Skip to content
HabitSeek
Back to results

Yoga practice

Practising physical yoga postures, typically in a class or guided session, as a structured form of mind-body exercise.

At a glance

  • Evidence: Yoga reduces back pain and improves back-related function compared with doing nothing, but the improvement is too small to be clinically important and disappears when the comparison is another active exercise.Chronic non-specific low back pain (pain intensity and back-related function) · Moderate certainty · Randomized intervention
  • What’s uncertain: Trials rely on self-reports and cannot blind participants; improvements over inactive controls may not persist against active exercise.
  • First step: Pick a form you can repeat: the trials mostly used hatha, iyengar or viniyoga, delivered as classes of roughly 60 minutes, once or twice a week, for 8 to 12 weeks.
  • Before you try it: Back-pain trials reported more adverse events than no exercise, mainly increased pain; start gently and stop postures that worsen pain. All precautions
Read the full takeaway

Outcome findings

Yoga helps a little with chronic low back pain, anxiety and self-reported sleep, but the back-pain effects are small and do not survive comparison with another back-focused exercise. The anxiety and sleep findings come mainly from comparisons with inactive controls, and one 2026 meta-analysis found yoga better than active exercise for physical function and emotional wellbeing, on two studies for function and with I2 = 95% for pain, so that advantage is preliminary rather than settled.

Comparison with other exercise

The best synthesis in back pain is a Cochrane review that found improvement against no exercise which it judged clinically unimportant, and no difference from other back-focused exercise; the same pattern holds for anxiety and for balance, where yoga beat inactive controls but the improvements were modest and have not been shown to prevent falls.

Adverse events and clinical limits

Yoga also carried more reported adverse events than no exercise in the back pain trials, mostly increased back pain. Nothing here supports yoga as a general health intervention, a treatment for a diagnosed anxiety disorder, or a substitute for physiotherapy, and the impossibility of blinding participants inflates every estimate that does exist.

A woman doing yoga on a mat in a park
ExerciseEnergyStress & wellbeing20+ min · Moderate · Moderate
Reviewed Sep 23, 2026
Save

Scientific evidence

What the research shows

Outcome assessments based on research, including its uncertainties and limitations.

Chronic non-specific low back pain (pain intensity and back-related function)

Finding: Mixed evidenceCertainty: Moderatestudies disagree, or the effect depends on who is studied

Evidence basis: randomised trials

Yoga reduces back pain and improves back-related function compared with doing nothing, but the improvement is too small to be clinically important and disappears when the comparison is another active exercise.

Population: Adults with chronic non-specific low back pain, mostly women in their 40s and 50s, recruited from clinical and community settings in the USA, India, the UK, Croatia, Germany, Sweden and Turkey (21 trials in the Cochrane review); 27 individual studies with 2702 participants in the second meta-analysis; 13 systematic reviews in the overview.

Research details

Summary

Yoga reduces back pain and improves back-related function compared with doing nothing, but the improvement is too small to be clinically important and disappears when the comparison is another active exercise. The Cochrane review of 21 trials (2223 participants) found a mean difference of -4.53 points on a 0 to 100 pain scale against an MCID of 15 points (moderate-certainty evidence) and -1.69 points on the 0 to 24 Roland-Morris Disability Questionnaire against an MCID of 5 points (low-certainty evidence) compared with no exercise at three months, and reported little or no difference from back-focused exercise. A separate meta-analysis of 27 studies (2702 participants) found short-term pain improvement against passive control (MD -0.74 on a numeric rating scale) with no significant difference against an active comparator in either the short or long term. The same review group's overview of 13 systematic reviews rated most pooled outcomes low or very low certainty.

Why this certainty

Rated Moderate: the Cochrane review is the most rigorous synthesis available, uses GRADE, and its conclusion of small, clinically unimportant benefit against no exercise plus no difference against other exercise is echoed by an independent meta-analysis and by an overview of 13 systematic reviews. It is not rated higher because all trials are unblinded and at high risk of performance and detection bias, several pooled outcomes are low or very low certainty, and the improvement falls well below the published clinically important thresholds on both pain and function.

Limitations

Every included trial was at high risk of performance and detection bias because participants and providers cannot be blinded and outcomes are self-assessed, and the Cochrane authors state that blinded comparisons would be unlikely to find a clinically important benefit. No trial compared yoga with sham yoga. The overview of reviews found the methodological quality of the contributing systematic reviews to be relatively low (AMSTAR-2), with GRADE ratings of moderate, low and very low for 30, 42 and 20 outcomes respectively, and heterogeneity is high in several pooled estimates (Li 2026, the 2026 yoga-versus-exercise synthesis, reports I2 = 95% for pain and 92% for disability, although it also found yoga better than active exercise for physical function on two studies and for emotional wellbeing).

Effect estimate

Versus no exercise at three months: pain MD -4.53 (95% CI -6.61 to -2.46) on a 0 to 100 scale, MCID 15 points, moderate certainty; back-specific function MD -1.69 (95% CI -2.73 to -0.65) on the 0 to 24 RMDQ, MCID 5 points, low certainty. Versus back-focused exercise at three months: function MD -0.38 (95% CI -1.33 to 0.62) on the RMDQ and pain MD 2.68 (95% CI -2.01 to 7.36), little or no difference. Versus passive control (Anheyer 2022): pain MD -0.74 points (95% CI -1.04 to -0.44; SMD -0.37, 95% CI -0.52 to -0.22) and pain-related disability MD -2.28 (95% CI -3.30 to -1.26; SMD -0.38, 95% CI -0.55 to -0.21), with no significant difference against an active comparator. Adverse events versus no exercise at six to 12 months: RR 4.76 (95% CI 2.08 to 10.89; 43 per 1000 with yoga versus 9 per 1000 with no exercise, low certainty); versus other exercise: RR 0.93 (95% CI 0.56 to 1.53).

Sources for this outcome:

Anxiety symptoms and perceived stress

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: randomised trials

Meta-analyses agree that yoga reduces anxiety and perceived stress, with small-to-moderate effects that are largest in people with elevated but undiagnosed anxiety.

Population: Adults with elevated anxiety symptoms or a formal anxiety disorder (8 RCTs, 319 participants, mean age 30.0 to 38.5 years, in the first meta-analysis); 36 studies of adults measuring perceived stress with the Perceived Stress Scale; 23 randomised trials in adults with diagnosed anxiety or depressive disorders; 42 trials in the physiological meta-analysis, which covered mixed populations including people with cardiac and other physical conditions.

Research details

Summary

Meta-analyses agree that yoga reduces anxiety and perceived stress, with small-to-moderate effects that are largest in people with elevated but undiagnosed anxiety. A meta-analysis of 8 randomised trials (319 participants) found a small effect on anxiety versus no treatment (SMD -0.43, 95% CI -0.74 to -0.11) and a large effect versus active comparators (SMD -0.86, 95% CI -1.56 to -0.15), but no effect in participants whose anxiety disorder was diagnosed by DSM criteria. A meta-analysis of 36 studies found a moderate effect on Perceived Stress Scale scores (g = 0.48, 95% CI 0.29 to 0.66). A larger review of mind-body exercise trials in diagnosed anxiety and depressive disorders (23 studies, 22 samples, n = 1420) found yoga superior to controls for anxiety symptoms but judged the certainty of that evidence very low.

Why this certainty

Rated Limited: several independent meta-analyses agree on direction and statistical significance, but the anxiety evidence base is small, the more rigorous synthesis in diagnosed disorders was graded very low certainty, the benefits are not seen in DSM-diagnosed anxiety, and blinding is impossible so expectancy and attention effects cannot be excluded. The physiological stress-marker meta-analysis supports a plausible mechanism but reported heterogeneous interventions and no pooled effect sizes in the abstract retrieved.

Limitations

The anxiety meta-analysis is small (8 trials, 319 participants), had unclear risk of selection bias in most trials, and found no effect where anxiety was diagnosed by DSM criteria, so the positive result rests on people without a formal diagnosis. The larger review of diagnosed disorders rated the certainty of all outcomes very low because of high risk of bias, indirectness, inconsistency and imprecision, and noted marked heterogeneity in the yoga styles and self-report instruments used. The stress meta-analysis reported substantial heterogeneity (Q = 117.33, p < .001) with 74.90% of the variation in effect sizes attributed to study characteristics, and its planned moderator hypotheses about breathwork, relaxation, hours practised and country were not supported. Passive and active comparators, scales and session doses all differ across trials, and participants cannot be blinded.

Effect estimate

Anxiety after yoga versus no treatment: SMD -0.43 (95% CI -0.74 to -0.11, P = .008); versus active comparators: SMD -0.86 (95% CI -1.56 to -0.15, P = .02); depression versus no treatment: SMD -0.35 (95% CI -0.66 to -0.04, P = .03). Perceived stress (Perceived Stress Scale, 36 studies): g = 0.48 (95% CI 0.29 to 0.66), Q = 117.33, p < .001. Physiological measures across 42 trials: reduced evening cortisol, waking cortisol, ambulatory systolic blood pressure, resting heart rate, fasting blood glucose, cholesterol and low-density lipoprotein compared with active control, and reduced high-frequency heart rate variability, which reflects lower vagal tone and is not a favourable direction (the wording is the source's; no pooled numeric estimates are available in the abstract retrieved).

Sources for this outcome:

Sleep quality in women with sleep problems

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: randomised trials

Yoga improves self-reported sleep quality in women with sleep problems compared with inactive controls, but the pooled effect is small and does not extend to insomnia severity or to every subgroup.

Population: Women with sleep problems (19 studies, 1832 participants) including subgroups with breast cancer and peri- or postmenopausal women; separately, 140 adults aged 18 to 64 with chronic low back pain in a two-arm randomised trial.

Research details

Summary

Yoga improves self-reported sleep quality in women with sleep problems compared with inactive controls, but the pooled effect is small and does not extend to insomnia severity or to every subgroup. A meta-analysis of 19 studies including 1832 participants found a significant improvement in sleep (SMD -0.327, 95% CI -0.506 to -0.148), driven by the 16 trials using the Pittsburgh Sleep Quality Index (SMD -0.54, 95% CI -0.89 to -0.19). There was no effect on the Insomnia Severity Index (SMD -0.13, 95% CI -0.74 to 0.48), no effect in women with breast cancer, and no effect in peri- or postmenopausal women. In one randomised trial in adults with chronic low back pain, a sleep-disturbance score improved by 0.4 points on a 0 to 4 scale at both 12 and 24 weeks relative to a wait-list control.

Why this certainty

Rated Limited: the direction is consistent across a meta-analysis and a randomised trial, and the meta-analysis measured safety with no serious adverse events reported, but effects are small, one subjective instrument carries the result, key subgroups are null, no trial used an active control or blinded assessment, and the only randomised-trial evidence retrieved in a chronic pain population used a secondary single-item outcome.

Limitations

Almost all sleep measurement is by self-report questionnaire, and participants cannot be blinded, so expectancy effects are unavoidable. The pooled improvement is small and the confidence interval for the overall estimate is narrow only because of the number of small trials. The three subgroups that showed no benefit (insomnia severity, breast cancer, peri- and postmenopausal women) mean the positive result cannot be generalised to the women most likely to seek help for sleep. The back pain trial's sleep outcome was a secondary or exploratory single-item PROMIS measure in a population recruited for back pain rather than for sleep problems, and its comparator was a wait list rather than an active intervention.

Effect estimate

Sleep quality (any instrument): SMD -0.327 (95% CI -0.506 to -0.148, P < 0.001) across 19 studies, 1832 participants. Pittsburgh Sleep Quality Index in 16 RCTs: SMD -0.54 (95% CI -0.89 to -0.19, P = 0.003). Insomnia Severity Index in 3 RCTs: SMD -0.13 (95% CI -0.74 to 0.48, P = 0.69). Breast cancer subgroup (7 RCTs): SMD -0.15 (95% CI -0.31 to 0.01, P = 0.5). Peri- or postmenopausal subgroup (4 RCTs): SMD -0.31 (95% CI -0.95 to 0.33, P = 0.34). Chronic low back pain trial: PROMIS sleep quality mean change 0.4 points (95% CI 0.1 to 0.7) at 12 weeks (P = .008) and 0.4 points (95% CI 0.1 to 0.7) at 24 weeks (P = .005) versus wait list.

Sources for this outcome:

Balance and physical mobility in older adults

Finding: Benefit supportedCertainty: Limitedstudies found a benefit from doing this

Evidence basis: randomised trials

Yoga-based exercise produces small improvements in balance and moderate improvements in physical mobility in people aged 60 and over, but no retrieved trial shows that this reduces falls.

Population: Community-dwelling people aged 60 years and over (6 trials, 307 participants); trials were included only if they evaluated physical yoga, excluding meditation and breathing exercises alone.

Research details

Summary

Yoga-based exercise produces small improvements in balance and moderate improvements in physical mobility in people aged 60 and over, but no retrieved trial shows that this reduces falls. A systematic review and meta-analysis of six trials of relatively high methodological quality (307 participants) found a small effect on balance performance (Hedges g = 0.40, 95% CI 0.15 to 0.65) and a medium effect on physical mobility (Hedges g = 0.50, 95% CI 0.06 to 0.95), and its authors state that further research is required to determine whether these improvements translate to prevention of falls.

Why this certainty

Rated Limited: the pooled estimates are statistically significant and come from trials the reviewers judged relatively high quality on the PEDro scale, but the evidence base is small and dated, the effect on balance is small, the mobility estimate is imprecise, and the clinically important outcome of falls was not assessed. This is an intermediate performance outcome and is reported separately from the pain and mental health outcomes rather than being presented as a general physical-function benefit.

Limitations

Only six trials with 307 participants contributed, the mobility estimate rests on three trials and has a confidence interval that approaches no effect, and the review explicitly could not determine whether the gains prevent falls. The outcome is a performance measure rather than a fall or fracture, so it is a surrogate for the outcome that matters. The evidence was searched to February 2015, so no newer trial is reflected, and intervention content and dose varied across the included trials.

Effect estimate

Balance performance: Hedges g = 0.40 (95% CI 0.15 to 0.65) across 6 trials, 307 participants. Physical mobility: Hedges g = 0.50 (95% CI 0.06 to 0.95) across 3 trials. No falls outcome was pooled or reported.

Sources for this outcome:

What people say

Who discusses this?

Public statements, with positions specific to each source and topic. These views do not determine the scientific evidence assessment.

What the research shows ↑

No sourced public statements added for this habit yet. This does not mean no one discusses it.

Practical steps and precautions

How to try it

  1. Pick a form you can repeat: the trials mostly used hatha, iyengar or viniyoga, delivered as classes of roughly 60 minutes, once or twice a week, for 8 to 12 weeks. That schedule is an editorial starting point drawn from the trial protocols, not a validated dose.
  2. The one randomised trial that tested virtual delivery compared live-streamed classes with a wait list, not with in-person classes; its authors note that their effect sizes were at least as large as earlier in-person trials, and that improvement occurred even though only 36.6% of participants attended at least half the classes. Format has not been compared head-to-head, so choose the one you will actually use.
  3. In the low back pain trials, benefit appeared within the first weeks and persisted for several months in the reviews that followed people up, so treat the first month as the point at which to judge whether it is worth continuing.
  4. If you are using yoga to move more, pair it with the aerobic and strength work you would otherwise do; there is no evidence that yoga replaces other exercise, only that it matches it in back pain at three months.

Precautions

  • In a Cochrane review of 21 back pain trials, yoga was associated with more adverse events than no exercise at six to 12 months, mainly increased back pain; start gently and stop a posture that worsens pain rather than working through it.
  • Yoga was not better than physiotherapy or other back-focused exercise in any head-to-head comparison retrieved, so it is not a substitute for treatment that has been prescribed for you.
  • In the largest meta-analysis, no effect was found in people whose anxiety disorder was diagnosed by DSM criteria (Cramer 2018); a later review found yoga superior to controls in diagnosed anxiety disorders but graded all outcomes very low certainty (Martinez-Calderon 2023), so yoga is not an established treatment for a diagnosed anxiety or depressive disorder.
  • Avoid or heavily modify inverted, deep-twist and extreme spinal positions if you have uncontrolled hypertension, glaucoma, osteoporosis, a recent injury or surgery, or are pregnant, and take advice from a clinician or an experienced teacher first.

References

Yoga for chronic non-specific low back pain

Wieland LS, Skoetz N, Pilkington K, Harbin S, Vempati R, Berman BM · Cochrane Database of Systematic Reviews · 2022

Cochrane systematic review and meta-analysis of randomised controlled trials with GRADE certainty assessment (21 trials, 2223 participants; search to 31 August 2021)Verified from abstract
Open original source

Yoga for treating low back pain: a systematic review and meta-analysis

Anheyer D, Haller H, Lauche R, Dobos G, Cramer H · Pain · 2022

Systematic review and meta-analysis of randomised controlled trials comparing yoga with passive control or an active comparator, with Cochrane risk of bias assessment; search to 26 May 2020Verified from abstract
Open original source

Efficacy and safety of yoga for the management of chronic low back pain: an overview of systematic reviews

Zhang X, Chang T, Hu W, Shi M, Chai Y, Wang S, Zhou G, Han M, Zhuang M, Yu J, Yin H, Zhu L, Zhao C, Li Z, Liao X · Frontiers in Neurology · 2023

Overview of systematic reviews with AMSTAR-2 methodological appraisal, PRISMA-2020 reporting appraisal and GRADE certainty grading; nine databases searched to September 2023Verified from abstract
Open original source

Effectiveness of Virtual Yoga for Chronic Low Back Pain: A Randomized Clinical Trial

Tankha H, Gaskins D, Shallcross A, Rothberg M, Hu B, Guo N, Roseen EJ, Dombrowski S, Bar J, Warren R, Wilgus H, Tate P, Goldfarb J, Drago VG, Saper R · JAMA Network Open · 2024

Single-blinded, 24-week, two-arm randomised clinical trial (live-streamed yoga versus wait-list control) with intention-to-treat analysis; NCT05319691Verified from abstract
Open original source

Is yoga more beneficial than exercise-based interventions for patients with chronic low back pain? A systematic review and meta-analysis.

Li T, Wen Q, Zhang L · Frontiers in medicine · 2026

Systematic review and meta-analysis of randomised controlled trialsVerified from abstract
Open original source

Yoga for anxiety: A systematic review and meta-analysis of randomized controlled trials

Cramer H, Lauche R, Anheyer D, Pilkington K, de Manincor M, Dobos G, Ward L · Depression and Anxiety · 2018

Systematic review and meta-analysis of randomised controlled trials of yoga for anxiety disorders or elevated anxiety, with Cochrane risk of bias assessment; searches to October 2016Verified from abstract
Open original source

Yoga as an intervention for stress: a meta-analysis

Rhoads MC, Barber ME, Grevstad N, Kirkland RA, Myers S, Gruidel KA, Greenwood E · Health Psychology Review · 2025

Meta-analysis of 36 studies measuring yoga's effect on perceived stress with the Perceived Stress Scale, with planned and exploratory moderator analysesVerified from abstract
Open original source

Yoga-based interventions may reduce anxiety symptoms in anxiety disorders and depression symptoms in depressive disorders: a systematic review with meta-analysis and meta-regression

Martinez-Calderon J, Casuso-Holgado MJ, Munoz-Fernandez MJ, Garcia-Munoz C, Heredia-Rizo AM · British Journal of Sports Medicine · 2023

Systematic review with meta-analysis and meta-regression of randomised clinical trials of qigong, tai chi or yoga with anxiety or depression symptoms as outcomes, with risk of bias and certainty of evidence assessment; five databases to July 2022Verified from abstract
Open original source

Yoga, mindfulness-based stress reduction and stress-related physiological measures: A meta-analysis

Pascoe MC, Thompson DR, Ski CF · Psychoneuroendocrinology · 2017

Systematic review and meta-analysis of randomised controlled trials comparing practices including yoga asanas, with or without mindfulness-based stress reduction, with an active control on stress-related physiological measures (42 studies)Verified from abstract
Open original source

The effect of yoga on sleep quality and insomnia in women with sleep problems: a systematic review and meta-analysis

Wang WL, Chen KH, Pan YC, Yang SN, Chan YY · BMC Psychiatry · 2020

Systematic review and meta-analysis of randomised controlled trials of yoga versus control in women with sleep problems, with Cochrane risk of bias assessment; six databases searched in June 2019Verified from abstract
Open original source

Yoga-based exercise improves balance and mobility in people aged 60 and over: a systematic review and meta-analysis

Youkhana S, Dean CM, Wolff M, Sherrington C, Tiedemann A · Age and Ageing · 2016

Systematic review and meta-analysis of trials of physical yoga on balance in people aged 60 and over, with PEDro scale methodological quality assessment and random-effects pooling; searches to February 2015 (PROSPERO CRD42015015872)Verified from abstract
Open original source

Community

Discussion

Have you tried this? What helped or made it difficult?

Reader experiences are separate from the scientific evidence. Explore all discussions

No discussions yet.

Be the first to ask a question or share your experience.

Sign in to vote or join the discussion.

Related habits

Explore other habits in a shared category. Each has its own evidence and precautions.

Editorial history and review method

Revision 4 · Approved for publication · Codex (publication approval explicitly authorized by site owner; category migration)

Prepared and checked with AI. Publication approval does not imply independent expert review. Request a correction.

Applied the group-2 audit. The card and verdict no longer claim that no effect survives comparison with an active intervention, which the entry's own anxiety outcome (SMD -0.86 versus active comparators) and Li 2026 contradict; Li 2026 was retrieved and is now cited from chronic-low-back-pain, where its I2 of 95% was already quoted without a source. Implementation 2 was rewritten: the one randomised trial of virtual delivery compared live-streamed classes with a wait list, not with in-person classes, and its authors report that improvement occurred even though only 36.6% of participants attended at least half the classes, which is the opposite of the old 'attendance was the main thing' claim. The anxiety effectEstimate no longer lists reduced high-frequency heart-rate variability as if it were a benefit: the wording is the source's, and lower HF-HRV means lower vagal tone, which is now stated in the field rather than silently corrected. Precaution 3 now attributes the DSM-diagnosed null to Cramer 2018 and sets it beside Martinez-Calderon 2023's very-low-certainty positive finding. Second reviewer must re-check: (a) whether precaution 4's list of cautions should be presented at all, since no retrieved source supports it; (b) whether the entry should adopt Li 2026's function and wellbeing findings in the card headline; (c) the adverse-event analysis in Wieland 2022 used 8 of 21 trials, so the entry's 'a Cochrane review of 21 back pain trials' is accurate about the review but loose about that analysis. Publication release 2026-10-04: site owner authorized a bounded AI audit and release of this exact revision. See docs/release-audit-2026-10-04.md for source-check findings and limits. No independent health-professional or peer review is claimed. Category migration 2026-10-04: removed the separate goals field; categories: Exercise, Energy → Exercise, Energy, Stress & wellbeing. Research assessments, sources, guidance, precautions, and evidence review dates are unchanged. This revision awaits publication approval; no new evidence review is claimed. Publication approval 2026-10-04: site owner explicitly accepted all current category-migration revisions for deployment and requested that remaining limitations be retained in docs/content-review-backlog.md. Approval applies to this exact content revision. No new evidence review, independent expert review, or clinical validation is claimed; existing precautions and disclosures remain in effect.