of efficacy and safety.
Limitations
There are some limitations to the current project. First of all, because search strategies were developed to identify systematic reviews of yoga it is likely that systematic reviews focused solely on yoga were comprehensively identified, but possible that some systematic reviews including yoga as one of a class of interventions (e.g., exercise) but not possessing text terms or indexing related to yoga were not retrieved. We cannot know in what ways other than indexing and title words those systematic reviews might differ from the reviews that were retrieved by our searches. Another limitation of our searches was that we did not specifically search non-English-language databases or seek out non-English-language studies. We did not have resources for translation and, while we were able to assess the eligibility of some papers in languages other than English, we were unable to assess six studies in Chinese, Dutch or Spanish and these, therefore, were not included in the review. Therefore, there may be additional non-English-language systematic reviews that are missing from this analysis.
Second, while pairs of authors independently assessed records for inclusion against explicit criteria defining systematic reviews of randomized controlled trials, yoga, and health outcomes only a single author extracted information from the included reviews. Data were not checked by a second author and it is therefore possible that there are some errors in interpretation and extraction that might have been avoided by dual data extraction or checking.
Conclusions
Research on the health effects of yoga is a growing area in systematic reviews. We have outlined the breadth of systematic review evidence on this topic and provided insight into the characteristics of this evidence and where it may be found as well as potential research gaps and open research questions. We consider this research to be a foundation for further research into systematic reviews of yoga. Further examination of this evidence, including gaps, duplication of effort, quality, and consensus in findings, may advance the use of systematic reviews to develop and clarify the evidence base for yoga to improve health.
Supplementary Material
Highlights.
Systematic reviews of yoga for health have increased sharply since 2011
Half of systematic reviews of yoga for health also cover broader interventions (e.g., exercise)
Most systematic reviews of yoga for health are published in specialty journals and indexed in MEDLINE
Systematic reviews of yoga for health commonly cover cancer, mental health, or musculoskeletal conditions
In some areas, there are multiple systematic reviews on the same or similar topics (e.g., low back pain)
Funding:
LSW is supported by the National Center for Complementary and Integrative Health of the National Institutes of Health (R24 AT001293). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Competing interests: None declared.
Data statement: This is a secondary analysis of publicly available literature. Data sharing is not applicable as no datasets were generated and/or analyzed for this study.
References
- 1.Sarbacker S, Kimple K. The Eight Limbs of Yoga: A Handbook for Living Yoga Philosophy: North Point Press; 2015. [Google Scholar]
- 2.DeMichelis E A History of Modern Yoga. Patanjali and Western Esotericism. London: Continuum; 2005. [Google Scholar]
- 3.Singleton M Yoga Body. The Origins of Modern Posture Practice. . Oxford: Oxford University Press; 2010. [Google Scholar]
- 4.Clarke T, Barnes P, Black L, Stussman B, Nahin R. Use of yoga, meditation, and chiropractors among U.S. adults aged 18 and over. NCHS data brief, no 325. Hyattsville, MD: National Center for Health Statistics. November 2018. [PubMed] [Google Scholar]
- 5.Xue CC, Zhang AL, Lin V, Da Costa C, Story DF. Complementary and alternative medicine use in Australia: a national population-based survey. Journal of alternative and complementary medicine (New York, NY). 2007;13(6):643–50. [DOI] [PubMed] [Google Scholar]
- 6.Cramer H [Yoga in Germany - Results of a Nationally Representative Survey]. Forschende Komplementarmedizin (2006). 2015;22(5):304–10. [DOI] [PubMed] [Google Scholar]
- 7.Ding D, Stamatakis E. Yoga practice in England 1997-2008: prevalence, temporal trends, and correlates of participation. BMC research notes. 2014;7:172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Clarke TC, Black LI, Stussman BJ, Barnes PM, Nahin RL. Trends in the use of complementary health approaches among adults: United States, 2002-2012. National health statistics reports. 2015(79):1–16. [PMC free article] [PubMed] [Google Scholar]
- 9.Gupta B, Mueen Ahmed K, Dhawan S, Gupta R. Yoga Research a Scientometric Assessment of Global Publications Output during 2007-16. Parmacogn J. 2018;10(3):394–402. [Google Scholar]
- 10.Stussman BJ, Black LI, Barnes PM, Clarke TC, Nahin RL. Wellness-related Use of Common Complementary Health Approaches Among Adults: United States, 2012. National health statistics reports. 2015(85):1–12. [PubMed] [Google Scholar]
- 11.Singh Khalsa S, Cohen L, McCall T, Telles S, editors. The Principles and Practice of Yoga in Health Care: Handspring Publishing Limited; 2016. [Google Scholar]
- 12.Cramer H, Lauche R, Dobos G. Characteristics of randomized controlled trials of yoga: a bibliometric analysis. BMC Complement Altern Med. 2014;14:328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.McCall MC, Ward A, Roberts NW, Heneghan C. Overview of systematic reviews: yoga as a therapeutic intervention for adults with acute and chronic health conditions. Evid Based Complement Alternat Med. 2013;2013:945895. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.OECD. Glossary of Statistical Terms: Bibliometrics. https://statsoecdorg/glossary/detailasp?ID=198 [accessed 12 March 2021].
- 15.Moher D, Liberati A, Tetzlaff J, Altman DG, Group P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kraftsow G The distinction between a yoga class and a yoga therapy session. International Journal of Yoga Therapy. 2014;24:17–8. [PubMed] [Google Scholar]
- 17.Wallace BC, Small K, Brodley CE, Lau J, Trikalinos TA. Deploying an interactive machine learning system in an evidence-based practice center: abstrackr. Proceedings of the ACM International Health Informatics Symposium. 2012:819–24. [Google Scholar]
- 18.Cramer H, Ward L, Steel A, Lauche R, Dobos G, Zhang Y. Prevalence, Patterns, and Predictors of Yoga Use: Results of a U.S. Nationally Representative Survey. Am J Prev Med. 2016;50(2):230–5. [DOI] [PubMed] [Google Scholar]
- 19.Cramer H, Lauche R. Yoga therapy: Efficacy, mechanisms and implementation. Complement Ther Med. 2018;40:236. [DOI] [PubMed] [Google Scholar]
- 20.Sullivan MB, Moonaz S, Weber K, Taylor JN, Schmalzl L. Toward an Explanatory Framework for Yoga Therapy Informed by Philosophical and Ethical Perspectives. Altern Ther Health Med. 2018;24(1):38–47. [PubMed] [Google Scholar]
- 21.Taylor MJ, McCall T. Implementation of Yoga Therapy into U.S. Healthcare Systems. Int J Yoga Therap. 2017;27(1):115–9. [DOI] [PubMed] [Google Scholar]
- 22.Ross A, Williams L, Pappas-Sandonas M, Touchton-Leonard K, Fogel D. Incorporating Yoga Therapy into Primary Care: The Casey Health Institute. Int J Yoga Therap. 2015;25(1):43–9. [DOI] [PubMed] [Google Scholar]
- 23.Rioux JG. Yoga Therapy Research: A Whole-Systems Perspective on Comparative Effectiveness and Patient-Centered Outcomes. Int J Yoga Therap. 2015;25(1):9–19. [DOI] [PubMed] [Google Scholar]
- 24.Pearson N, Prosko S, Sullivan M, Taylor MJ. White Paper: Yoga Therapy and Pain-How Yoga Therapy Serves in Comprehensive Integrative Pain Management, and How It Can Do More. Int J Yoga Therap. 2020;30(1):117–33. [DOI] [PubMed] [Google Scholar]
- 25.Stephens I Medical Yoga Therapy. Children (Basel, Switzerland). 2017;4(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Noggle JJ. Scientific Svadhyaya: developing guidelines for yoga therapy research and practice. Int J Yoga Therap. 2013(23):7–8. [PubMed] [Google Scholar]
- 27.Bar J Bringing yoga therapy into mainstream health care: lessons from the Cleveland Clinic and their relationship to emotional well-being. Int J Yoga Therap. 2013(23):67. [PubMed] [Google Scholar]
- 28.Kimura K Yoga Therapy in Japan. Int J Yoga Therap. 2017;27(1):127–9. [DOI] [PubMed] [Google Scholar]
- 29.Cramer H Yoga Therapy in the German Healthcare System. Int J Yoga Therap. 2018;28(1):133–5. [DOI] [PubMed] [Google Scholar]
- 30.Jeter PE, Slutsky J, Singh N, Khalsa SB. Yoga as a Therapeutic Intervention: A Bibliometric Analysis of Published Research Studies from 1967 to 2013. Journal of alternative and complementary medicine (New York, NY). 2015;21(10):586–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Cramer H, Langhorst J, Dobos G, Lauche R. Associated Factors and Consequences of Risk of Bias in Randomized Controlled Trials of Yoga: A Systematic Review. PLoS One. 2015;10(12):e0144125. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.