Concern about yoga-related injuries seems to be at an all-time high at the moment, with articles in the mainstream media pointing toward increasing numbers of injuries resulting from the practice, systems of yoga arising that claim to teach a “safer & more sustainable” style of practice, and widespread discussions among yoga teachers about the injurious nature of yoga and how best to keep students safe on the mat.

Much of this discussion is rooted in anecdotal evidence, however. As someone with a great interest in helping to bring a solid grounding in science to the practice of yoga, I am thrilled to present a special guest post today that will add to the greater dialog about yoga injuries with a look at the actual scientific evidence on this important topic.

Jari Karppinen, PT, MSc, and PhD student in Sport & Exercise Medicine, provides us with an up-to-date review of the scientific research on yoga injuries. This piece is essential reading for any yoga practitioner interested in an evidence-based perspective on yoga injuries, as well as anyone interested in improving their scientific literacy skills in general.

I hope you appreciate this high-quality content offered by Jari. Be sure to read the conclusion at the end of the post because it contains some major takeaways for our yoga community! And consider following Jari and his work on Instagram at @jarikarppinen.

Thank you for reading! --Jenni

Jari Karppinen, PT, MSc, PhD student in Sport & Exercise Medicine

Introduction

Exercise, despite all its benefits, possesses also risks – from sudden cardiac death to minor strains and sprains. Naturally, asana-based yoga is no exception. The topic is widely discussed, and most yoga professionals have their own opinion on the matter; however, these opinions are often based on anecdotal evidence.

When estimating the potential risks of yoga practice and deciding whether special injury prevention strategies are needed, we first need information about 1) how common adverse events are and 2) why do they happen, including information about mechanisms and possible risk and preventative factors [1]. This blog post is intended to provide an up to date review of scientific evidence concerning yoga-related adverse events to aid constructive discussion.

Basic concepts

Before diving into the results, some basic concepts need to be addressed. I want to highlight that study design has a huge impact on the type information a study can provide and how valid that information is [2]. Some of the benefits and limitations of each study design are discussed later. Also, the generalizability of the findings is not always straightforward as yoga is a heterogeneous discipline practiced by different demographics. For example, if the study participants practice ashtanga, it is far-reaching to extrapolate the findings directly to kundalini or yin yoga practitioners. Similarly, participants with a medical condition might respond to yoga differently when compared with healthy participants.

“Study design has a huge impact on the type information a study can provide and how valid that information is.”

A good thing to also keep in mind is that sports injury classification is notoriously problematic [3], and the outcome measures vary between studies. Thus, in this blog post, injuries are just called “adverse events”. In yoga studies, adverse events are typically any aches or pains associated with yoga practice reported by the participant. In more rigorous designs, medical confirmation may be required. Studies often also classify the adverse events based on the need for medical treatment, whether the event caused loss-of-time from practice and how long it lasted. Naturally, the classification system used greatly influences the study’s findings.

Evidence from case studies

When investigating exercise-related adverse events, case studies are the lowest form of evidence. Case studies are typically medically confirmed rare events, which are then reported in scientific journals. Case studies may provide mechanistic evidence and aid in building research hypotheses that can be then investigated with more rigorous study designs. However, they provide only low evidence as they lack repeatability and generalizability.

“When investigating exercise-related adverse events, case studies are the lowest form of evidence.”

For example, two very interesting case studies exist where HIV infection was cured with an allogeneic stem cell transplantation [4,5]. Still, it is obvious that this high-risk treatment cannot be routinely used to treat HIV. Another limitation of case studies is the heightened risk of researcher bias – it is very possible that the researcher's own ideas and opinions influence how he or she interprets and reports the case.

In 2013, Cramer et al. provided a thorough synthesis of case studies concerning yoga-related adverse events [6]. In total, they identified 77 cases of which 27 affected the musculoskeletal system, 14 the nervous system and 9 the eyes. Over half of the subjects reached full recovery after the incident. One case of death was reported, which resulted from air embolism after voluntary mouth-to-mouth breathing exercise. Typical poses or practices associated with the adverse events were headstand (10 cases) and shoulder stand (3 cases), which mostly led to glaucomatous symptoms. Also, forceful breathing exercises (3 cases) and poses requiring extreme flexibility (3) were reported as causes of adverse events.

The case studies often raise a lot of interest as we as humans are drawn into stories and anecdotes. However, I want to highlight that these are rare occurrences and they do not provide generalizable evidence.

Evidence from observation studies

Cross-sectional studies

When searching for evidence on how common yoga-related adverse events are, we need to look beyond case studies. The research largely focused on cross-sectional studies, which is understandable as they are relatively easy to conduct compared to longitudinal setups. The cross-sectional design means that the researchers collect data only at one point in time and in the case of yoga studies, usually with surveys. Cross-sectional studies can give information about prevalence (e.g., how a large proportion of the study participants currently have or have experienced a yoga-related adverse event). Some studies also estimate rates and risks; however, these should be treated with caution as longitudinal designs are better suited for this.

Cross-sectional studies have some limitations. First, the risk of inaccurate data is relatively high as participants are asked to remember past events. Second, the studies commonly suffer from selection bias, which means that the actual study sample is not representative of the accessible study population. For example, participants who have experienced an adverse event may be more motivated to take part in a study investigating the topic. Third, cross-sectional studies are influenced by a myriad of confounding factors. Thus, any association found by the study should not be treated as causal.

“ To no surprise, prevalence seemed to be higher among participants with more frequent and intense practice habits, such as among Finnish Ashtanga practitioners [8]”

A good source to dive into the topic is a systematic review by Cramer at al. (2018) where the researchers gathered together all observational evidence on yoga-related adverse events published before October 2016 [7]. In total, they located nine cross-sectional studies, which included 9,129 yoga practitioners from all over the world. The number of yoga practitioners in individual studies ranged from 86 to 2,567. The overall methodological quality was mixed and generally on the weaker side. For example, only two studies included an unbiased representative population sample.

Full synthesis of the studies was impossible as the included studies varied in study populations and outcome measures. Thus, the results presented in the review and here are based only on some of the studies. The lifetime prevalence of adverse events ranged from 21.3 to 61.8% [8–12]. To no surprise, prevalence seemed to be higher among participants with more frequent and intense practice habits, such as among Finnish Ashtanga practitioners [8]. Still, the estimated adverse event risk in the population was low (1.45 events per 1,000 hrs of practice) [8]. More serious events (e.g. requiring discontinuation of a yoga class or resulting in discontinued yoga use) were rare as their lifetime prevalence ranged from 0 to 2 %. The most common adverse events reported were sprains and strains. Poses associated with adverse events were head-, shoulder- and handstand, lotus pose and forward- or backward bends [9].

“Lauche et al. showed that regular yoga practice was not associated with joint issues among 9,103 upper middle-aged Australian women [13].”

Since data collection of Cramer et al. (2018) ended, four cross-sectional studies, that deserve a mention, have been published. In their study published in 2017, Lauche et al. showed that regular yoga practice was not associated with joint issues among 9,103 upper middle-aged Australian women [13]. An interesting comparison between exercise disciplines was made possible by Bueno et al. (2018) who investigated past 12-month prevalence of sports injuries among 3,498 Danish adults [14]. In this study, adverse events were classified as events that prevented exercise participation for at least seven days, and/or which required contact with a health professional. In total, 303 participants reported regular yoga practice and only two of them had experienced a yoga-associated adverse event. Thus, the injury prevalence of yoga was less than 1% when compared to e.g. 38% among soccer players, 19% among runners and 9% among subjects participating in strength training.

“As many other exercise disciplines have higher injury rates... both research groups concluded that yoga seems to be a low-risk exercise and at least as safe or even safer than other exercise types.”

During fall 2019, two new large cross-sectional surveys were published. In a study by Cramer et al. (2019) lifetime prevalence of yoga-related acute or chronic adverse events among 1,702 German yoga practitioners was 21.4% and 10.2%, respectively [15]. In total, 0.6 injuries were estimated to happen for every 1,000 hrs of yoga practice. The lifetime prevalence of at least one yoga-related adverse event was 65% in a study by Wiese et al. (2019) whose 2,630 respondents were mainly from North America or Europe [16]. As many other exercise disciplines have higher injury rates (e.g. 2.5 injuries per 1,000 hrs of general cardiovascular exercise), both research groups concluded that yoga seems to be a low-risk exercise and at least as safe or even safer than other exercise types. The studies also concluded that the adverse events were mostly mild sprains or strains.

In Cramer et al., 16 serious events were reported (e.g. one case of cerebral hemorrhage, fractures, spinal or nerve injuries) and in Wiese et al. (2019), 4% of participants reported a more severe event (requiring a medical procedure or the injury persisted over a year). In Cramer at al. (2019), events were more common in participants with a pre-existing medical condition and among participants who did yoga on their own without prior or current supervision by a teacher. Of different yoga styles, Power yoga was associated with the highest and Viniyoga with the lowest adverse event risk (1.5 and 0.54 per 1,000 hrs of yoga practice, respectively). Interestingly, the practice of yoga philosophy and the use of props were associated with slightly higher acute adverse event risk. However, this association can be probably explained with confounding factors. In Wiese et al. (2019), the strongest predictor of adverse events was years spent practicing yoga. Also, yoga teacher status was associated with a slightly increased risk. Somewhat contrary to Cramer et al. (2019), adverse events were more commonly reported to happen during a guided class than during self-practice. The poses most often associated with adverse events were hand-, shoulder- and headstands and the injured body locations mostly reported were knees, low back, shoulders and hips.

Longitudinal studies

Prospective cohort studies are the most informative studies when assessing exercise-related adverse events. In the design, a group of participants (i.e. a cohort) is recruited and followed for a certain period. Thus, the design can provide information about injury incidence (rate of new injury cases during the follow-up) which is much more informative than prevalence. Still, like cross-sectional studies, cohort studies are influenced by confounding factors and it is difficult to fully prove causality with them.

“As the number of new pain cases was low, the study lacked the power to find any associations between pain cases and potential risk factors.”

The study by Campo et al. (2018) is, to my knowledge, the only prospective cohort study concerning yoga-related adverse events [17]. The cohort was recruited via a U.S. yoga studio, with a wide offering of different yoga styles. Data were collected with a questionnaire in the beginning and after a one-year follow-up. The participants were asked about the pain they experienced in the past 12 months and if yoga caused, exacerbated, unaffected or improved the pain. Participants also provided descriptive information about the pain experience.

In total, 354 participants completed the study. On average, the participants were middle-aged women with a regular yoga practice. During the follow-up, 38 of the participants (10.7%) experienced at least one new case of pain caused by yoga. As the researchers collected data only twice, it is possible that minor pains were forgotten during the follow-up. The reported pain cases were mild to moderate and situated most often in the wrist or hand area. Poses associated with pain ranged widely; the most common reported pose was Down Dog (3 cases). Fifteen pain cases resulted in lost participation time and 16 lasted longer than three months. As the number of new pain cases was low, the study lacked the power to find any associations between pain cases and potential risk factors. Thus, in the future, a much larger cohort and/or a longer follow-up are needed to investigate the question, as adverse event incidence in yoga is low.

Evidence from experimental studies

Of experimental studies, randomized controlled trials (RCTs) in particular can give detailed information about exercise-related adverse events as they are carefully monitored, and the events can be easily medically confirmed. However, the primary aim of RCTs is to investigate intervention effects and thus adverse effects are often poorly reported [18]. Also, limited sample sizes and short follow-ups usually lead to a low number of adverse events, which limits the use of RCTs in injury research. Furthermore, carefully supervised interventions may not reflect how yoga is practiced in “real life”.

Cramer et al. (2015) conducted a meta-analysis of the safety of yoga in RCTs where they identified 301 RCTs published before February 2014 of which 94 RCTs reported adverse events [18]. Thus, the quality of adverse event reporting was considered poor. The researchers classified adverse events as serious (required medical intervention) or non-serious (all other events) and as intervention-related only if the authors reported so. Also, dropouts due to injuries were collected.

“Pooled results showed that odds of adverse events or dropouts due to these events were similar between yoga and no treatment or usual care, and between yoga and exercise interventions.”

The included studies were conducted all over the world, in heterogeneous populations and with varying interventions. Yoga was compared with no treatment or usual care in 53 studies, with exercise in 26 studies and with psychological or educational interventions in 22 studies. The median intervention duration was 10 weeks. In total, 2.2% of the subjects participating in yoga interventions reported an intervention-related adverse event. Non-serious and serious adverse events (not necessarily intervention-related) were reported by 10.9 and 0.6% of the yoga intervention participants. Pooled results showed that odds of adverse events or dropouts due to these events were similar between yoga and no treatment or usual care, and between yoga and exercise interventions. The odds of non-serious adverse events were higher when yoga was compared with psychological or educational interventions; however, there were no differences in serious adverse events or dropouts due to the events between groups.

Conclusions and my two cents on the injury discussion

“The conclusion is not surprising as a typical yoga class can be classified as low-load and low-to-moderate intensity exercise [20–22]”

Based on observational and experimental evidence, yoga is a safe exercise discipline and practicing yoga is as safe as taking part in general exercise. This is repeatedly reported by studies and it is very unlikely that this conclusion will change in the future. The conclusion is not surprising as a typical yoga class can be classified as low-load and low-to-moderate intensity exercise [20–22]. When compared with high-load activities, adverse event risk is much lower in yoga. When adverse events do happen, they are usually mild sprains and strains that heal on their own. Serious complications do happen; however, they are rare. Participants with an existing medical condition should discuss the safety of their practice with a medical professional.

Evidence of adverse event mechanisms and risk factors is inconclusive. However, no pose can be deemed unsafe. As the risk of adverse events in yoga is low, typically performed yoga poses seem well tolerated. Furthermore, there seem to be no safe and unsafe yoga styles. The adverse events are probably more common in physically demanding styles due to higher loading, but the styles do not come through in the scientific literature as injurious. The most important risk factor seems to be the time spent on practicing yoga, which is not surprising as the more you do something, the higher the risk of adverse events is even due to chance. In the future, if researchers want to study the potential risk factors of yoga-related adverse events, I would encourage them to devote their time to conducting a large enough prospective cohort study with a long enough follow-up. Maybe in the future this is possible as funding opportunities may improve as yoga becomes more and more mainstream. Also, RCTs investigating yoga interventions should monitor and report adverse events more rigorously.

“My subjective view is that in the yoga world injury prevention has mainly revolved around optimizing alignment in poses. I would like to focus the attention to a different direction as injuries do not happen in a vacuum and their prevention is much more than how you perform a certain exercise.”

What about injury prevention? Currently no studies exist and, as the adverse event risk in yoga is low, research is not desperately needed. My subjective view is that in the yoga world injury prevention has mainly revolved around optimizing alignment in poses. I would like to focus the attention to a different direction as injuries do not happen in a vacuum and their prevention is much more than how you perform a certain exercise. For example, if you specialize in yoga and devote your time getting better at more and more advanced physical practice, training should be progressive and recovery must be adequate. If a high training load is combined with a restrictive diet and possibly with psychosocial stress, the risk of adverse events probably elevates. I dare to hypothesize that these practice characteristics and lifestyle factors are much more important than specific alignment rules. Perhaps practicing poses with varying ways might even make the practice more sustainable?

I also want to raise awareness of the language yoga teachers use. I understand that many see yoga as a healing practice and ahimsa (non-violence) is one of the key values of yoga. Thus, teachers may try to protect students from all possible adverse events by repeating warnings after warnings during a class. However, words aimed to protect can injure too! Endless cues on how to prevent adverse events can lead to a thought of humans as fragile and unadaptable machinelike systems. This may even sensitize some students to experience more pain at some point in their life [23]. After all, this mindset is not justified with scientific evidence.

For further reading, see 7 Prominent Yogis Weigh in on Yoga Injuries and What to Do About Them

References

Finch, C. 2008. A new framework for research leading to sports injury prevention. J Sci Med Sport. 9(1-2):3-9.
Brooks, J. H. & Fuller, C. W. 2006. The influence of methodological issues on the results and conclusions from epidemiological studies of sports injuries: illustrative examples. Sports Med. 36(6):459-72.
Fuller, C & Drawer, S. 2004. The application of risk management in sport. Sports Med. 34(6):349-56.
Hütter, G., Nowak, D., Mossner, M., Ganepola, S., Müssig, A., Allers, K., Schneider, T., Hofmann, J., Kücherer, C., Blau, O., Blau, I. W., Hofmann, W. K. & Thiel, E. 2009. Long-term control of HIV by CCR5 Delta32/Delta32 stem-cell transplantation. N Engl J Med. 360(7):692-8. doi: 10.1056/NEJMoa0802905.
Gupta, R. K., Abdul-Jawad, S., McCoy, L. E., Mok, H. P., Peppa, D., Salgado, M., Martinez-Picado, J., Nijhuis, M., Wensing, A. M. J., Lee, H., Grant, P., Nastouli, E., Lambert, J., Pace, M., Salasc, F., Monit, C., Innes, A. J., Muir, L., Waters, L., Frater, J., Lever, A. M. L., Edwards, S. G., Gabriel, I. H. & Olavarria, E. 2019. HIV-1 remission following CCR5Δ32/Δ32 haematopoietic stem-cell transplantation. Nature. 568(7751):244-248. doi: 10.1038/s41586-019-1027-4.
Cramer, H., Krucoff, C. & Dobos, G. 2013. Adverse events associated with yoga: a systematic review of published case reports and case series. PLoS One. 8(10):e75515. doi: 10.1371/journal.pone.0075515.
Cramer, H., Ostermann, T. & Dobos, G. 2018. Injuries and other adverse events associated with yoga practice: A systematic review of epidemiological studies. J Sci Med Sport. 21(2):147-154. doi: 10.1016/j.jsams.2017.08.026.
Mikkonen, J., Pedersen, P. & McCarthy, P. W. 2008. A Survey of Musculoskeletal Injury among Ashtanga Vinyasa Yoga Practitioners. Int J Yoga Therap. 18(1):59-64.
Penman, S., Cohen, M., Stevens, P. & Jackson, S. 2012. Yoga in Australia: Results of a national survey. Int J Yoga. 5(2):92-101. doi: 10.4103/0973-6131.98217.
Uebelacker, L. A., Weinstock, L. M., Kraines, M. A. 2014. Self-reported benefits and risks of yoga in individuals with bipolar disorder. J Psychiatr Pract. 20(5):345-52. doi: 10.1097/01.pra.0000454779.59859.
Mace, C. & Eggleston, B. 2016. Self-Reported Benefits and Adverse Outcomes of Hot Yoga Participation. Int J Yoga Therap. 26(1):49-53.
Park, C. L., Riley, K. E. & Braun, T. D. 2016. Practitioners' perceptions of yoga's positive and negative effects: Results of a National United States survey. J Bodyw Mov Ther. 20(2):270-9. doi: 10.1016/j.jbmt.2015.11.005.
Lauche, R., Schumann, D., Sibbritt, D., Adams, J. & Cramer, H. 2017. Associations between yoga practice and joint problems: a cross-sectional survey among 9151 Australian women. Rheumatol Int. 37(7):1145-1148. doi: 10.1007/s00296-017-3744-z.
Bueno, A. M., Pilgaard, M., Hulme, A., Forsberg, P., Ramskov, D., Damsted, C. & Nielsen, R. O. 2018. Injury prevalence across sports: a descriptive analysis on a representative sample of the Danish population. Inj Epidemiol. 5(1):6. doi:10.1186/s40621-018-0136-0.
Cramer, H., Quinker, D., Schumann, D., Wardle, J., Dobos, G. & Lauche, R. 2019. Adverse effects of yoga: a national cross-sectional survey. BMC Complement Altern Med. 19(1):190. doi: 10.1186/s12906-019-2612-7.
Wiese, C., Keil, D., Rasmussen, A. S. & Olesen, R. 2019. Injury in yoga asana practice: Assessment of the risks. J Bodyw Mov Ther. 23(3):479-488. doi:10.1016/j.jbmt.2018.09.151.
Campo, M., Shiyko, M. P., Kean, M. B., Roberts, L. & Pappas, E. 2018. Musculoskeletal pain associated with recreational yoga participation: A prospective cohort study with 1-year follow-up. J Bodyw Mov Ther. 22(2):418-423. doi:10.1016/j.jbmt.2017.05.022.
Niemeijer, A., Lund, H., Stafne, S. N., Ipsen, T., Goldschmidt, C. L., Jørgensen, C. T. & Juhl, C. B. 2019. Adverse events of exercise therapy in randomised controlled trials: a systematic review and meta-analysis. Br J Sports Med. 2019 Sep 28. pii: bjsports-2018-100461. doi: 10.1136/bjsports-2018-100461.
Cramer, H., Ward, L., Saper, R., Fishbein, D., Dobos, G. & Lauche R. 2015. The Safety of Yoga: A Systematic Review and Meta-Analysis of Randomized Controlled Trials. Am J Epidemiol 182(4):281-93. doi: 10.1093/aje/kwv071.
Wilcox, S. J., Hager, R., Lockhart, B. & Seeley, M. K. 2012. Ground Reaction Forces Generated by Twenty-eight Hatha Yoga Postures. Int J Exerc Sci. 5(2):114-126.
Sherman, S. A., Rogers, R. J., Davis, K. K., Minster, R. L., Creasy, S. A., Mullarkey, N. C., O'Dell, M., Donahue, P. & Jakicic, J. M. 2017. Energy Expenditure in Vinyasa Yoga Versus Walking. J Phys Act Health. 14(8):597-605. doi: 10.1123/jpah.2016-0548.
Pryor, J. L., Christensen, B., Jackson, C. G. R. & Moore-Reed, S. 2019. Metabolic Demands of Yoga at Varying Tempos and Compared With Walking. J Phys Act Health. 16(7):575-580. doi: 10.1123/jpah.2018-0283.
Stewart, M & Loftus, S. 2018. Sticks and Stones: The Impact of Language in Musculoskeletal Rehabilitation. J Orthop Sports Phys Ther. 48(7):519-522. doi: 10.2519/jospt.2018.0610.