Open access
Editor's Choice

Effect of yoga on health-related outcomes in people at risk of fractures: a systematic review

Publication: Applied Physiology, Nutrition, and Metabolism
16 December 2021


We summarized the effects of yoga on health-related outcomes and adverse events in men and postmenopausal women ≥50 years-old at increased risk of fracture, to inform the updated Osteoporosis Canada clinical practice guidelines. Six databases were searched for observational studies, randomized controlled trials and case series. Certainty of evidence was assessed using the Grading of Recommendations, Assessment, Development and Evaluation handbook. Nine studies were included and reported using narrative syntheses due to the limited available evidence. Overall, the available evidence was of very low certainty. There was no effect of yoga on health-related quality of life in randomized trials. Effects on other health-related outcomes were mixed or not available in the literature. Five studies reported no adverse events directly related to the study intervention, and 2 studies did not report whether adverse events occurred. However, 2 case series reported vertebral fractures related to yoga participation, possibly due to excessive spinal flexion. Due to the limited and very low certainty evidence, guideline developers will need to draw indirect evidence from yoga studies among middle aged or older adults that are not at fracture risk. PROSPERO: CRD42019124898.
Evidence in general was of very low certainty.
Yoga had no effect on health-related quality of life in randomized trials. Evidence was mixed or unavailable for other outcomes.
Case studies reported yoga poses involving spinal flexion coincided with incidents of vertebral compression fracture among older adults with increased fracture risk.


Nous avons résumé les effets du yoga sur les résultats liés à la santé et les événements indésirables chez les hommes et les femmes postménopausées de ≥50 ans présentant un risque accru de fracture, et ce, afin de préciser les lignes directrices de pratique clinique mises à jour d’Ostéoporose Canada. Six bases de données ont été consultées pour des études d’observation, des essais contrôlés randomisés et des séries de cas. La certitude des données probantes a été évaluée à l’aide du manuel Grading of Recommendations, Assessment, Development and Evaluation. Neuf études ont été incluses et rapportées à l’aide de synthèses narratives en raison du peu de données probantes disponibles. Dans l’ensemble, les preuves disponibles étaient de très faible certitude. Il n’y avait aucun effet du yoga sur la qualité de vie liée à la santé dans les essais randomisés. Les effets sur d’autres résultats liés à la santé étaient mitigés ou n’étaient pas disponibles dans la documentation. Cinq études n’ont rapporté aucun événement indésirable directement lié à l’intervention dans l’étude et deux études n’ont pas rapporté si des événements indésirables se sont produits. Cependant, deux séries de cas ont rapporté des fractures vertébrales liées à la pratique du yoga, probablement dues à une flexion vertébrale excessive. En raison des données probantes limitées et de très faible certitude, les élaborateurs de lignes directrices devront tirer des données probantes indirectes d’études sur le yoga auprès d’adultes d’âge moyen ou d’adultes plus âgés qui ne présentent pas de risque de fracture. PROSPERO : CRD42019124898. [Traduit par la Rédaction]
Les nouveautés :
Les données probantes en général étaient d’une très faible certitude.
Le yoga n’a eu aucun effet sur la qualité de vie liée à la santé dans les essais randomisés. Les données probantes étaient mitigées ou indisponibles pour d’autres critères de jugement.
Des études de cas ont rapporté que les postures de yoga impliquant une flexion de la colonne vertébrale coïncidaient avec des incidents de fracture vertébrale par compression chez les personnes âgées présentant un risque accru de fracture.


Osteoporosis is a skeletal disorder characterized by compromised bone strength and a consequent increased risk of fragility fractures (Klibanski et al. 2001). Chronic disabling pain, fear, anxiety and depression are common co-morbidities of osteoporosis (Nagae et al. 2006; Bhattacharya et al. 2014; Sozen et al. 2017). According to data from Statistics Canada from 2015 to 2016, approximately 2.2 million Canadians 40 years and older were estimated to be living with osteoporosis and women were 1.3 to 3 times more likely to sustain a fracture at the arm, spine or pelvis and hip (Statistics Canada 2020). Furthermore, the economic burden of osteoporosis in Canada after adjusting for inflation has increased over the recent years, from $2.5 billion CAD in 2008 to $4.6 billion CAD in 2011 (Hopkins et al. 2016). Given the high prevalence of osteoporosis among Canadians and the economic burden of osteoporotic fractures, exercise interventions that can safely help to maintain bone health and prevent osteoporotic fractures should be implemented.
Yoga is an increasingly popular activity among older adults (Clarke et al. 2015) and studies have suggested that yoga can improve back pain, posture, body composition and health-related quality of life (HRQoL) in healthy adult populations (Shanb and Youssef 2014; Gupta and March 2016; Kelly and Gilman 2017). However, some yoga poses involve spinal flexion or twisting, and a recent national survey of >1000 patients who are members of the Canadian Osteoporosis Patient Network revealed that patients often have questions about the safety and efficacy of yoga for older adults with low bone mass (Morin et al. 2020). The survey also revealed that the health outcomes that are important to patients include physical functioning, quality of life and autonomy (Morin et al. 2020).
The present review is part of a series of systematic reviews (Ponzano et al. 2021a, 2021b; Rodrigues et al. 2021a, 2021b) investigating the effects of different types of exercise on health-related outcomes in people with low bone mass and will inform the development of the Osteoporosis Canada Clinical Practice Guidelines for Management of Osteoporosis and Fracture Prevention in Canada. The purpose of the current systematic review is to report on what we know about the effect of yoga interventions on falls, fractures, and other health-related outcomes in men and postmenopausal women at least 50 years old with low bone mass or a history of a fragility fracture.

Materials and methods

Protocol and registration

This systematic review was reported according to the Preferred Reporting Items for Systematic Review and Meta-Analysis 2020 statement (Page et al. 2021). The protocol was informed by the Cochrane Handbook for Systematic Review of Interventions (Cochrane 2020), and registered with the International Prospective Register of Systematic Reviews at (registration number CRD42019124898, submitted and last updated on February 20th, 2020). The protocol was designed by a working group comprised of researchers, physiotherapists, physicians, a patient partner, and graduate students.

Search strategy

A librarian (J.S.), in collaboration with the research team, performed searches within the following databases: MEDLINE (Ovid), EMBASE (Ovid), Cochrane CENTRAL (clinical trials), Cochrane Database of Systematic Reviews, CINAHL (allied health journal content), Epistemonikos, and Web of Science. The search strategies incorporated a combination of subject headings (i.e., Medical Subject Headings) and author keywords for the following concepts: “osteoporosis”, “exercise”, and “older adults”. The full search strategies are reported in Appendix A. There were no restrictions on gender, ethnicity, exercise setting, or if the research originated from a low-income country. We included human studies, cohort studies, case-control studies, cross-sectional studies, case reports, randomized controlled trials (RCTs) or quasi-RCTs. We included publications written in English, Portuguese, Spanish, Italian, or Farsi, because members of the research team had sufficient communication skills in those languages. Literature searches were performed in August/September 2018 and updated in December 2020. Search results were de-duplicated in Endnote and then records were imported into Covidence (Veritas Health Innovation, Melbourne, Australia).

Study selection

Pairs of reviewers independently assessed the eligibility criteria at each phase: title and abstract screening of studies for all exercises investigated in our series of systematic reviews (I.B.R., M.P., J.T., N.T., K.V.K., M.C.A.), full-text review of eligible yoga studies (J.B., I.B.R., K.V.K.), and data extraction (M.C.A., K.V.K.). Screening for eligible studies was performed in Covidence (Veritas Health Innovation, Melbourne, Australia) and data extraction was completed using a template created on Microsoft Excel (2007). Any conflicts between reviewers were resolved by discussion and if an agreement could not be reached, by a third reviewer (L.G.).

Eligibility criteria


We included studies with men and postmenopausal women aged 50 years or older with either: (1) a diagnosis of low bone mass or osteoporosis at the femoral neck or the lumbar spine (T-score ≤ –1) measured with dual-energy X-ray absorptiometry (DXA); (2) a history of a fragility fracture (i.e., fracture of the spine, hip, wrist, or humerus caused by minimal trauma); or (3) a score of moderate or high-risk of a fracture based on a 10-year risk using either the CAROC, FRAX, or GARVAN calculators. Studies with postmenopausal women were not included unless a subgroup analysis was conducted in individuals with low bone mass, or at least 80% of the participants had low bone mass. Authors were contacted if the inclusion criteria of the studies were unclear. We excluded studies of individuals diagnosed with secondary osteoporosis, glucocorticoid-induced osteoporosis, or traumatic fractures.


We included studies that investigated an intervention that was labeled as yoga or as a specific branch of yoga (e.g., Hatha, Bikram, Yogasana). Trials where the intervention was solely meditation-based were excluded. Exercise could be home-delivered or performed in a centre outside the home, group-based or individual programs, and either supervised or non-supervised. Trials combining yoga with other types of exercise were included; we planned to report results from studies that applied an intervention that consisted of yoga only or that consisted of yoga in combination with other exercises. Studies were excluded if they involved a pharmacological therapy intervention that was not administered to all the study groups.


For RCTs or non-randomized clinical trials, we originally intended to include studies if at least 1 group received a placebo, a non-exercise, or a non-physical therapy intervention (e.g., educational intervention or stretching), but had to revise our inclusion criteria to include studies with an active control due to the limited number of studies. If the study was not an RCT or quasi-RCT, a control group was not required for inclusion.


We established a list of outcomes deemed critical or important to individuals living with osteoporosis (Morin et al. 2020). The outcomes considered critical for decision making were: (1) mortality, due to any cause such as natural, disease, or injury-related circumstances that resulted in a fatal injury or in death; (2) fracture-related mortality defined as deaths attributed to a fragility fracture; (3) fragility fractures, fracture of the spine, wrist, humerus, or pelvis caused by minimal trauma; (4) hip fractures, fracture at the femoral neck or trochanter; (5) number of falls experienced or number of people who experienced 1 or more falls during the study or fall-related injuries; (6) physical functioning and disability, any validated performance-based measure of physical functioning (e.g., gait speed, 5× sit-to-stand, Timed Up and Go (TUG)) but not including isolated measures of muscle strength (e.g., knee extensor strength); (7) HRQoL, determined using any validated generic QoL questionnaire or osteoporosis-specific QoL questionnaire; and (8) serious adverse events, defined as any untoward medical occurrence, that at any dose, results in death, inpatient hospitalization, prolongation of existing hospitalization, persistent or significant disability/incapacity, or is life-threatening, (Health Canada 2018) or non-serious adverse events, defined as any reaction related to the intervention such as musculoskeletal injuries (e.g., sprains, strains, joint pain, overuse injuries) not requiring immediate medical attention.
Falls and bone mineral density (BMD) were considered indirect outcomes for fall-related injuries and fracture risk, respectively. We included studies that reported BMD (g/cm2) at the lumbar spine, total hip, and femoral neck, or bone strength outcomes measured using peripheral quantitative computed tomography. Pain was not voted as a critical outcome for the guidelines but was included in our review, measured using any validated questionnaire such as a pain intensity scale (e.g., Visual Analog Scale), a global measurement scale (e.g., overall improvement, proportion of participants recovered), or a generic functional status (e.g., SF-36, Nottingham Health Profile) (Haefeli and Elfering 2006).

Time frame

Our working group decided via consensus that we would include studies where the intervention lasted at least 4 weeks (Ponzano et al. 2021b; Rodrigues et al. 2021a), because we wanted to exclude studies of acute effects of exercise.

Study design

Cohort studies, case-control studies, cross-sectional studies, case reports, RCTs, quasi-randomized and non-randomized controlled clinical trials were considered for inclusion. We decided to consider non-RCTs as we anticipated that there would be very few eligible studies and wanted to use all available evidence for decision-making.

Risk of bias

The Cochrane Risk of Bias tool was used to assess RCTs and quasi-RCTs. The Cochrane Risk of Bias in Non-randomized Studies – of Interventions tool (ROBINS-I tool) (Sterne et al. 2016) was originally planned to be used to assess observational studies. However, the ROBINS-I tool could not be applied to all observational studies included in this review because they did not have a comparator group (Sterne et al. 2016). Thus, we rated all observational studies as serious/very serious risk of bias, and low or very low certainty evidence, in accordance with guidance from the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) handbook (The GRADE Working Group 2013).

Data synthesis and statistical analysis

We collected participant demographic characteristics identified in PROGRESS-Plus (O’Neill et al. 2014), which is an acronym for place of residence, race/ethnicity/culture/language, occupation, gender/sex, religion, education, socioeconomic status, social capital, personal characteristics associated with discrimination, features of relationships, and time-dependent relationships.
We intended to conduct meta-analyses as reported in our protocol, but meta-analyses could not be performed due to the low number of RCTs included in the review; therefore, we reported the results using narrative syntheses. For each outcome, we reported between-group differences from studies with a comparator, and within-group differences from studies without a comparator. We used tables to present study characteristics and results from individual studies. In the tables reporting our outcomes, we used footnotes to identify study designs.


Study selection and characteristics

Our initial search strategy yielded 7286 records of any type of exercise included in our series of systematic reviews, with 4243 records remaining after de-duplication (Fig. 1). Dual screening at the title/abstract and full text screening stages resulted in the inclusion of 3 RCTs (Tüzün et al. 2010; Guo et al. 2019; Grahn Kronhed et al. 2020), 1 serial controlled repeated measure pilot study (Fishman 2009), 3 pre-post studies (Vardar Yağlı and Ülger 2012; Lu et al. 2016; Motorwala et al. 2016), and 2 case series (Sinaki 2013; Sfeir et al. 2018) exploring the effects of yoga on health-related outcomes in adults with osteoporosis (Table 1). All participants in the study by Guo et al. (2019) had sustained a hip fracture prior to the study and completed their respective interventions as in-bed exercises. The study by Fishman (2009) published the preliminary results from the experimental pre-post study by Lu et al. (2016) (Table 1). Two of the cases reported by Sinaki (2013) were also reported by Sfeir et al. (2018), so we included the 2 cases only once, for a total number of 10 unique cases.
Fig. 1.
Fig. 1. PRISMA 2020 flow diagram for new systematic reviews, which included searches of databases and registers only.
Table 1.
Table 1. Summary of study characteristics (total number of analyzed participants N = 232).

Study participants

The majority of participants were female, and 3 studies only recruited female participants (Tüzün et al. 2010; Vardar Yağlı and Ülger 2012; Motorwala et al. 2016). Four studies did not state age-related inclusion criteria but were included in this review for the following reasons: Motorwala et al. (2016) were contacted and confirmed that ≥80% of the participants were aged ≥50 years; in 3 other studies, the mean age (Fishman 2009) or individual age of participants (Sinaki 2013; Sfeir et al. 2018) reported in the studies satisfied our age inclusion criterion. Three of the included studies specified that study participants were taking a medication that would affect bone metabolism. In the study by Tüzün et al. (2010), all participants received risedronate 5 mg/day, 100 mg elementary calcium and active vitamin D metabolites, and 5 participants from the case series (Sinaki 2013; Sfeir et al. 2018) had received low dose prednisone for 2 months (n = 1), a bisphosphonate (n = 2), or a bisphosphonate and hormone replacement therapy (n = 2). One observational study did not have low bone mass as an inclusion criterion, but was included in this review because 83% of study participants were reported to have osteoporosis or osteopenia (Lu et al. 2016). Reported PROGRESS-Plus characteristics (O’Neill et al. 2014) included personal characteristics associated with discrimination (i.e., age) and time dependent relationships (i.e., time leaving hospital).


Four studies included a comparator group in the study design. One of these studies was a 3-arm RCT that did not report results from the exercise control group and only presented results from the yoga intervention group and non-active control group (Grahn Kronhed et al. 2020). One serial controlled pre-post study had an inactive control group (Fishman 2009). Two studies had active control groups that performed deep abdominal breathing exercises (Guo et al. 2019) or strength and balance exercises (Tüzün et al. 2010).


Most yoga programs included a warm-up, a series of various yoga poses and cool-down period (Fishman 2009; Vardar Yağlı and Ülger 2012; Motorwala et al. 2016). Two studies specified the yoga type (Tüzün et al. 2010; Lu et al. 2016) while others listed specific yoga poses performed (Fishman 2009; Vardar Yağlı and Ülger 2012; Motorwala et al. 2016; Guo et al. 2019; Grahn Kronhed et al. 2020). The yoga sequence designed by Guo et al. (2019) was completed in bed and only involved the upper body, unlike the other studies that performed full-body yoga. The length and frequency of yoga sessions varied from 12 to 60 min, once a day, for 2 to 7 days per week (Table 1). One study asked participants to perform yoga twice daily, 7 days per week (Guo et al. 2019). The duration of each study ranged from 4 weeks to 2 years (Table 1). The level of supervision during yoga varied from no supervision (Fishman 2009; Lu et al. 2016), partial supervision (Guo et al. 2019; Grahn Kronhed et al. 2020), to full supervision (Tüzün et al. 2010; Vardar Yağlı and Ülger 2012; Motorwala et al. 2016). Participants received supervision from a physiotherapist or physical therapist (Tüzün et al. 2010; Motorwala et al. 2016; Guo et al. 2019; Grahn Kronhed et al. 2020), yoga instructor (Tüzün et al. 2010; Motorwala et al. 2016), or nurse (Guo et al. 2019). The length, frequency and duration of yoga participation were not reported in detail in the case reports but the poses claimed to have caused vertebral compression fractures are included in Table 2 (Sinaki 2013; Sfeir et al. 2018).
Table 2.
Table 2. Adverse events reported by case series.

Adherence rate

Adherence to a yoga program varied across the included studies. In order of increasing adherence, a serial controlled pre-post study reported 15% adherence (Fishman 2009), an observational study reported 75% adherence (Vardar Yağlı and Ülger 2012), an RCT reported 80% adherence (Grahn Kronhed et al. 2020), an RCT reported 93% adherence (Guo et al. 2019), and 1 RCT (Tüzün et al. 2010) and 1 observational study (Motorwala et al. 2016) reported full adherence to a yoga program. Adherence could not be calculated for the ongoing pilot study by Lu et al. (2016) because most of the participants had not completed the full trial. Adherence could not be determined for the case series due to the type of study design.

Risk of bias

The overall risk of bias of the included studies was very serious. The majority of studies used an observational design with no comparator group or used an active control that may have influenced 1 or more outcomes. Only 3 studies provided reasons for loss to follow-up or exclusion of participants at follow-up. Reasons included incomplete data, low compliance, withdrawal due to an adverse event unrelated to study protocol, early discharge of participants from the hospital, disappointment of being randomized to the control group, economic reasons, and change in location (Lu et al. 2016; Guo et al. 2019; Grahn Kronhed et al. 2020). Data were also missing for some measured outcomes, raising concern for selection bias (Tüzün et al. 2010; Grahn Kronhed et al. 2020).

Mortality, fractures, falls, and bone mineral density

There were no studies reporting on mortality or fracture-related mortality, hip fractures, fragility fractures, fall-related injuries, or falls as outcomes, with the exception of a case series reporting fractures as adverse events (see below). Three observational studies presented BMD as DXA T-scores (Fishman 2009; Motorwala et al. 2016) or as mean monthly BMD change values (Lu et al. 2016), rather than mean group BMD in g/cm2, and could not be interpreted with certainty.

Health-Related quality of life

The effect of yoga on HRQoL in people at risk of fracture was uncertain (3 studies, 53 participants, very low certainty evidence). HRQoL was reported by 2 RCTs (Tüzün et al. 2010; Grahn Kronhed et al. 2020) and 1 observational study (Vardar Yağlı and Ülger 2012). The RCT by Grahn Kronhed et al. (2020) did not provide post-intervention between-group effect estimates, and stated that there were no statistically significant effects of yoga on HRQoL. Therefore, our observations for between-group differences were limited to results from 1 RCT by Tüzün et al. (2010), which reported no statistically significant between-group differences for the HRQoL questionnaire of the European Foundation for Osteoporosis (QUALEFFO-41) (Lips et al. 1997) total or sub-domain scores (Table 3). There was 1 observational study (Vardar Yağlı and Ülger 2012) that did not have a control group and statistically significant within-group differences were reported for the Turkish Nottingham Health Profile (NHP) questionnaire total score and all domain-specific scores (Table 3).
Table 3.
Table 3. Studies reporting health-related quality of life (n = 3).

Physical functioning and disability

The effect of yoga on physical functioning in people at risk of fracture was uncertain (3 studies, 117 participants, very low certainty evidence). Physical functioning was reported by 2 RCTs (Tüzün et al. 2010; Guo et al. 2019) and 1 observational study (Vardar Yağlı and Ülger 2012).
The RCT by Tüzün et al. (2010) assessed balance and mobility using a neuromuscular battery test (Kerschan-Schindl et al. 2001) consisting of a 1-leg stance, tandem walk and a body sway activity, but only 1-leg stance test scores were reported by the authors. There were no between-group differences at the end of study for the left (meanyoga = 14.18 (1.50) seconds vs. meancontrol = 12.79 (3.78) seconds; p = 0.762) and right (meanyoga = 13.85 (2.30) seconds vs. meancontrol = 11.85 (3.32) seconds; p = 0.153) 1-leg stance scores.
The RCT by Guo et al. (2019) reported mean Barthel Index (BI) (Hsieh et al. 2007) scores at timepoints T1, T2 and T3, which corresponded to participants’ scores at hospital admission, pre-intervention and post-intervention, respectively. Participants received 7 days of training to become familiar with the intervention from T1 to T2 and completed either the control group or yoga exercise program for 4 weeks from T2 to T3. There was no between-group difference at T1 (meanyoga = 18.08 (6.85) vs. meancontrol = 19.00 (9.62); p = 0.665), but significant between-group differences were reported at T2 (meanyoga = 38.59 (8.66) vs. meancontrol = 33.00 (9.32); p = 0.009) and T3 (meanyoga = 70.77 (10.23) vs. meancontrol = 65.75 (11.30); p = 0.019) (Guo et al. 2019).
One observational study showed statistically-significant performance improvements on the Timed Up and Go test from 12.26 (1.81) seconds to 9.94 (1.19) seconds (p < 0.001) (Vardar Yağlı and Ülger 2012).


The effect of yoga on pain in individuals at risk of fractures was uncertain (2 studies, 27 participants, very low certainty evidence). Pain was reported by 1 RCT (Grahn Kronhed et al. 2020) and 1 observational study (Vardar Yağlı and Ülger 2012). Grahn Kronhed et al. (2020) assessed present pain, pain last week and worst pain using the numerical pain rating scale. The authors did not publish or provide between-group post-intervention pain scores, and stated that there were no statistically significant effects of yoga on pain (Grahn Kronhed et al. 2020). Participants from the observational study significantly improved their mean visual analog scale score from 8.22 (SD 1.44) to 4.46 (SD 0.79) (p < 0.001) after completing a 12-week yoga program (Vardar Yağlı and Ülger 2012).

Adverse events

Adverse events were difficult to describe due to inadequate reporting and risk of bias (7 studies, 304 participants, very low certainty evidence). One RCT (Tüzün et al. 2010) and 1 observational study (Vardar Yağlı and Ülger 2012) did not report whether or not adverse events occurred. Two RCTs and 3 observational studies reported no adverse events directly related to the study intervention (Fishman 2009; Lu et al. 2016; Motorwala et al. 2016; Guo et al. 2019; Grahn Kronhed et al. 2020). Of these studies, 1 study reported 109 participants had a fracture before the study and an additional 19 participants had a fracture after study enrolment but did not attribute the fractures to participation in yoga (Guo et al. 2019). Two case series reported incidents of vertebral compression fractures (VCFs) that were reported by patients to have occurred during yoga poses involving spinal flexion, and the majority of the fractures were localized to the thoracolumbar region of the spine (Table 2) (Sinaki 2013; Sfeir et al. 2018).


There is very low certainty evidence from RCTs that yoga does not improve HRQoL in individuals at risk of fracture compared with education, or active control. The effects on physical functioning and pain are uncertain. While case studies reported incidents of fractures potentially caused by yoga, it is difficult to confirm that the fractures were attributable to yoga, or to ascertain point estimates for risk of fracture. It may be prudent for high-risk individuals to avoid yoga poses that involve extreme spinal flexion. Although the number of studies is limited, it is important for guideline developers to understand the scope of the existing literature to make decisions and inform future research.
Studies of yoga in people at risk of fracture are limited and 2 RCTs were available to provide very low certainty evidence that the effect of yoga on physical functioning is uncertain. Therefore, we may need to use additional indirect evidence from yoga studies in older adults not at risk of fracture to inform guidelines. A systematic review by Sivaramakrishnan et al. (2019) included 5 RCTs (n = 377) that investigated the effect of yoga on physical functioning and health related outcomes in older adults, including balance and walking speed. Compared with inactive controls, yoga significantly improved balance (Hedges’ g = 0.7; 95% CI: 0.19, 1.22) but not walking speed (Hedges’ g = 0.38; 95% CI: -0.02, 0.78) among older adults (Sivaramakrishnan et al. 2019). It is possible that the presence of pain, gait impairments or hyperkyphosis may affect the generalizability of the findings in older adults to those with osteoporosis. However, guideline developers can still draw indirect evidence from older adult populations to inform guidelines for adults with low bone mass with thoughtful considerations of the population’s unique characteristics.
While we did not observe a consistent positive effect of yoga on HRQoL, studies in older adults suggest that yoga may improve HRQoL. Nine RCTs included in the systematic review by Sivaramakrishnan et al. (2019) reported that yoga had a positive effect on perceived mental health (n = 508, SMD= 0.60; 95% CI: 0.33, 0.87) and perceived physical health (n = 354; SMD= 0.61; 95% CI: 0.29, 0.94) among older adults, when compared with inactive controls. Thus, there is some indirect evidence that yoga may improve HRQoL in older adults, which may be used in guidelines for individuals at risk of fracture.
Whether yoga can reduce pain in individuals at risk of fracture is uncertain. Our work suggests that effects on pain were mixed. A systematic review of 9 RCTs including 640 adults aged 50 to 80 years and with osteoarthritis of the knee, hip, hand, feet or spine found very low-quality evidence for a favourable effect of yoga on pain when compared with participants who did not exercise (control group) (SMD = –0.75; 95% CI: –1.18, –0.31; p < 0.001), as measured with various validated pain measurement tools (Lauche et al. 2019). However, effects on arthritis pain may not be generalizable to individuals at risk of fracture. Future studies interested in examining effects of yoga on pain in people at risk of fracture should target individuals with pain at baseline.
People living with osteoporosis have expressed the desire to be informed about the safety of various physical activities, including yoga (Morin et al. 2020). Our work suggests that there is very low certainty evidence to make any conclusions about the safety of yoga for individuals with low bone mass. Two case series reported a potential connection between VCFs and yoga postures involving spinal flexion. However, it was not possible to confirm a causal link with the available information, nor to determine the relative risk of sustaining a VCF between non-exercise controls and yoga participants. In a systematic review of yoga in older adults, reports of non-serious adverse events during yoga were infrequent and limited to musculoskeletal injuries and 1 fall (Sivaramakrishnan et al. 2019). Until there is higher certainty evidence that confirms or refutes the potential risks associated with yoga, instructors and individuals with low bone mass should proceed with caution when performing poses that involve spinal flexion or torsion. Specifically, compressive forces on the spine have been observed to be greatest when one’s centre of mass is anteriorly shifted either by change in position or application of weights, and peak thoracolumbar loading has been observed to increase with increased thoracic kyphosis (Bruno et al. 2017). Individuals with low bone mass or thoracic kyphosis who want to participate in yoga could modify poses involving spinal flexion to maintain a neutral spine. Balance poses should also be performed with caution or with a supportive object nearby.
There were some limitations of our review, and in the available evidence. The number of studies retrieved according to our eligibility criteria was limited and not all outcomes of interest were reported by included studies. Specifically, included studies did not report on the effect of yoga on mortality, falls or BMD. Adherence to yoga as an intervention varied and the acceptability of yoga among older adults with low bone mass could not be determined. However, Sivaramakrishnan et al. (2019) reported 63%–95% adherence to yoga programs that were at least 4 weeks long, among older adults (Sivaramakrishnan et al. 2019). While we placed more emphasis on data from RCTs when making judgements about the certainty of evidence, the number of RCTs was limited and the risk of bias in those studies was high, so they may not provide more certainty than within group analyses or observational studies. A meta-analysis could not be performed due to the small number studies and lack of studies with a comparator group. The generalizability of our review is limited by the fact that the majority of study participants were female. It is possible that our search strategy and inclusion criteria may have omitted relevant studies that were published in languages of countries where yoga practice is more common. During level 1 screening, we excluded 1 abstract that was not published in a language that we could review, i.e., English, Spanish, Portuguese or Italian.


Our review revealed knowledge gaps related to the effect of yoga on health-related outcomes determined as important by patients and health care professionals. There is very low certainty evidence yoga does not improve measures of HRQoL among older adults with low bone mass compared with education or active control, and the effect of yoga on physical functioning and pain was uncertain based on the studies included in our review. No information was available to establish yoga’s effect on fracture-related mortality, hip fractures, fragility fractures, fall-related injuries, mortality, and falls in older adults with low bone mass. The studies included in this review reported no adverse events directly related to the study intervention or did not collect information on adverse events. However, a small number of case reports suggest individuals with low bone mass, health professionals and yoga instructors should practice caution when performing yoga poses involving spinal flexion or torsion, as these may place individuals at risk of vertebral compression fractures. Recommendations for patients should consider the balance between the patient’s desire and ability to do an activity, and the potential risk. Individuals at risk of fracture who wish to practice yoga should seek advice from or attend a yoga class taught by a qualified instructor with knowledge of how to adapt postures for older adults with low bone mass.

Competing interests statement

The authors declare there are no competing interests.

Contributors’ statement

All authors contributed to the design of the study, the interpretation of the findings and the finalization of this paper. The data were collected by K.V.K. and M.C.A., and analyzed by K.V.K., M.C.A. and L.G. K.V.K. led the writing of the manuscript.

Data availability statement

Materials and data may be available upon request.


The authors would like to acknowledge and thank Zachary Fielding, Karolina Godwinska, Emily McLaughlin, Jeff Templeton, Nora Thorpe, Justin Wagler, and Jordan Wolanski for their support and contribution. Maureen C. Ashe gratefully acknowledges support from the Canada Research Chairs program.


Bhattacharya R., Shen C., and Sambamoorthi U. 2014. Excess risk of chronic physical conditions associated with depression and anxiety. BMC Psychiatry, 14: 10.
Bruno A.G., Burkhart K., Allaire B., Anderson D.E., and Bouxsein M.L. 2017. Spinal Loading Patterns From Biomechanical Modeling Explain the High Incidence of Vertebral Fractures in the Thoracolumbar Region. J. Bone Miner. Res. 32(6): 1282–1290.
Clarke T.C., Black L.I., Stussman B.J., Barnes P.M., and Nahin R.L. 2015. Trends in the use of complementary health approaches among adults: United States, 2002–2012. Natl. Health Stat. Report, 10(79): 1–16.
Cochrane. 2020. Cochrane Handbook for Systematic Reviews of Interventions Version 6.1. Edited by J. Higgins.
Fishman L.M. 2009. Yoga for osteoporosis: A pilot study. Top. Geriatr. Rehabil. 25(3): 244–250.
Grahn Kronhed A., Enthoven P., Spångeus A., and Willerton C. 2020. Mindfulness and Modified Medical Yoga as Intervention in Older Women with Osteoporotic Vertebral Fracture. J. Altern. Complement. Med. 26(7): 610–619.
Guo J., Gao C., Xin H., Li J., Li B., Wei Z., and Yue Y. 2019. The application of “upper-body yoga” in elderly patients with acute hip fracture: A prospective, randomized, and single-blind study. J. Orthop. Surg. Res. 14(1): 250.
Gupta A. and March L. 2016. Treating osteoporosis. Aust. Prescr. 39(2): 40–46.
Haefeli M. and Elfering A. 2006. Pain assessment. Eur. Spine J. 15: S17–S24.
Health Canada. 2018. Reporting Adverse Reactions to Marketed Health Products. Guidance Document for Industry. Cat. no. H164-33/2018E-PDF. Health Canada, Ottawa, ON, Canada. Available from
Hopkins R.B., Burke N., Von Keyserlingk C., Leslie W.D., Morin S.N., Adachi J., et al. 2016. The current economic burden of illness of osteoporosis in Canada. Osteoporos. Int. 27(10): 3023–3032.
Hsieh Y., Wang C., Wu S., Chen P., Sheu C., and Hsieh C. 2007. Establishing the minimal clinically important difference of the barthel index in stroke patients. Neurorehabil. Neural Repair, 21(3): 233–238.
Kelly O.J. and Gilman J.C. 2017. Can Unconventional Exercise be Helpful in the Treatment, Management and Prevention of Osteosarcopenic Obesity? Curr. Aging Sci. 10(2): 106–121.
Kerschan-Schindl K., Uher E., Grampp S., Kaider A., Ghanem A.H., Fialka-Moser V., and Preisinger E. 2001. A neuromuscular test battery for osteoporotic women: A pilot study. Am. J. Phys. Med. Rehabil. 80(5): 351–357.
Klibanski A., Adams-Campbell L., Bassford T., Blair S.N., Boden S.D., Dickersin K., et al. 2001. Osteoporosis prevention, diagnosis, and therapy’. JAMA. 285: 785–795.
Lauche R., Hunter D.J., Adams J., and Cramer H. 2019. Yoga for osteoarthritis: a systematic review and meta-analysis. Curr. Rheumatol. Rep. 21: 47.
Lips P., Cooper C., Agnusdei D., Caulin F., Egger P., Johnell O., et al. 1997. Quality of life as outcome in the treatment of osteoporosis: The development of a questionnaire for quality of life by the European Foundation for Osteoporosis. Osteoporosis. Int. 7(1): 36–38.
Lu Y., Rosner B., Chang G., and Fishman L.M. 2016. Twelve-minute daily yoga regimen reverses osteoporotic bone loss. Top. Geriatr. Rehabil. 32(2): 81–87.
Morin S.N., Djekic-Ivankovic M., Funnell L., Giangregorio L., Rodrigues I.B., Ridout R., et al. 2020. Patient engagement in clinical guidelines development: input from > 1000 members of the Canadian Osteoporosis Patient Network. Osteoporos. Int. 31(5): 867–874.
Motorwala Z.S., Kolke S., Panchal P.Y., Bedekar N.S., Sancheti P.K., and Shyam A. 2016. Effects of Yogasanas on osteoporosis in postmenopausal women. Int. J. Yoga, 9(1): 44–48.
Nagae M., Hiraga T., Wakabayashi H., Wang L., Iwata K., and Yoneda T. 2006. Osteoclasts play a part in pain due to the inflammation adjacent to bone. Bone, 39(5): 1107–1115.
O’Neill J., Tabish H., Welch V., Petticrew M., Pottie K., Clarke M., et al. 2014. Applying an equity lens to interventions: Using PROGRESS ensures consideration of socially stratifying factors to illuminate inequities in health. J. Clin. Epidemiol. 67(1): 56–64.
Page M.J., McKenzie J.E., Bossuyt P.M., Boutron I., Hoffmann T.C., Mulrow C.D., et al. 2021. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372: n71.
Ponzano M., Tibert N., Bansal S., Katzman W., and Giangregorio L. 2021a. Exercise for improving age-related hyperkyphosis: a systematic review and meta-analysis with GRADE assessment. Arch Osteoporos. 16: 140.
Ponzano M., Rodrigues I.B., Hosseini Z., Ashe M.C., Butt D.A., Chilibeck P.D., et al. 2021b. Progressive resistance training for improving health-related outcomes in people at risk of fracture: a systematic review and meta-analysis of randomized controlled trials. Phys Ther. 101: pzaa221.
Rodrigues I.B., Ponzano M., Hosseini Z., Thabane L., Chilibeck P.D., Butt D.A., et al. 2021a. The effect of impact exercise (alone or multicomponent intervention) on health-related outcomes in individuals at risk of fractures: a systematic review and meta-analysis of randomized controlled trials. Sports Med. 51: 1273–1292.
Rodrigues I.B., Ponzano M., Butt D.A., Bartley J., Bardai Z., Ashe M.C., et al. 2021b. The effects of walking or nordic walking in adults 50 years and older at elevated risk of fractures: a systematic review and meta-analysis. J. Aging Phys. Act. 29(5): 886–899.
Sfeir J.G., Drake M.T., Sonawane V.J., and Sinaki M. 2018. Vertebral compression fractures associated with yoga: A case series. Eur. J. Phys. Rehabil. Med. 54(6): 947–951.
Shanb A. and Youssef E. 2014. The impact of adding weight-bearing exercise versus nonweight bearing programs to the medical treatment of elderly patients with osteoporosis. J. Family Community Med. 21(3): 176.
Sinaki M. 2013. Yoga spinal flexion positions and vertebral compression fracture in osteopenia or osteoporosis of spine: case series. Pain Practice, 13(1): 68–75.
Sivaramakrishnan D., Fitzsimons C., Kelly P., Ludwig K., Mutrie N., Saunders D.H., and Baker G. 2019. The effects of yoga compared to active and inactive controls on physical function and health related quality of life in older adults- systematic review and meta-analysis of randomised controlled trials. Int. J. Behav. Nutr. Phys. Act, 16: 33.
Sozen T., Ozisik L., and Calik Basaran N. 2017. An overview and management of osteoporosis. Eur. J. Rheumatol. 4(1): 46–56.
Statistics Canada. 2020. Osteoporosis and related fractures in Canada: Report from the Canadian Chronic Disease Surveillance System 2020. Available from [Accessed 14 October 2021.]
Sterne J.A.C., Hernán M.A., Reeves B.C., Savović J., Berkman N.D., Viswanathan M., et al. 2016. ROBINS-I: A tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 355: i4919.
The GRADE Working Group. 2013. GRADE handbook for grading quality of evidence and strength of recommendations. Edited by H. Schünemann Available from
Tüzün S., Aktas I., Akarirmak U., Sipahi S., and Tüzün F. 2010. Yoga might be an alternative training for the quality of life and balance in postmenopausal osteoporosis. Eur. J. Phys. Rehabil. Med. 46(1): 69–72.
Vardar Yağlı N. and Ülger Ö. 2012. Effects of yoga on the quality of life and mobility in geriatric patients with osteoporosis. J. Yoga Phys. Ther. 01(S2).

Appendix A. Search strategies for osteoporosis review

Table A1.
Table A1. MEDLINE search strategy.
Table A2.
Table A2. EMBASE search strategy.
Table A3.
Table A3. CINAHL search strategy.
Table A4.
Table A4. Web of Science search strategy.
Table A5.
Table A5. Cochrane Library search strategy.
Table A6.
Table A6. Epistemonikos search strategy.

Information & Authors


Published In

cover image Applied Physiology, Nutrition, and Metabolism
Applied Physiology, Nutrition, and Metabolism
Volume 47Number 3March 2022
Pages: 215 - 226


Received: 10 November 2021
Accepted: 9 December 2021
Published online: 16 December 2021

Key Words

  1. osteoporosis
  2. bone health
  3. low bone mass
  4. older adults
  5. yoga
  6. health related outcome


  1. ostéoporose
  2. santé des os
  3. faible masse osseuse
  4. personnes âgées
  5. yoga
  6. résultats liés à la santé



Kawon V. Kim
Department of Kinesiology and Health Sciences, University of Waterloo, Waterloo, ON N2L 3G1, Canada.
Joan Bartley
Patient partner, Osteoporosis Canada.
Maureen C. Ashe
Department of Family Practice, The University of British Columbia, Vancouver, BC V6T 1Z3, Canada.
Zahra Bardai
Department of Medicine, McMaster University, Hamilton, ON L8S 4L8, Canada.
Debra A. Butt
Department of Family and Community Medicine, University of Toronto, Toronto, ON M5G 1V7, Canada.
Philip D. Chilibeck
College of Kinesiology, University of Saskatchewan, Saskatoon, SK S7N 5B2, Canada.
Matteo Ponzano
Department of Kinesiology and Health Sciences, University of Waterloo, Waterloo, ON N2L 3G1, Canada.
Isabel B. Rodrigues
Department of Kinesiology and Health Sciences, University of Waterloo, Waterloo, ON N2L 3G1, Canada.
Jackie Stapleton
University of Waterloo Library, University of Waterloo, Waterloo, ON N2L 3G1, Canada.
Lehana Thabane
Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, ON L8S 4L8, Canada.
John D. Wark
Department of Medicine, Bone and Mineral Medicine, Department of Diabetes and Endocrinology, University of Melbourne, Royal Melbourne Hospital, Victoria 3050, Australia.
Department of Kinesiology and Health Sciences, University of Waterloo, Waterloo, ON N2L 3G1, Canada.
Schlegel Research Institute for Aging, Waterloo, ON N2J 0E2, Canada.

Funding Information

Funding for this project was provided by Osteoporosis Canada.

Metrics & Citations


Other Metrics


Cite As

Export Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Cited by

1. Exercise and physical activity in individuals at risk of fracture

View Options

View options


View PDF

Get Access

Login options

Check if you access through your login credentials or your institution to get full access on this article.


Click on the button below to subscribe to Applied Physiology, Nutrition, and Metabolism

Purchase options

Purchase this article to get full access to it.

Restore your content access

Enter your email address to restore your content access:

Note: This functionality works only for purchases done as a guest. If you already have an account, log in to access the content to which you are entitled.





Share Options


Share the article link

Share with email

Email a colleague

Share on social media