Skip to main content

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

Abstract

Background

Yoga has been recommended as a muscle strengthening and balance activity in national and global physical activity guidelines. However, the evidence base establishing the effectiveness of yoga in improving physical function and health related quality of life (HRQoL) in an older adult population not recruited on the basis of any specific disease or condition, has not been systematically reviewed. The objective of this study was to synthesise existing evidence on the effects of yoga on physical function and HRQoL in older adults not characterised by any specific clinical condition.

Methods

The following databases were systematically searched in September 2017: MEDLINE, PsycInfo, CINAHL Plus, Scopus, Web of Science, Cochrane Library, EMBASE, SPORTDiscus, AMED and ProQuest Dissertations & Theses Global. Study inclusion criteria: Older adult participants with mean age of 60 years and above, not recruited on the basis of any specific disease or condition; yoga intervention compared with inactive controls (example: wait-list control, education booklets) or active controls (example: walking, chair aerobics); physical function and HRQoL outcomes; and randomised/cluster randomised controlled trials published in English. A vote counting analysis and meta-analysis with standardised effect sizes (Hedges’ g) computed using random effects models were conducted.

Results

A total of 27 records from 22 RCTs were included (17 RCTs assessed physical function and 20 assessed HRQoL). The meta-analysis revealed significant effects (5% level of significance) favouring the yoga group for the following physical function outcomes compared with inactive controls: balance (effect size (ES) = 0.7), lower body flexibility (ES = 0.5), lower limb strength (ES = 0.45); compared with active controls: lower limb strength (ES = 0.49), lower body flexibility (ES = 0.28). For HRQoL, significant effects favouring yoga were found compared to inactive controls for: depression (ES = 0.64), perceived mental health (ES = 0.6), perceived physical health (ES = 0.61), sleep quality (ES = 0.65), and vitality (ES = 0.31); compared to active controls: depression (ES = 0.54).

Conclusion

This review is the first to compare the effects of yoga with active and inactive controls in older adults not characterised by a specific clinical condition. Results indicate that yoga interventions improve multiple physical function and HRQoL outcomes in this population compared to both control conditions. This study provides robust evidence for promoting yoga in physical activity guidelines for older adults as a multimodal activity that improves aspects of fitness like strength, balance and flexibility, as well as mental wellbeing.

Trial registration

PROSPERO registration number: CRD42016038052.

Background

The World Health Organization’s physical activity (PA) recommendations for older adults (aged 65 years and over) include aerobic, muscle strengthening and balance components [1]. Physical activity levels worldwide decrease with age [2], and the percentage of older adults meeting these recommendations remains low. The United Kingdom (UK) PA guidelines for this age group include the accumulation of at least 150 min of moderate intensity activity or 75 min of vigorous activity per week (MVPA guidelines), as well as activities to improve muscle strength, and balance and coordination on at least two days a week [3]. Thirty-one percent of adults aged 65–74 years and 54% of adults aged 75+ years in England (2015–2016) [4], and 53% of men and 66% of women aged 65 years and over in Scotland (2012–2014) [5], did not meet the MVPA guidelines. The balance guidelines were met by 19% of older men and 12% of older women in Scotland [6]; and only 14% of men and 12% of women in the 65–74 age-group, and 9 % of men and 4 % of women over 75 years met the muscle strength guidelines [6]. Accordingly, the World Health Organization identifies older adults as a strategic priority area for the promotion of physical activity [7].

Yoga is an ancient practice and a way of life that originated in India, and includes the practice of postures, regulated breathing and meditation [8]. It is a mode of activity found to have multiple benefits for older adults [9,10,11]. Previous systematic reviews have provided evidence on the beneficial effects of yoga in older adults in terms of promoting cardiovascular health [12], balance and mobility [10], alleviating depression and improving quality of sleep [9]. A recent systematic review and meta-analysis by Tulloch et al. [13] found that yoga had a medium effect on health related quality of life (HRQoL), and a small effect on mental wellbeing in people aged 60+ years. In this review, HRQoL was measured by physical component summary scales, and mental wellbeing was assessed by mental component summary scales from questionnaires like SF-36 and WHOQOL. However, HRQoL has been described as a concept encompassing several aspects of overall quality of life that can be clearly shown to affect health [14], including anxiety, stress, depression, vitality, social health and sleep [15], which were not assessed in the review.

Physical function is another relevant outcome for the older adult population and includes aspects such as cardio-respiratory fitness, muscular strength, flexibility and balance [16, 17]. Benefits of performing muscle strength activities in older adults include the offsetting of age-related muscle loss (sarcopenia), enhanced functional performance, improved bone mineral density (BMD), and prevention of falls [18, 19]. Whilst yoga has been specifically recommended as a muscle strengthening activity as part of several national PA guidelines including the UK and United States (US) [20, 21], there have been no previous attempts to synthesise the evidence base to support this recommendation for the older adult population. Patel et al. [11] studied the effects of yoga on some physical function and HRQoL outcomes in older adults from randomised controlled trials (RCT) published between 1950 and 2010. Results of the meta-analysis showed that yoga may be significantly better than controls in improving self-rated health status and aerobic fitness, but no significant differences were found for depression. However, the narrative and quantitative analysis in the Patel et al. review [11] combined data in which yoga was compared with active (example: walking, Tai chi, stretching exercises) and inactive controls (example: usual care, socialisation, education group), making it difficult to draw conclusions on whether any true effects (statistically significant) of yoga compared to other exercise programmes exist, and the strength (magnitude) of these effects.

Tulloch et al. [13] and Patel et al. [11] included studies involving older participants with clinical conditions. Other systematic reviews have focused on yoga in specific clinical groups such as cancer [22], Type 2 Diabetes [23, 24] and rheumatic diseases [25], and found some evidence that yoga has beneficial effects on physiological, physical function and psychosocial outcomes in these populations. Results from studies which only recruited participants with specific diseases or conditions cannot be generalised to all older adults. The yoga interventions used in studies involving clinical populations may have been specially developed to address particular symptoms (example: dyspnea related distress in older adults with chronic obstructive pulmonary disease [26]). It is also difficult to disentangle the effects of yoga when data from heterogeneous groups with different clinical conditions are merged in a review.

Therefore, the present systematic review aims to address limitations in previous reviews and expand on existing evidence in three ways: i) including a comprehensive list of physical function and HRQoL outcome measures; ii) comparing yoga against distinct active and inactive controls so that the relative benefits of yoga can be assessed; and iii) reviewing the effectiveness of yoga in studies where older adult participants were not recruited on the basis of a specific disease or condition. The objective of this review was to assess the effectiveness of yoga compared to active and inactive controls on physical function and HRQoL in older adults not characterised by a specific clinical condition, based on randomised/cluster randomised controlled trials.

Methods

The review was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [27], and recommendations of the Cochrane collaboration [28]. The protocol was developed in advance of the study and registered on PROSPERO (Registration number: CRD42016038052).

Search and selection criteria

The inclusion and exclusion criteria for studies were as follows: (i) Participants: older adults defined as mean age 60 years and above, not recruited on the basis of a specific disease or condition were included; (ii) Intervention and comparison: studies comparing yoga interventions with active and inactive controls were included. Studies in which yoga was specified as a control condition or where yoga was combined with other practices or exercise forms were excluded; (iii) Outcomes: only studies reporting physical function and/or HRQoL outcomes were included; (iv) Study type: studies with a randomised (including cluster randomised) controlled study design published in English were included.

A mean age of 60 years and above was a criterion for inclusion. The retirement age in countries like India and China is 60 years [29, 30], and the United Nations defines older persons as those aged 60 years or over [31]. To accommodate these definitions of old age, the age criterion for inclusion in this review was set as a mean of 60+ years. Another criterion was the inclusion of participants who were not recruited based on a disease or condition, and this meant excluding studies in which participants were recruited specifically if they had a particular disease or clinical condition. However, studies with frail, inactive older adults, and those with poor balance were included in the review.

Studies with yoga as a control group were excluded from the review (n = 6) [32,33,34,35,36,37]. In these studies, the yoga group was used to control for aspects such as social stimulation and attention from trainers, without producing an aerobic response. The reporting for the controls was not rigorous, and the yoga programmes were not described in detail. Some studies dated back to 1989, making it difficult to procure the necessary data for them.

Search and screening

Database searches were conducted in September 2017. The following databases were searched (from inception till September 2017): Medline, PsycInfo, CINAHL Plus, Scopus, Web of Science, Cochrane Library, Embase, SPORTDiscus, AMED, ProQuest Dissertations & Theses Global. The search was conducted using key words related to “yoga” and “older adults”. A detailed list of the search terms used is presented in the supplementary section (Additional file 1). The outcome and study type criteria were applied at the screening stage. The reference lists of included studies were also checked for additional relevant studies [38].

Screening was carried out in three stages using reference management software (EndNote X7.2.1). First, a preliminary title and abstract screening was performed by one researcher (DiS) where duplicates and obviously irrelevant studies were removed. Five percent of the search results were cross-checked by another researcher (KL). Second, titles and abstracts of all studies were screened by two researchers (DiS, KL) with studies categorised as “Yes” (satisfied eligibility criteria), “No” (did not satisfy eligibility criteria) and “Maybe” (uncertain, and need further scrutiny). Finally, full texts of studies in the “Yes” and “Maybe” categories were screened in further detail by two researchers (DiS, KL). Disagreements were resolved by a third researcher (CF or GB).

Data extraction

A custom data extraction form for descriptive characteristics (Additional file 2) was developed and piloted by three researchers (DiS, GB, CF). Descriptive data were extracted for all included studies by one researcher (DiS), and 33% of these were cross-checked by another researcher (GB or KL). Outcome data were extracted by one researcher (DiS), and 100% cross-checked by another researcher (KL). Discrepancies were resolved through discussions among the researchers (DiS, KL). Authors of studies for which outcome data were not available were contacted and requested to provide the data, and were asked for clarifications if required. One study only reported median, minimum and maximum values for outcome variables [39]. Means and standard deviation were imputed from these data [40,41,42], and the study was included in the meta-analysis.

Quality assessment

Risk of bias was assessed independently by two researchers (DiS, KL) using the Cochrane risk of bias tool [43]. The following domains were assessed for physical function and HRQoL outcomes separately: selection bias (random sequence generation, allocation concealment), detection bias (blinding of outcome assessment), attrition bias (incomplete outcome data), reporting bias (selective reporting), and other bias (sample selection bias [44,45,46,47], contamination bias [45, 46], compliance bias [46] and response bias [48]). Performance bias (blinding of participants and personnel) was not assessed as it is impossible to blind participants and personnel in a yoga intervention study. Under each domain, studies were classified as low, high or unclear risk of bias. Discrepancies were resolved through discussion between the two researchers.

Analysis

For the physical function and HRQoL variables, separate analyses comparing yoga with active and inactive groups were conducted. Other sub-group analyses such as types of yoga and gender were not explored. Though different yoga types have been used in the included studies, there is similarity between types in terms of the structure and postures followed and hence, it was not considered appropriate to compare them. Further, the requisite outcome data were not readily available by gender for a majority of studies.

Vote-counting

As a preliminary analysis, a ‘vote counting’ approach was adopted [49], where three categories were created for each outcome: statistically significant (as reported by authors) positive effects favouring the yoga group, statistically significant negative effects (i.e., favouring the control group), and no significant difference between groups. For every outcome, effects of yoga was based on the category with the highest number of vote counts. For example, for strength, if the majority of studies found significant positive results favouring yoga, then yoga was considered to have a positive effect [49]. Vote-counting has been critiqued as crude and flawed as it does not give due weight to sample size and effect size (ES). However, when used in conjunction with a meta-analysis, the method can provide a comprehensive understanding of the studies and outcomes included, and the effects of the intervention [49]. The vote-counting approach helped create a catalogue of all results from every study included in the systematic review, providing a foundational structure based on which the data for the meta-analysis were generated. The vote counting analysis included all outcomes assessed by more than one study, and the included outcomes are presented in Table 1.

Table 1 List of outcomes included and not included in vote-counting and meta-analysis

Meta-analysis

For outcomes where quantitative data from three or more studies were available, a meta-analysis was conducted using the Comprehensive Meta-Analysis Version 3, Professional software. The outcomes included in the meta-analysis are presented in Table 1. Some studies used more than one test or instrument to measure an outcome. Since only one of these could be included in the meta-analysis, the test most commonly reported by the included studies was chosen. For balance, only functional assessments [50] such as one leg stand test, Berg balance scale, standing balance tests from the Short Physical Performance Battery, and Performance Oriented Mobility Assessment (POMA) were included in the meta-analysis. Objective measures like static and dynamic posturography [50] were not included due to the lack of a composite index and difficulties in interpreting the data. A random effects model was used as it better models data from heterogeneous populations [51]. Data at pre-intervention and immediately following the intervention were analysed, and effect sizes were calculated based on change (post minus pre) scores. Since various different instruments and units were used by studies to measure outcomes, calculation of mean differences was not possible, and standardised mean differences (SMD) were computed instead [51]. Hedges’ g statistic was used to compute effect sizes, and Forest plots were created with 95% confidence intervals (CI). Effect sizes were categorised as small (0.2 to 0.5), moderate (0.5 to 0.8) and large (> 0.8) using Cohen’s categories [52]. Statistical heterogeneity between studies was assessed using the I2 statistic. Heterogeneity was categorised as low (I2 = 0 to 40%), moderate (I2 = 30 to 60%), substantial (I2 = 50 to 90%) and considerable (I2 = 75 to 100%) [51].

One study had two yoga intervention groups and one control group [53]. Both yoga groups were included in the meta-analysis, each one compared with half the number of participants in the control group [54]. Four studies [55,56,57,58] had one yoga intervention group and two control groups. In these cases the result was included twice in the meta-analysis with half the number of participants for the yoga group each time [54]. Following this, two sensitivity analyses were also conducted: (i) comparing the full yoga intervention arm and the first control group, and (ii) comparing the full yoga intervention arm and the second control group. Five included studies [53, 57, 59,60,61] used cluster randomisation, and an iteration of the meta-analysis was run after adjusting the number of participants in the studies to account for this (adjusted sample size = original sample size /design effect, where design effect = 1+ (Average cluster size - 1)*Intracluster Correlation Coefficient; calculations presented in Additional file 3) [54]. Four studies [53, 59,60,61] had the requisite data for cluster randomisation adjustment, and one study [57] was removed from this analysis due to lack of data. There were insufficient studies (less than 10) in the meta-analyses to test for publication bias using funnel plots [62].

Results

7996 records were identified through the data searches, and after the three stages of screening, 27 records from 22 RCTs (Fig. 1) were included in the systematic review. Seventeen RCTs with 967 participants assessed physical function, and 20 RCTs with 1567 participants assessed HRQoL.

Fig. 1
figure 1

Flow of studies through the review. Legend: *Studies could have been excluded for more than one reason

Study and participant characteristics (Table 2)

Ten studies were from the USA [55, 56, 63,64,65,66,67,68,69,70], four from Australia [39, 71,72,73], two from Taiwan [53, 59], two from India [57, 74], and one each from Brazil [75], Iran [76], UK [77] and Portugal [78]. The number of participants in studies ranged from 18 to 410, and the mean size was 77 ± 84.

Table 2 Participant, intervention and study characteristics

The mean age of participants in the studies ranged from 61.0 years to 83.8 years. In 15 studies, more than 70% of the participants were female. The attendance rates for class-based yoga sessions ranged from 67 to 100%, and for active controls it was 62 to 91%. Four studies reported adverse events in the yoga group (groin muscle strain [64], fall during yoga session [71], and musculoskeletal pain [73, 77]). Four studies reported that there were no adverse events during the course of the yoga intervention [53, 59, 68, 69].

Intervention characteristics

Eight types of yoga (Table 3) were used in the studies including Hatha yoga (4 studies) [67,68,69, 76], chair yoga (3 studies) [55, 63, 78], Iyengar yoga (3 studies) [39, 64, 73], Silver Yoga (2 studies) [53, 59], The Easy Does It Yoga Programme (2 studies) [66, 70], balance yoga programme [56], Thai Yoga [72] and the British Wheel of Yoga (BWY) Gentle Years Yoga programme [77]. Five studies did not mention the type of yoga programme conducted [57, 65, 71, 74, 75]. The most common class structure for the yoga intervention adopted by included studies was a warm up, followed by the main postures, and ending with relaxation, breathing and meditation. Some common postures (used in four or more included studies) are: Cat and cow pose, Tree position, Triangle position, Seated twists, Mountain pose, Warrior 1, Cobra, Chair pose, Eagle or Half eagle, Locust posture, Downward dog, Wind relieving pose, Child’s pose, Standing hands on feet pose, Cow face pose, and Corpse pose (used for relaxation). The length of interventions ranged from four to 28 weeks, the most predominant being 24 weeks (6 studies) [53, 57, 59, 63, 64, 74] followed by 12 weeks (5 studies) [56, 72, 73, 75, 77], and eight weeks (4 studies) [39, 67, 68, 76]. The most common frequency of intervention was two sessions per week (9 studies) [39, 56, 65, 67, 69, 71,72,73, 76], followed by three sessions per week (6 studies) [53, 55, 59, 66, 68, 75]. Eight studies also encouraged practicing yoga at home in addition to class based sessions [39, 64, 69, 70, 72,73,74, 77]. Duration of classes ranged from 30 min to 90 min. A 60 min class duration was reported most frequently (9 studies) [56, 57, 65,66,67, 69, 73, 74, 76]. One study did not report class duration [70]. Inactive controls used in the studies were wait-list control (8 studies) [39, 53, 57, 59, 63, 64, 74, 77], playing games like Dominoes, Chinese Checkers and Scrabble [55], fall risk awareness [67], socialisation [65], education on osteoporosis and fitness [69], fall prevention education booklet [73], herbal preparation [57], telephone counselling [72], film series or art class [70], and usual care where no intervention was provided but participants could continue to use the facilities provided by the residential care centre like bingo, story-telling, exercise classes and gym [71]. Active controls included were Tai Chi (3 studies) [56, 71, 72], chair aerobics/exercise (2 studies) [55, 63], a walking programme (2 studies) [55, 64], balance training (2 studies) [56, 67], and stretching–strengthening exercises [68].

Table 3 Types of yoga used in included studies

Results of vote counting

The vote-counting tables with all results for both physical function and HRQoL outcomes are presented in the supplementary section (Additional file 4).

Physical function

For yoga vs inactive controls, the “favouring yoga” category received most votes for the following outcomes (presented as: number of results where yoga had significantly positive effects compared with control / total number of results): cardio-respiratory fitness (2/3), flexibility (17/23, with lower body flexibility (5/7), ROM (10/13), upper body flexibility (2/3)), and walking speed (3/5). On no occasion did the inactive controls group receive more votes than yoga.

While comparing yoga and active controls, the “no significant difference” category got the highest number of votes for all outcomes.

HRQoL

For yoga vs inactive controls, the “favouring yoga” category received most votes for the following outcomes: quality of life (2/3), and sleep quality (3/4). In the yoga vs active controls analysis, the “favouring yoga” category did not receive the highest number of votes for any of the outcomes. The “favouring control” category received no votes for both yoga vs active and yoga vs inactive controls for any HRQoL outcomes.

Meta-analysis

Sixteen studies assessing physical function and 17 assessing HRQoL variables (from 18 records) were included in the meta-analysis (Table 4). Data used for meta-analysis are attached as supplementary tables (Additional file 5).

Table 4 Meta-analysis results- effect sizes and heterogeneity

Physical function

Yoga vs inactive controls

Yoga was found to significantly improve balance (ES (Hedges’ g) = 0.7, 95% CI 0.19 to 1.22, p = 0.01), lower limb strength (ES = 0.45, 95% CI 0.22 to 0.68, p < 0.001), and lower body flexibility (ES = 0.50, 95% CI 0.30 to 0.69, p < 0.001) compared to inactive controls (Fig. 2). No significant difference between yoga and inactive controls was found for body composition (ES = 0.16, 95% CI -0.06 to 0.38, p = 0.16), upper body flexibility (ES = 0.28, 95% CI -0.02 to 0.58, p = 0.07) or walking speed (ES = 0.38, 95% CI -0.02 to 0.78, p = 0.06).

Fig. 2
figure 2

Effect sizes and 95% confidence intervals for yoga compared to inactive controls on physical function outcomes. Legend: Bal: Balance; M: Yoga group with meditation; SSY: Short Silver Yoga; BC: Body composition; LFlex: Lower body flexibility; LST: Lower limb strength; UFlex: Upper body flexibility; WS: Walking speed

Yoga vs active controls

There was a significant effect favouring yoga for lower limb strength (ES = 0.49, 95% CI 0.10 to 0.88, p = 0.01) and lower body flexibility (ES = 0.28, 95% CI 0.01 to 0.54, p = 0.04) (Fig. 3). No significant difference between yoga and active controls was found for balance (ES = 0.32, 95% CI -0.02 to 0.66, p = 0.07), mobility (ES = 0.31, 95% CI -0.25 to 0.87, p = 0.28) or walking speed (ES = − 0.29, 95% CI -0.79 to 0.22, p = 0.26).

Fig. 3
figure 3

Effect sizes and 95% confidence intervals for yoga compared to active controls on physical function outcomes. Legend: Bal: balance; TC: Tai Chi; SBT: Standard balance training; LFlex: Lower body flexibility; LST: Lower limb strength; Mob: Mobility; WS: Walking speed

HRQoL

Yoga vs. inactive controls

There was a significant effect favouring yoga for depression (ES = 0.64, 95% CI 0.32 to 0.95, p < 0.001), perceived mental health (ES = 0.60, 95% CI 0.33 to 0.87, p < 0.001), perceived physical health (ES = 0.61, 95% CI 0.29 to 0.94, p < 0.001), sleep quality (ES = 0.65, 95% CI 0.41 to 0.88, p < 0.001), and vitality (ES = 0.31, 95% 0.03 CI to 0.59, p = 0.03) (Fig. 4). No significant effect was found for fear of falls (ES = 0.39, 95% CI -0.45 to 1.24, p = 0.36) or social health (ES = 0.27, 95% CI -0.15 to 0.69, p = 0.20).

Fig. 4
figure 4

Effect sizes and 95% confidence intervals for yoga compared to inactive controls on HRQoL outcomes. Legend: Dep: Depression; FOF: Fear of falls; Per mental health/PMH: Perceived mental health; Per physical health/PPH: Perceived physical health; Sleep: Sleep quality; Soc: Social health; Vit: Vitality; Ay: Ayurveda (herbal preparation); WL: Wait-list control

Yoga vs. active controls

A significant effect favouring yoga was found for depression (ES = 0.54, 95% CI 0.25 to 0.83, p < 0.001) (Fig. 5). No significant effect was found for anxiety (ES = 0.43, 95% CI -0.03 to 0.88, p = 0.06) and perceived mental health (ES = 0.26, 95% CI -0.03 to 0.55, p = 0.08).

Fig. 5
figure 5

Effect sizes and 95% confidence intervals for yoga compared to active controls on HRQoL outcomes. Legend: Anx: Anxiety; CA: Chair aerobics; W: Walking programme; Dep: Depression; Per mental health/PMH: Perceived mental health

Heterogeneity

Statistically significant heterogeneity was found only when comparing yoga and inactive controls (Table 4). For physical function, significant substantial heterogeneity was found for balance (I2 = 72.15, p = 0.001), and walking speed (I2 = 72.69, p = 0.003). For HRQoL outcomes, statistically significant considerable heterogeneity was found for fear of falls (I2 = 75.64, p = 0.02). Significant substantial heterogeneity was found for depression (I2 = 57.09, p = 0.02), perceived mental health (I2 = 54.87, p = 0.02), and perceived physical health (I2 = 58.55, p = 0.05).

Combining data from different measurement instruments could introduce heterogeneity. For example, significant heterogeneity arose in the comparison of yoga and inactive controls when balance data from one-leg-stand test, Berg balance scale, standing balance tests and POMA were combined. In contrast, when lower body flexibility was measured using a single instrument (sit-and-reach/chair sit-and-reach test) no significant heterogeneity occurred.

Sensitivity analyses and cluster randomisation adjustment

Sensitivity analyses were conducted for four studies which had two controls, introducing one full control arm and then the other (Additional file 6). For one study [56], yoga was compared with active controls, and the sensitivity analysis affected three outcomes (balance, mobility and walking speed). For the second study [55], yoga was compared to active controls affecting two HRQoL outcomes (anxiety and depression). The third [57] and fourth study [58] compared yoga with inactive controls and the sensitivity analysis affected sleep quality [57] and depression [58]. While there were small changes in effect sizes and p values, none of the variables crossed the significance thresholds, and conclusions derived from the original analysis were not altered.

Meta-analysis results were not greatly altered after taking into account cluster randomization (Additional file 7). While there was a small reduction in effect sizes for some outcomes, significance was not affected.

Risk of bias

For physical function, relatively few studies had high risk of bias (selection bias: random sequence generation (6%) and allocation concealment (18%), detection bias (6%), attrition bias (24%), reporting bias (18%) and other bias (41%)) (Fig. 6). Sample selection bias was evident for many studies and a small number were also at risk of contamination bias. Similarly, only few studies assessing HRQoL outcomes had high risk of bias (selection bias: random sequence generation (5%) and allocation concealment (10%), detection bias (10%), attrition bias (24%) and reporting bias (5%)) (Fig. 7). Other bias included response bias which emanates from the use of questionnaires and interviews, including social desirability response, acquiescence response and Hawthorne effect [48]. Since all studies assessing HRQoL used subjective self-report instruments, the risk of other bias is 100% for HRQoL outcomes. Detailed information on sources of bias is provided as supplementary material (Additional file 8).

Fig. 6
figure 6

Risk of bias table and graph for physical function outcomes

Fig. 7
figure 7

Risk of bias table and graph for HRQoL outcomes

Discussion

Summary of main findings

The results of this systematic review demonstrate that compared to inactive controls, it is possible for older adults to improve many aspects of their physical function and HRQoL through participating in a yoga intervention. Findings suggest that small to moderate sized beneficial effects can be achieved for balance, lower body flexibility, lower limb strength, depression, perceived mental health, perceived physical health, sleep quality, and vitality. When yoga was compared with active controls, statistically significant small to moderate effects favouring yoga were found for lower body strength, lower body flexibility and depression. Yoga was found to be as good as the activity undertaken by active controls in improving outcomes such as mobility, walking speed, balance, anxiety and perceived mental health. The yoga group was never significantly worse than the active or inactive group for any of the outcomes. With high attendance rates for class-based sessions, yoga is a feasible intervention that can be recommended to older adults as an activity that improves physical and mental wellbeing.

Comparison to previous literature

While other systematic reviews have included or focused on studies that recruited older adults with clinical conditions, this review is the first to provide a comprehensive overview of the effects of yoga on physical function and HRQoL in an older adult population not characterised by a specific disease or condition. Outcomes such as depression, perceived mental and physical health, balance and mobility have been evaluated by other meta-analysis of RCTs in an older adult population (5), and are described in the section below.

Physical function

Youkhana et al. [10] conducted a systematic review to assess the effects of yoga on balance and mobility. Since the control groups in the review consisted of no intervention, waitlist control/usual care and provision of an education booklet, the study results can be contrasted with the inactive controls groups of the present study. Although the direction of the effect for balance is similar between the two reviews, the effect size in the inactive control group in this study is much higher than in the Youkhana et al. review (Table 5). The meta-analysis for balance in their review included six studies, with three studies in common with the inactive control group. The difference in effect size could be because of the extremely high effect size in one study included only in the current review, in which participants with poor balance were recruited, and saw great benefits from the yoga intervention [76]. Heterogeneity was lower and non-significant in their review for balance compared to the inactive control group in the current study. This could be attributed to more variation in the yoga types, as well as the inclusion of participants with poor balance at base-line in the current review.

Table 5 Comparison of effect sizes from previous reviews and the current review

The two reviews used different tests to assess mobility. Hence, a meta-analysis was conducted for mobility in the review by Youkhana et al. [10], but not in the current review while comparing yoga with inactive controls. The current study assessed mobility using the timed-up-and-go test (Additional file 9), which was measured only by two studies, and hence no meta-analysis was conducted. In the meta-analysis by Youkhana et al. [10], mobility was measured in three studies using the timed-eight-foot-walk, sit-to-stand test and the 4-m-walk. Two of the three studies [64, 73] were also included in the current review, with the sit-to-stand test included under strength and the 4-m-walk included under walking speed (Additional file 9).

HRQoL

Two reviews conducted meta-analyses to assess the effects of yoga on perceived mental and physical health in older adults [11, 13], and found a significant positive effect favouring yoga.

A smaller effect size was found for these outcomes in the Tulloch et al. review [13] compared to the current study (Table 5). The effect size in the present study for perceived physical and mental health in the inactive control group can be compared to HRQoL and mental wellbeing in the meta-analysis by Tulloch et al. correspondingly. The smaller effect size may be attributed to differences in inclusion criteria (studies which specifically recruited clinical populations were excluded in the current study), and only four of the 12 studies in the Tulloch et al. meta-analysis overlapped with the inactive control group of the present study. Some studies included in the current review [39, 65] were not captured by the Tulloch review due to differing search strategies, and search dates. The effect sizes for perceived physical and mental health in the meta-analysis by Patel et al. [11] were comparable to that of the inactive control group in the current study. Their review also assessed depression, and although a moderate effect size was found, it was not significant. The current meta-analysis for depression included more studies and may have the power to detect differences between groups. In line with the results of the current review, another systematic review published in Chinese [9] concluded that yoga significantly reduced depressive symptoms and improved quality of sleep in older adults.

Strengths and limitations

This systematic review and meta-analysis offers a comprehensive view of the effectiveness of yoga on both physical and psychological outcomes. The method of segregating controls into active and inactive groups has not been adopted by any other systematic review for this age group, and is a significant strength of this study. The review provides novel and valuable information on the effects of yoga on some salient outcomes like strength, vitality, and social health in an older adult population. No yoga RCT has directly assessed strength in older adults using techniques like isokinetic dynamometry (gold standard) or hand-held dynamometry [79]. To our knowledge this is the first study to conduct a meta-analysis to comment on the effectiveness of yoga in improving strength albeit using a functional fitness measure as a proxy (sit-to-stand test). The sit-to-stand test is a reliable and valid indicator of lower body strength in older adults [80]. Used in conjunction with measures of flexibility, balance, mobility and walking speed, the sit-to-stand test is a fitting indicator of functional fitness and the ability to perform everyday activities in older adults [81].

This study had a broad search strategy, and criteria other than yoga and older adults were applied only at the screening stage, making it less likely to miss out studies. The review also included dissertations, which were not included in some previous reviews [11], leading to more robust results. However, the authors had difficulties in securing quantitative data for non-significant outcomes for some included studies (selective reporting bias) [53, 59], and these could not be incorporated in the meta-analysis. Consideration of this bias is critical since the primary studies test numerous outcomes, increasing the chance of type 2 errors. The inclusion of articles only published in English can be considered a limitation of the review. However, the review has captured studies from across the world including non-English speaking countries such as India, Taiwan, Brazil, and Iran. Only three studies [71, 73, 77] actually included adverse events as an outcome at the onset of the intervention. While eight studies reported on adverse events in the yoga group, it is not evident if there were no adverse events in the other studies, or if they were not reported. In one study [67] it is not clear if the injuries reported can be attributed to the yoga intervention. Ambiguous or no reporting of adverse events is a deficiency in yoga research, which future studies should address. While only a small proportion of included studies have been rated as high risk of bias, several studies have unclear risk of bias for random sequence generation, allocation concealment and blinding of outcome assessment. Future studies should ensure that randomisation and data collection procedures are reported in detail to allow for accurate assessment of bias and reliability of intervention effects.

The classification of test and instruments into broad physical function and HRQoL categories was carried out in a structured manner, referring to literature when available, to support the decisions made. However, this process can be subjective, and could be the root of differences in effect sizes between reviews (for example, sit-to-stand test was classified as assessing mobility in the Youkhana et al. review [10], but was categorised as evaluating lower limb strength in the present review).

Implications for policy and practice

The study offers clear evidence that compared to no activity, yoga improves physical function and psychological wellbeing in older adults. It can be inferred from the meta-analysis results that yoga improves muscle strength and balance. Previous systematic reviews have highlighted the potential of yoga in improving balance in healthy adults [82], and PA policy should continue to promote yoga within muscle strength and balance guidelines to enhance and maintain health. Approximately 15% of older adults are likely to suffer from a mental health disorder [83], with depression affecting 22% of older men and 28% of older women in the UK [84]. Mental wellbeing is critical for an older adult population, and this review highlights the beneficial effects of yoga in improving perceived physical and mental health, vitality, and alleviating depressive symptoms.

The findings from this review could be used to challenge older adults’ perceptions of yoga. Older adults have the impression that yoga only improves flexibility, and the lack of an aerobic component has been cited as a barrier to yoga participation [85]. The older population and yoga teachers need to be educated on the muscle strength and balance guidelines, and also made aware of the physical function and HRQoL benefits of yoga as evidenced by this study. Information from the studies included in this review (e.g. common yoga postures and class structure) should be shared with yoga teachers. Although not directly examined in relation to effectiveness, the cross tabulation of frequency and duration of class-based sessions (Additional file 10) showed that 60 min on two days a week was the most common, which can be easily translated to practice.

Yoga is a recognised and accepted form of activity in India where it originated. In western countries, although an increasing trend in older adult participation in yoga/pilates has been observed [86,87,88], yoga participation rates still remain low [87, 89, 90]. This review adds to the growing evidence on the benefits of yoga, and researchers should work closely with yoga teachers, studios, fitness centres and policy makers to develop and implement strategies to encourage yoga participation among older adults, tying in with the final aim of increasing participation in muscle strength and balance activities.

Future research

Future intervention studies should include an active control arm, so that conclusions can be drawn with respect to the effectiveness of yoga compared to different exercise programmes. Upper limb strength, hand grip strength, fall frequency, balance confidence, stress and self-efficacy are relevant and important outcomes for this population. The effects of yoga on these outcomes could not be computed through a meta-analysis due to lack of studies, and future research with robust experimental designs should focus on these outcomes. Future systematic reviews for the older adult population should aim to comment on dose-response relationships. The current review assessed the effects of yoga immediately after the intervention, and 28-weeks was the longest follow-up period. Future reviews should assess effects over a longer period, taking into account post-intervention follow-up data. Moreover, this review did not include physiological (e.g. cholesterol, indicators of immune function) and cognitive outcomes (e.g. memory and executive functions) and future reviews could aim to assess these outcomes.

There is a need to develop an appropriate framework for assessing physical function in an older adult population. Health Related Physical Fitness is defined in the American College of Sports Medicine manual as consisting of those specific components of physical fitness that have a relationship with good health, and includes cardio-respiratory fitness, body composition, muscular strength and flexibility [16]. However, it does not include mobility, walking speed, balance and frequency of falls which are important parameters of health for this population. Moreover, clear guidance is needed on the tests and instruments that assess these aspects, with details on whether they are a valid measure of the outcomes assessed. A study may have more than one instrument assessing the same outcome, and there is no standard procedure for choosing which one measure to include in the meta-analysis. This is a potential source of bias, and guidance for this process should be developed to reduce subjectivity.

Conclusion

Results of this systematic review and meta-analysis show that yoga improves multiple physical function and HRQoL outcomes in older adults not characterised by any specific disease or condition. Compared to inactive controls, small to moderate significant effects favouring yoga were found for balance, lower body flexibility, lower limb strength, depression, perceived mental health, perceived physical health, sleep quality, and vitality. When yoga was compared with active controls, significant small to moderate effects were also found for lower body strength, lower body flexibility and depression. Yoga is a multimodal activity that improves muscle strength, balance and flexibility in older adults, and physical activity policy should continue to promote yoga as an activity that enhances physical and mental wellbeing in this population.

Abbreviations

BMI:

Body Mass Index

CI:

Confidence Interval

ES:

Effect Size

HRQoL:

health related quality of life

MS:

Muscle Strength

PA:

Physical Activity

POMA:

Performance Oriented Mobility Assessment

PRISMA:

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

RCT:

Randomised Controlled Trials

ROM:

Range of Motion

SMD:

Standardised Mean Difference

UK:

United Kingdom

US:

United States

References

  1. World Health Organization. Global Recommendations on Physical Activity for Health. https://www.who.int/dietphysicalactivity/physical-activity-recommendations-65years.pdf?ua=1. Accessed 12 March 2018.

  2. Hallal PC, Andersen LB, Bull FC, Guthold R, Haskell W, Ekelund U. Group LPASW: global physical activity levels: surveillance progress, pitfalls, and prospects. Lancet. 2012;380:247–57.

    Article  Google Scholar 

  3. Department of Health. Start Active, Stay Active. A report on physical activity for health from the four home countries' Chief Medical Officers. http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_128209. Accessed 23 March 2018.

  4. National statistics: Statistics on Obesity, Physical Activity and Diet - England: 2017. 2017.

  5. Strain T, Fitzsimons C, Foster C, Mutrie N, Townsend N, Kelly P. Age-related comparisons by sex in the domains of aerobic physical activity for adults in Scotland. Preventive Medicine Reports. 2016;3:90–7.

    Article  Google Scholar 

  6. Strain T, Fitzsimons C, Kelly P, Mutrie N: The forgotten guidelines: cross-sectional analysis of participation in muscle strengthening and balance & co-ordination activities by adults and older adults in Scotland.(Report). BMC Public Health 2016, 16.

  7. Regional Committee for Europe 65th Session: Physical activity strategy for the WHO European Region 2016–2025. World Health Organization Regional Office for Europe; 2015.

  8. Saraswati SS. Asana Pranayama Mudra Bandha. Munger, Bihar, India: Yoga Publications Trust; 2008.

    Google Scholar 

  9. Wang YY, Chang HY, Lin CY. Systematic review of yoga for depression and quality of sleep in the elderly. Journal of Nursing. 2014;61:85–92.

    PubMed  Google Scholar 

  10. Youkhana S, Dean CM, Wolff M, Sherrington C, Tiedemann A. Yoga-based exercise improves balance and mobility in people aged 60 and over: a systematic review and meta-analysis. Age Ageing. 2016;45:21–9.

    Article  Google Scholar 

  11. Patel NK, Newstead AH, Ferrer RL. The effects of yoga on physical functioning and health related quality of life in older adults: a systematic review and meta-analysis. Journal of Alternative & Complementary Medicine. 2012;18:902–17.

    Article  Google Scholar 

  12. Barrows JL, Fleury J. Systematic review of yoga interventions to promote cardiovascular health in older adults. West J Nurs Res. 2015;38:753–81.

    Article  Google Scholar 

  13. Tulloch A, Bombell H, Dean C, Tiedemann A. Yoga- based exercise improves health- related quality of life and mental well-being in older people: a systematic review of randomised controlled trials. Age Ageing. 2018;47:537–44.

    Article  Google Scholar 

  14. Centers for Disease Control and Prevention. Health-Related Quality of Life (HRQoL). http://www.cdc.gov/hrqol/concept.htm. Accessed 23 March 2018.

  15. Rejeski WJ, Mihalko SL. Physical activity and quality of life in older adults. J Gerontol Ser A Biol Med Sci. 2001;56:23–35.

    Article  Google Scholar 

  16. Kaminsky LA: ACSM's health-related physical fitness assessment manual. Third edition / editor, Leonard Kaminsky.. edn. Philadelphia: Philadelphia : Wolters Kluwer Health/Lippincott Williams & Wilkins Health; 2010.

  17. Halter JB, Reuben DB. Indicators of function in the geriatric population. Cells and surveys: Should biological measures be included in social science research. 2000:159–79.

  18. Kelley AG, Kelley SK, Tran VZ. Resistance training and bone mineral density in women: a Meta- analysis of controlled trials. American Journal of Physical Medicine & Rehabilitation. 2001;80:65–77.

    Article  CAS  Google Scholar 

  19. Lang T, Streeper T, Cawthon P, Baldwin K, Taaffe DR, Harris T. Sarcopenia: etiology, clinical consequences, intervention, and assessment. Osteoporos Int. 2010;21:543–59.

    Article  CAS  Google Scholar 

  20. Department of Health. Physical activity benefits infographic for adults and older people. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/541233/Physical_activity_infographic.PDF. Accessed 23 March 2018.

  21. United States Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans. https://health.gov/paguidelines/pdf/paguide.pdf. Accessed 22 July 2018.

  22. Buffart LM, van Uffelen JGZ, Riphagen II, Brug J, van Mechelen W, Brown WJ, Chinapaw MJM. Physical and psychosocial benefits of yoga in cancer patients and survivors, a systematic review and meta- analysis of randomized controlled trials. BMC Cancer. 2012;12:559.

    Article  Google Scholar 

  23. Innes KE, Vincent HK. The influence of yoga-based programs on risk profiles in adults with type 2 diabetes mellitus: a systematic review. Evid Based Complement Alternat Med. 2007;4:469–86.

    Article  Google Scholar 

  24. Aljasir B, Bryson M, Al-shehri B. Yoga practice for the management of type II diabetes mellitus in adults: a systematic review. Evid Based Complement Alternat Med. 2010;7:399–408.

    Article  Google Scholar 

  25. Cramer H, Lauche R, Langhorst J, Dobos G. Yoga for rheumatic diseases: a systematic review. Rheumatology. 2013;52:2025–30.

    Article  Google Scholar 

  26. Donesky-Cuenco D, Nguyen HQ, Paul S, Carrieri-Kohlman V. Yoga therapy decreases dyspnea-related distress and improves functional performance in people with chronic obstructive pulmonary disease: a pilot study. Journal of Alternative & Complementary Medicine. 2009;15:225–34.

    Article  Google Scholar 

  27. Moher D, Liberati A, Tetzlaff J, Altman D. Preferred reporting items for systematic reviews and meta- analyses: the PRISMA statement. Br Med J. 2009;339:332.

    Article  Google Scholar 

  28. Higgins J, Green S, Cochrane C, Higgins JPT, Wiley I. Cochrane handbook for systematic reviews of interventions. Chichester, England. Wiley-Blackwell: Hoboken, N.J; 2008.

    Book  Google Scholar 

  29. Trading Economics. Retirement age men. https://tradingeconomics.com/country-list/retirement-age-men. Accessed 23 March 2018.

  30. Trading Economics. Retirement age women. https://tradingeconomics.com/country-list/retirement-age-women. Accessed 23 March 2018.

  31. United Nations. Department of Economic and Social Affairs. World Population Ageing: Population Division; 2015. http://www.un.org/en/development/desa/population/publications/pdf/ageing/WPA2015_Report.pdf. Accessed 22 July 2018

    Google Scholar 

  32. Blumenthal JA, Emery CF, Madden DJ, George LK, Coleman RE, Riddle MW, McKee DC, Reasoner J, Williams RS. Cardiovascular and behavioral effects of aerobic exercise training in healthy older men and women. Journals of Gerontology. 1989;44:M147–57.

    Article  CAS  Google Scholar 

  33. Blumenthal JA, Emery CF, Madden DJ, Schniebolk S, Riddle MW, Cobb FR, Higginbotham M, Coleman RE. Effects of exercise training on bone density in older men and women. J Am Geriatr Soc. 1991;39:1065–70.

    Article  CAS  Google Scholar 

  34. Bowman AJ, Clayton RH, Murray A, Reed JW, Subhan MMF, Ford GA. Effects of aerobic exercise training and yoga on the baroreflex in healthy elderly persons. Eur J Clin Investig. 1997;27:443–9.

    Article  CAS  Google Scholar 

  35. Emery CF, Blumenthal JA. Perceived change among participants in an exercise program for older adults. Gerontologist. 1990;30:516–21.

    Article  CAS  Google Scholar 

  36. Madden DJ, Blumenthal JA, Allen PA, Emery CF. Improving aerobic capacity in healthy older adults does not necessarily lead to improved cognitive performance. Psychology & Aging. 1989;4:307–20.

    Article  CAS  Google Scholar 

  37. Stachenfeld NS, Mack CW, DiPietro L, Morocco TS, Jozsi AC, Nadel ER. Regulation of blood volume during training in post-menopausal women. Med Sci Sports Exerc. 1998;30:92–8.

    Article  CAS  Google Scholar 

  38. Horsley T, Dingwall O, Tetzlaff J, Sampson M. Checking reference lists to find additional studies for systematic reviews. Cochrane Database Syst Rev. 2009;1.

  39. Vogler J, O'Hara L, Gregg J, Burnell F. The impact of a short-term iyengar yoga program on the health and well-being of physically inactive older adults. International Journal of Yoga Therapy. 2011:61–72.

  40. Walter SD, Yao X. Effect sizes can be calculated for studies reporting ranges for outcome variables in systematic reviews. J Clin Epidemiol. 2007;60:849–52.

    Article  CAS  Google Scholar 

  41. Weir CJ, Butcher I, Assi V, Lewis SC, Murray GD, Langhorne P, Brady MC. Dealing with missing standard deviation and mean values in meta-analysis of continuous outcomes: a systematic review. BMC Med Res Methodol. 2018;18:25.

    Article  Google Scholar 

  42. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol. 2005;5:13.

    Article  Google Scholar 

  43. Higgins JPT, Altman DG: Assessing Risk of Bias in Included Studies. In Cochrane Handbook for Systematic Reviews of Interventions. John Wiley & Sons, Ltd; 2008: 187–241.

  44. Pannucci CJ, Wilkins EG. Identifying and avoiding bias in research. Plast Reconstr Surg. 2010;126:619.

    Article  CAS  Google Scholar 

  45. Kate Ann L: Study design II. Issues of chance, bias, confounding and contamination. Evidence-Based Dentistry 2005, 6:102.

  46. Krishna R, Maithreyi R, Surapaneni K. Research bias: a review for medical students. J Clin Diagn Res. 2010;4:2320–4.

    Google Scholar 

  47. Henderson M, Page L. Appraising the evidence: what is selection bias? Evidence Based Mental Health. 2007;10:67.

    Article  Google Scholar 

  48. Furnham A. Response bias, social desirability and dissimulation. Personal Individ Differ. 1986;7:385–400.

    Article  Google Scholar 

  49. Bushman B, Wang MC: Vote-counting procedures in meta-analysis. In The handbook of research synthesis and meta-analysis. Edited by Cooper HM, Hedges LV, Valentine JC; 2009: 207–220.

  50. Mancini M, Horak FB. The relevance of clinical balance assessment tools to differentiate balance deficits. European journal of physical and rehabilitation medicine. 2010;46:239.

    CAS  PubMed  PubMed Central  Google Scholar 

  51. Deeks JJ, Higgins JP, Altman DG: Analysing Data and Undertaking Meta-Analyses. In Cochrane Handbook for Systematic Reviews of Interventions. Chichester, England: John Wiley & Sons, Ltd; 2008.[Higgins JP, Green S (Series Editor).

  52. Cohen J: Statistical power analysis for the behavioral sciences. Second edition.. edn. Hillsdale, N.J. : Lawrence Erlbaum; 1988.

  53. Chen KM, Chen MH, Hong SM, Chao HC, Lin HS, Li CH. Physical fitness of older adults in senior activity centres after 24-week silver yoga exercises. J Clin Nurs. 2008;17:2634–46.

    Article  Google Scholar 

  54. Higgins JP, Deeks JJ, Altman DG: Special Topics in Statistics. In Cochrane Handbook for Systematic Reviews of Interventions. Edited by P. HJ, S. G. Chichester, UK.: John Wiley & Sons, Ltd; 2008.

  55. Bethany KH. The effects of selected exercise modalities on stress, anxiety, and depression responses in the elderly: M.S. The Florida State University; 2005.

  56. Ni M, Mooney K, Balachandran A, Signorile JF. Comparative impacts of tai chi, balance training, and a specially-designed yoga program on balance in older fallers. Arch Phys Med Rehabil. 2014;95:1620–8.

    Article  Google Scholar 

  57. Manjunath NK, Telles S. Influence of yoga and Ayurveda on self-rated sleep in a geriatric population. Indian J Med Res. 2005;121:683–90.

    CAS  PubMed  Google Scholar 

  58. Krishnamurthy MN, Telles S. Assessing depression following two ancient Indian interventions: effects of yoga and ayurveda on older adults in a residential home. J Gerontol Nurs. 2007;33:17–23.

    PubMed  Google Scholar 

  59. Chen KM, Fan JT, Wang HH, Wu SJ, Li CH, Lin HS. Silver yoga exercises improved physical fitness of transitional frail elders. Nurs Res. 2010;59:364–70.

    Article  Google Scholar 

  60. Chen KM, Chen MH, Lin MH, Fan JT, Lin HS, Li CH. Effects of yoga on sleep quality and depression in elders in assisted living facilities. J Nurs Res. 2010;18:53–61.

    Article  CAS  Google Scholar 

  61. Chen KM, Chen MH, Chao HC, Hung HM, Lin HS, Li CH. Sleep quality, depression state, and health status of older adults after silver yoga exercises: cluster randomized trial. Int J Nurs Stud. 2009;46:154–63.

    Article  Google Scholar 

  62. Sterne JA, Egger M, Moher D: Addressing reporting biases. In Cochrane handbook for systematic reviews of interventions. Edited by Green SH, Julian. Chichester, England: Hoboken, NJ : Wiley-Blackwell; 2008.

  63. Bonura KB, Tenenbaum G. Effects of yoga on psychological health in older adults. J Phys Act Health. 2014;11:1334–41.

    Article  Google Scholar 

  64. Oken BS, Zajdel D, Kishiyama S, Flegal K, Dehen C, Haas M, Kraemer DF, Lawrence J, Leyva J. Randomized, controlled, six-month trial of yoga in healthy seniors: effects on cognition and quality of life. Altern Ther Health Med. 2006;12:40–7.

    PubMed  PubMed Central  Google Scholar 

  65. Wang D. Feasibility of a yoga intervention for enhancing the mental well-being, and physical functioning of older adults living in the community. Activities, Adaptation and Aging. 2010;34:85–97.

    CAS  Google Scholar 

  66. Haber D. A health promotion program in ten nursing homes. Activities, Adaptation and Aging. 1988;11:75–84.

    Google Scholar 

  67. Morris DM. An evaluation of yoga for the reduction of fall risk factors in older adults. Ph.D: The Florida State University; 2008.

    Google Scholar 

  68. Gothe NP. The effects of an 8-week yoga intervention on cognition and functional fitness in older adults: Ph.D. University of Illinois at Urbana-Champaign; 2013.

  69. Leininger P. Physical and psychological effects of yoga exercise on healthy community-dwelling older adult women. Ph.D: Temple University; 2006.

    Google Scholar 

  70. Haber D. Yoga as a preventive health care program for white and black elders: an exploratory study. Int J Aging Hum Dev. 1983;17:169–76.

    Article  CAS  Google Scholar 

  71. Saravanakumar P, Higgins IJ, Van Der Riet PJ, Marquez J, Sibbritt D. The influence of tai chi and yoga on balance and falls in a residential care setting: a randomised controlled trial. Contemp Nurse. 2014;48:76–87.

    Article  Google Scholar 

  72. Noradechanunt C, Worsley A, Groeller H. Thai yoga improves physical function and well-being in older adults: a randomised controlled trial. J Sci Med Sport. 2017;20:494–501.

    Article  Google Scholar 

  73. Tiedemann A, O'Rourke S, Sesto R, Sherrington C. A 12-week Iyengar yoga program improved balance and mobility in older community-dwelling people: a pilot randomized controlled trial. Journals of Gerontology Series A-Biological Sciences & Medical Sciences. 2013;68:1068–75.

    Article  Google Scholar 

  74. Hariprasad VR, Sivakumar PT, Koparde V, Varambally S, Thirthalli J, Varghese M, Basavaraddi IV, Gangadhar BN. Effects of yoga intervention on sleep and quality-of-life in elderly: a randomized controlled trial. Indian J Psychiatry. 2013;55(Suppl 3):S364–8.

    CAS  PubMed  PubMed Central  Google Scholar 

  75. Bezerra LA, de Melo HF, Garay AP, Reis VM, Aidar FJ, Bodas AR, Garrido ND, de Oliveira RJ. Do 12-week yoga program influence respiratory function of elderly women? Journal of Human Kinetics. 2014;43:177–84.

    Article  Google Scholar 

  76. Nick N, Petramfar P, Ghodsbin F, Keshavarzi S, Jahanbin I. The effect of yoga on balance and fear of falling in older adults. PM&R. 2016;8:145–51.

    Article  Google Scholar 

  77. Tew GA, Howsam J, Hardy M, Bissell L. Adapted yoga to improve physical function and health-related quality of life in physically-inactive older adults: a randomised controlled pilot trial. BMC Geriatr. 2017;17.

  78. Marques M, Chupel MU, Furtado GE, Minuzzi LG, Rosado F, Pedrosa F, Ferreira JP, Teixeira AM. Influence of chair-based yoga on salivary anti-microbial proteins, functional fitness, perceived stress and well-being in older women: a randomized pilot controlled trial. European Journal of Integrative Medicine. 2017;12:44–52.

    Article  Google Scholar 

  79. Stark T, Walker B, Phillips JK, Fejer R, Beck R. Hand-held dynamometry correlation with the gold standard isokinetic dynamometry: a systematic review. PM&R. 2011;3:472–9.

    Article  Google Scholar 

  80. Jones CJ, Rikli RE, Beam WC. A 30- s chair- stand test as a measure of lower body strength in community- residing older adults. Res Q Exerc Sport. 1999;70:113–9.

    Article  CAS  Google Scholar 

  81. Rikli RE, Jones CJ. Development and validation of a functional fitness test for community-residing older adults. J Aging Phys Act. 1999;7:129–61.

    Article  Google Scholar 

  82. Jeter PE, Nkodo AF, Moonaz SH, Dagnelie G. A systematic review of yoga for balance in a healthy population. J Altern Complement Med. 2014;20:221–32.

    Article  Google Scholar 

  83. World Health Organization. Mental health of older adults. https://www.who.int/news-room/fact-sheets/detail/mental-health-of-older-adults. Accessed 28 January 2019.

  84. Craig R, Mindell J. Health survey for England 2005: the health of older people. Joint Health Surveys Unit: National Centre for Social Research, Department of Epidemiology and Public Health at the Royal Free, University College Medical School; 2007.

    Google Scholar 

  85. Sivaramakrishnan D, Fitzsimons C, Mutrie N, Baker G. Perceptions of yoga among older adults: a qualitative approach. Annals of Yoga and Physical Therapy. 2017;2.

  86. Vergeer I, Bennie JA, Charity MJ, Harvey JT, van Uffelen JGZ, Biddle SJH, Eime RM. Participation trends in holistic movement practices: a 10-year comparison of yoga/Pilates and t'ai chi/qigong use among a national sample of 195,926 Australians. BMC Complement Altern Med. 2017;17.

  87. Gill V. Chapter 5: physical activity. In The Scottish Health Survey 2014: Volume 1: Main Report. Edited by Campbell-Jack D, Hinchliffe S, Bromley C. The Scottish Government: Edinburgh; 2015.

  88. Currie E. Chapter 3: physical activity. In Scottish Health Survey 2016: Volume 1: Main Report. Edited by McLean J, Christie S, gray L. The Scottish Government: Edinburgh; 2017.

  89. Cramer H, Ward L, Steel A, Lauche R, Dobos G, Zhang Y. Prevalence, patterns, and predictors of yoga use results of a US nationally representative survey. Am J Prev Med. 2016;50:230–5.

    Article  Google Scholar 

  90. Ding D, Stamatakis E. Yoga practice in England 1997-2008: prevalence, temporal trends, and correlates of participation. BMC Research Notes. 2014;7:172.

    Article  Google Scholar 

  91. Gothe NP, McAuley E. Yoga is as good as stretching-strengthening exercises in improving functional fitness outcomes: results from a randomized controlled trial. Journals of Gerontology Series a-Biological Sciences and Medical Sciences. 2016;71:406–11.

    Article  Google Scholar 

  92. Krishnamurthy M, Telles S: Effects of Yoga and an Ayurveda preparation on gait, balance and mobility in older persons. Med Sci Monit 2007, 13:Le19–20.

  93. 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 Evidence-based Complementary and Alternative Medicine 2013, 2013.

  94. Chen KM, Tsen WS, Ting LF, Huang GF. Development and evaluation of a yoga exercise programme for older adults. J Adv Nurs. 2007;57:432–41.

    Article  Google Scholar 

Download references

Acknowledgements

Authors would like to thank Marshal Dozier who helped with database searching, and Dr. Anne Martin for her inputs on methodology and analysis.

Funding

DiS is funded by a College PhD Scholarship (University of Edinburgh).

Availability of data and materials

Data and results for the main meta-analyses presented in this study are included in this published article (and its supplementary information files). Results for sensitivity analysis and cluster randomisation adjustment analysis are included as supplementary files. The datasets for these analyses are available from the corresponding author on request.

Author information

Authors and Affiliations

Authors

Contributions

DiS, CF, GB, NM conceived the study. DiS, CF, GB, NM established eligibility criteria. DiS, KL, GB, CF worked on study selection, data extraction and quality assessment. DiS, CF, GB, PK and DaS contributed to analysis. PK and DaS provided input and support at all stages of the project. DiS drafted the manuscript. All authors contributed to manuscript development and read and approved the final draft.

Corresponding author

Correspondence to Divya Sivaramakrishnan.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Additional files

Additional file 1:

Search terms for Ovid databases (MEDLINE, PsycInfo, EMBASE, AMED). Detailed list of search terms used for OVID databases are provided in this file. (PDF 37 kb)

Additional file 2:

Data extraction template. This is the custom data extraction template used in the study. (XLSX 12 kb)

Additional file 3:

The formula used in the study for calculating the sample size after adjusting for cluster randomisation is provided in this file. The studies included in this analysis are also listed in this document. (PDF 1175 kb)

Additional file 4:

Vote count tables for physical function and HRQoL outcomes. The tables contain columns for outcome, study name, tests and instrument used, intervention and controls, and whether there were significant effects. (PDF 412 kb)

Additional file 5:

(i) Data for meta-analysis- yoga compared with inactive controls for physical function outcomes. (ii) Data for meta-analysis- yoga compared with active controls for physical function outcomes. (iii) Data for meta-analysis- yoga compared with inactive controls for HRQoL outcomes. (iv) Data for meta-analysis- yoga compared with active controls for HRQoL outcomes. This excel file has four tabs which contain the data used in the main meta-analyses analyses (physical function and HRQoL outcomes for yoga vs inactive controls and yoga vs active controls) presented in this review. (XLSX 29 kb)

Additional file 6:

Sensitivity analysis results. Results of the sensitivity analysis conducted are provided in this document. This includes Forest plots and homogeneity data. (PDF 307 kb)

Additional file 7:

Cluster randomisation adjustment results- Forest plots and homogeneity data. Results of the cluster randomisation adjustment analysis are provided in this document. This includes Forest plots and homogeneity data. (PDF 29 kb)

Additional file 8:

(i) Risk of bias details for physical function outcomes. (ii) Risk of bias details for HRQoL outcomes. This excel file has two tabs with the risk of bias details for physical function and HRQoL outcomes. (XLSX 25 kb)

Additional file 9:

This table consists of the tests/instrument used to measure each outcome for all studies included in the meta-analysis (XLSX 13 kb)

Additional file 10:

Frequency and duration of yoga sessions from studies included in the systematic review. This is a cross-tab of frequency and duration of yoga interventions from studies included in the meta-analysis. (XLSX 9 kb)

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sivaramakrishnan, D., Fitzsimons, C., Kelly, P. et al. 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 (2019). https://doi.org/10.1186/s12966-019-0789-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12966-019-0789-2

Keywords