body

Balance Improvement

See how Balance Improvement fits within the body category and what the term is intended to describe.

36Practices

These are research findings, not promises. Most options have little evidence, and results differ from person to person.

What may help

What studies found for each option, best evidence first.

36 options:1 good evidence5 some evidence12 mixed13 not enough evidence1 personal stories only4 no benefit found

What studies foundSize of effect

CardiopracticeGood evidence it helpsSize of effect: Large
Good human studies found a benefit, and they mostly agree. This is about what studies measured, not a cure.
All 3 studies with a clear result found a benefit. Strongest study: a pooled analysis of several trials.
A pooled analysis whose designs were not recorded found higher balance scores on the Berg Balance Scale; higher scores indicate better balance. Another pooled analysis found greater improvement among people with Parkinson's disease whose aerobic exercise dose had high compliance with ACSM recommendations.
Open the Cardio page
SwimmingpracticeSome evidence it helpsSize of effect: Not known
Some human studies found a benefit. More research could change this.
All 16 studies with a clear result found a benefit. Strongest study: a pooled analysis of several trials.
Open the Swimming page
BreathworkpracticeSome evidence it helpsSize of effect: Not known
Some human studies found a benefit. More research could change this.
All 2 studies with a clear result found a benefit. Strongest study: a pooled analysis of several trials.
A pooled analysis whose study designs were not recorded found a 3.25-point improvement on the Berg Balance Scale. A 2025 randomized trial of 40 people also reported improved dynamic balance.
Open the Breathwork page
Foam RollingpracticeSome evidence it helpsSize of effect: Not known
Some human studies found a benefit. More research could change this.
8 of 10 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.
Trials reported improved dynamic balance and reach-test scores, including a 4.61 cm between-group difference in one trial of 40 people. Another trial of 40 found no significant effects on balance, so results differ across the studied groups, which included people with Back Pain and Osteoarthritis.
Open the Foam Rolling page
CyclingpracticeSome evidence it helpsSize of effect: Not known
Some human studies found a benefit. More research could change this.
All 3 studies with a clear result found a benefit. Strongest study: a randomized controlled trial.
Open the Cycling page
MassagepracticeSome evidence it helpsSize of effect: Not known
Some human studies found a benefit. More research could change this.
All 2 studies with a clear result found a benefit. Strongest study: a randomized controlled trial.
A pooled analysis whose designs were not recorded found improved dynamic balance, particularly on the right foot, among Taekwondo athletes receiving a 10MMP or 15MMP rather than NMP; 12 people were included in this outcome’s analysis. A separate trial found better YBT composite scores on the dominant limb.
Open the Massage page
High-Intensity Interval Training (HIIT)practiceMixed resultsSize of effect: Small
Studies disagree: some found a benefit, others found none. Do not count on it.
9 of 12 studies found a benefit; 3 found none. Strongest study: a pooled analysis of several trials.
Pooled analyses of randomized trials reported improved balance, including a timed-up-and-go measure in an analysis of 125 people and the Berg Balance Scale in another analysis. The latter result had high heterogeneity.
Open the High-Intensity Interval Training (HIIT) page
Hatha YogapracticeMixed resultsSize of effect: Large
Studies disagree: some found a benefit, others found none. Do not count on it.
8 of 12 studies found a benefit; 4 found none. Strongest study: a pooled analysis of several trials.
A meta-analysis of randomized trials found improved static balance compared with NIC. A separate review described the evidence for balance and multicomponent physical-function measures as low-certainty.
Open the Hatha Yoga page

Mobility TrainingpracticeMixed resultsSize of effect: Not known
Studies disagree: some found a benefit, others found none. Do not count on it.
7 of 10 studies found a benefit; 3 found none. Strongest study: a randomized controlled trial.
A systematic review of randomized trials involving 457 people found that cognitive-motor balance training improved single-task balance function (standardized effect size = 0.27-1.82). In a randomized controlled trial of 32 patients with Parkinson's Disease, both synchronous telerehabilitation-based exercise protocols enhanced balance and gait, as well as activity level and quality of life.
Open the Mobility Training page
Tai ChipracticeMixed resultsSize of effect: Medium
Studies disagree: some found a benefit, others found none. Do not count on it.
51 of 57 studies found a benefit; 6 found none. Strongest study: a pooled analysis of several trials.
A pooled analysis whose study designs were not recorded found improved balance scores among older adults, including 190 people in its outcome analysis in 2025. Results were not consistent across studies: a pooled analysis of randomized trials involving 240 people found no statistically significant effect before one article was removed.
Open the Tai Chi page
Strength TrainingpracticeMixed resultsSize of effect: Medium
Studies disagree: some found a benefit, others found none. Do not count on it.
24 of 27 studies found a benefit; 3 found none. Strongest study: a pooled analysis of several trials.
Open the Strength Training page
PilatespracticeMixed resultsSize of effect: Medium
Studies disagree: some found a benefit, others found none. Do not count on it.
45 of 54 studies found a benefit; 9 found none. Strongest study: a pooled analysis of several trials.
A pooled analysis of randomized trials found a moderate improvement in balance among 930 people. Results conflict: a separate pooled analysis using the Berg Scale found no significant difference between Pilates and control groups.
Open the Pilates page
QigongpracticeMixed resultsSize of effect: Large
Studies disagree: some found a benefit, others found none. Do not count on it.
16 of 22 studies found a benefit; 6 found none. Strongest study: a pooled analysis of several trials.
A 2026 meta-analysis of randomized trials found a 3.98-point improvement on the Berg Balance Scale among 3,743 people in its analysis. Another 2026 meta-analysis found no statistically significant difference in proprioception, so the findings vary by the balance-related measure assessed.
Open the Qigong page
DancepracticeMixed resultsSize of effect: Small
Studies disagree: some found a benefit, others found none. Do not count on it.
20 of 22 studies found a benefit; 2 found none. Strongest study: a pooled analysis of several trials.
Open the Dance page
Social DancingpracticeMixed resultsSize of effect: Not known
Studies disagree: some found a benefit, others found none. Do not count on it.
8 of 10 studies found a benefit; 2 found none. Strongest study: a pooled analysis of several trials.
Open the Social Dancing page
Brisk WalkingpracticeMixed resultsSize of effect: Not known
Studies disagree: some found a benefit, others found none. Do not count on it.
5 of 7 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.
A 2022 review pooling randomized trials found better Get Up and Go test results in the walking intervention group. In a 2016 trial, the difference between groups was confirmed only for mean center-of-pressure velocity in the anterior-posterior direction with eyes closed.
Open the Brisk Walking page
TENS / EMS TherapypracticeMixed resultsSize of effect: Not known
Studies disagree: some found a benefit, others found none. Do not count on it.
6 of 8 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.
A review covering 243 people with conditions including stroke, multiple sclerosis, and peripheral neuropathy reported reduced spasticity and improved balance after TENS. In one randomized trial, low-frequency, high-amplitude TENS reduced envelope area by 20.7% under the eyes-open condition.
Open the TENS / EMS Therapy page
Red Light Therapy (Photobiomodulation)practiceMixed resultsSize of effect: Not known
Studies disagree: some found a benefit, others found none. Do not count on it.
3 of 5 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.
In studies involving people with Parkinson’s disease, assessments reported improvements in mobility and balance tests, including the Timed Up and Go, Step Test and Single-Leg Stance; 19 people were included in this outcome analysis. A separate randomized trial of 72 reported positive effects on posturography outcomes.
Open the Red Light Therapy (Photobiomodulation) page
YogapracticeNot enough evidence
Too few or too small studies to know if it works.
11 of 15 studies found a benefit; 4 found none. Strongest study: a pooled analysis of several trials.
A 2025 pooled analysis of randomized trials found improved static balance compared with no intervention among four trials. Results differed when yoga was compared with other exercise therapies: a 2026 pooled analysis found no statistically significant differences.
Open the Yoga page
CalisthenicspracticeNot enough evidence
Too few or too small studies to know if it works.
3 of 6 studies found a benefit; 3 found none. Strongest study: a randomized controlled trial.
A 2022 randomized trial of 17 participants found increased posterolateral reach on the Y Balance Test when standing on the left leg; a 2025 quasi-experimental study found static balance increased by 56% in the flexible flatfeet group and 41.2% in the normal feet group. In contrast, a 2023 randomized trial of 69 found no significant changes in the reported balance-related measures.
Open the Calisthenics page
AcupuncturepracticeNot enough evidence
Too few or too small studies to know if it works.
The one study found a benefit. Strongest study: a pooled analysis of several trials.
In people with stroke, a pooled analysis of randomized trials found balance scores favored acupuncture on the Berg Balance Scale. The result varied substantially across studies and was rated very low certainty.
Open the Acupuncture page
Vinyasa YogapracticeNot enough evidence
Too few or too small studies to know if it works.
All 6 studies with a clear result found a benefit. Strongest study: a randomized controlled trial.
An 80-person randomized trial in 2020 found an average Balance Index improvement of 5.6 mm/s. A separate randomized trial also found balance improved over time, including among participants studied in connection with Menopause.
Open the Vinyasa Yoga page
Ice Bath / Cold PlungepracticeNot enough evidence
Too few or too small studies to know if it works.
1 of 4 studies found a benefit; 3 found none. Strongest study: a systematic review of studies.
A review involving 225 healthy people found varied balance results: some studies suggested improvement, while others found no effect. Overall, the review described more positive than negative findings, but noted that results were contradictory.
Open the Ice Bath / Cold Plunge page
Latin Dance FitnesspracticeNot enough evidence
Too few or too small studies to know if it works.
2 of 3 studies found a benefit; 1 found none. Strongest study: a randomized controlled trial.
Open the Latin Dance Fitness page
Jump Rope TrainingpracticeNot enough evidence
Too few or too small studies to know if it works.
2 of 4 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.
Open the Jump Rope Training page
Brain Games / PuzzlespracticeNot enough evidence
Too few or too small studies to know if it works.
2 of 3 studies found a benefit; 1 found none. Strongest study: a randomized controlled trial.
Open the Brain Games / Puzzles page
Hot YogapracticeNot enough evidence
Too few or too small studies to know if it works.
All 2 studies with a clear result found a benefit. Strongest study: a non-randomized trial.
In a quasi-experimental study with 21 participants, balance time increased by an average of 228%, from 19.5 +/- 14 to 34.3 +/- 18 seconds; the control group did not change.
Open the Hot Yoga page
Forest Bathing (Shinrin-yoku)practiceNot enough evidence
Too few or too small studies to know if it works.
All 2 studies with a clear result found a benefit. Strongest study: a randomized controlled trial.
An RCT involving 88 people found small improvements in both groups over time: 1.16% in the HG group and 2.63% in the FTG group. A pilot study in people with cancer also reported changes in SLS and Y-Balance scores.
Open the Forest Bathing (Shinrin-yoku) page
Indoor RowingpracticeNot enough evidence
Too few or too small studies to know if it works.
1 of 2 studies found a benefit; 1 found none. Strongest study: a randomized controlled trial.
A randomized trial in 38 people with osteoarthritis found functional reach increased by 3.74 cm.
Open the Indoor Rowing page
Gardening TherapypracticeNot enough evidence
Too few or too small studies to know if it works.
All 2 studies with a clear result found a benefit. Strongest study: a systematic review of studies.
A 2021 systematic review reported that falls decreased from 288 to 200 in the year after a garden opened; the review examined people with dementia.
Open the Gardening Therapy page
Tabata TrainingpracticeNot enough evidence
Too few or too small studies to know if it works.
The one study found a benefit. Strongest study: a randomized controlled trial.
A 2021 randomized trial of 48 healthy participants found a time-by-group interaction on the Y-balance test composite score.
Open the Tabata Training page
Swedish (Classic) MassagepracticeOnly personal stories
Only people's own stories online. No study has tested it.
Based on one study.
Open the Swedish (Classic) Massage page
Thai MassagepracticeNo benefit found
Studies looked for a benefit and did not find one.
0 of 2 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.
Balance has been examined in people with obesity or overweight, but there is no human evidence supporting an improvement.
Open the Thai Massage page
Mindfulness MeditationpracticeNo benefit found
Studies looked for a benefit and did not find one.
The one study found no benefit. Strongest study: a pooled analysis of several trials.
Studies in people with Parkinson's disease did not support improved balance; a pooled analysis of randomized and non-randomized human studies described the evidence for benefit on balance and mobility as limited.
Open the Mindfulness Meditation page
MeditationpracticeNo benefit found
Studies looked for a benefit and did not find one.
The one study found no benefit. Strongest study: a systematic review of studies.
Open the Meditation page
Anusara YogapracticeNo benefit found
Studies looked for a benefit and did not find one.
1 of 2 studies found a benefit; 1 found none.
The studies did not support a balance benefit. A 2026 meta-analysis pooling randomized and non-randomized human studies involving 1,471 people in its balance analysis found that potential benefits for dynamic balance in unadjusted analyses became non-significant after trim-and-fill adjustment.
Open the Anusara Yoga page

Show all 36

Comparing practices

CardioGood evidence it helpsSize of effect: Large3 trials
Hatha YogaMixed resultsSize of effect: Large8 trials
QigongMixed resultsSize of effect: Large14 trials

The expert view shows every study, dose protocol and source.

What may help

Linked to improving it36 listed

Start here

Highest probability, strongest evidence

PracticesCardioCertaintyHighSourceMeta-analysisLarge effect3 other sources
A pooled analysis whose designs were not recorded found higher balance scores on the Berg Balance Scale (SMD = 1.15; 95% CI, 0.90 to 1.41; P < 0.00001); higher scores indicate better balance. Another pooled analysis found greater improvement among people with Parkinson's disease whose aerobic exercise dose had high compliance with ACSM recommendations.
How big is the effect
Large effectA big difference on average. Check how strong the evidence is.
Who was studied
People with a condition
Study result
Aerobic exercise produced significantly greater improvements in BBS scores than control (SMD = 1.15; 95% CI, 0.90 to 1.41; P < 0.00001; I² = 28%); because higher BBS scores indicate better balance, the positive SMD indicates a statistically significant improvement in favour of aerobic exercise.
Open Cardio
→

Medium probability(17)

PracticesHigh-Intensity Interval Training (HIIT)CertaintyMediumSourceMeta-analysisSmall effect12 other sources
Pooled analyses of randomized trials reported improved balance, including a timed-up-and-go measure in an analysis of 125 people and the Berg Balance Scale in another analysis. The latter result had high heterogeneity (I² = 93.13%).
How big is the effect
Small effectA small but noticeable difference on average.
Who was studied
Healthy people and patients
Coronary Artery Disease
Diabetes Type2
Elderly
Heart Disease
Heart Failure
Neuropathy Peripheral
+3 more
Study result
HIIT significantly improved balance (e.g., timed-up-and-go) (g = -0.79, 95%CI = [-1.19, -0.40], P = 0.035)
Open High-Intensity Interval Training (HIIT)
→
PracticesSwimmingCertaintyMediumSourceMeta-analysis16 other sources
For people recovering from Stroke, a pooled analysis whose designs were not recorded found aquatic exercise improved Berg Balance Scale scores compared with a control group (MD = 3.43, 95% CI = (0.23, 6.63)).
Who was studied
People with a condition
Stroke
Study result
AR, AUE, BT, CT, Dance, Exergame, and FT showed significant differences compared to Flexibility (p < 0.05)
Open Swimming
→
PracticesStrength TrainingCertaintyMediumSourceMeta-analysisModerate effectconflicting findings27 other sources
Findings disagree: one pooled analysis whose study designs were not recorded found a moderate improvement in balance among 2,097 people in its balance analysis, while another found isolated strength training had limited effects on dynamic balance (SMD 0.21, 95% CI [−0.05, 0.47]). A separate analysis described the reductions as small and below the MID threshold for clinical importance.
How big is the effect
Moderate effectA clear difference on average.
Who was studied
Healthy people and patients
Obesity
Sarcopenia
Stroke
Study result
Low-certainty evidence indicated that RT ... led to only small reductions that did not reach the MID threshold and are therefore unlikely to be clinically important.
Open Strength Training
→
PracticesTai ChiCertaintyMediumSourceMeta-analysisModerate effectconflicting findings58 other sources
A pooled analysis whose study designs were not recorded found improved balance scores among older adults, including 190 people in its outcome analysis in 2025. Results were not consistent across studies: a pooled analysis of randomized trials involving 240 people found no statistically significant effect before one article was removed.
How big is the effect
Moderate effectA clear difference on average.
Who was studied
Healthy people and patients
Alzheimers Disease
Dementia
Elderly
Intellectual Disability
Osteoarthritis
Parkinsons Disease
+2 more
Study result
The standardized mean difference (SMD) was −0.76, with a 95% confidence interval (CI) ranging from −1.46 to −0.07 (p=0.03). The overall effect was statistically significant, suggesting that Tai Chi has a significant effect on improving balance ability in the older adults.
Open Tai Chi
→
PracticesHatha YogaCertaintyMediumSourceRCTLarge effect13 other sources
A meta-analysis of randomized trials found improved static balance compared with NIC (SMD = 2.36; 95% CI 1.13-3.58). A separate review described the evidence for balance and multicomponent physical-function measures as low-certainty.
How big is the effect
Large effectA big difference on average. Check how strong the evidence is.
Who was studied
Healthy people and patients
Back Pain
Parkinsons Disease
Study result
there was ... low-certainty evidence for balance and multicomponent physical function measures
Open Hatha Yoga
→
PracticesMobility TrainingCertaintyMediumSourceRCT11 other sources
A systematic review of randomized trials involving 457 people found that cognitive-motor balance training improved single-task balance function (standardized effect size = 0.27-1.82). In a randomized controlled trial of 32 patients with Parkinson's Disease, both synchronous telerehabilitation-based exercise protocols enhanced balance and gait, as well as activity level and quality of life.
Who was studied
Healthy people and patients
Dementia
Neuropathy Diabetic
Parkinsons Disease
Stroke
Study result
Cognitive-motor balance training was effective in improving single-task balance function (standardized effect size = 0.27-1.82).
Open Mobility Training
→
PracticesBreathworkCertaintyMediumSourceRCT2 other sources
A pooled analysis whose study designs were not recorded found a 3.25-point improvement on the Berg Balance Scale. A 2025 randomized trial of 40 people also reported improved dynamic balance.
Who was studied
People with a condition
Overweight
Study result
breathing training significantly improved Berg Balance Scale (BBS) score (WMD, 3.25; 95%CI, 1.89 to 4.60, P < 0.00001, I^2 = 58%)
Open Breathwork
→
PracticesDanceCertaintyMediumSourceRCTSmall effectconflicting findings22 other sources
A review covering 1,259 people reported better balance scores with dance, including in comparisons with traditional therapeutic interventions. Findings differ across pooled analyses: one reported improved balance (SMD = 0.4090; 95% CI: [0.0983, 0.7198]), while another reported a confidence interval for static balance that crossed zero (95% CI -0.47; 1.32).
How big is the effect
Small effectA small but noticeable difference on average.
Who was studied
Healthy people and patients
Alzheimers Disease
Dementia
Mild Cognitive Impairment
Parkinsons Disease
Study result
Balance scores were significantly better in the experimental group... Indian folk dance as an intervention significantly improved balance, fall risk, functional ability, and QoL, with similar or even better results than traditional therapeutic interventions.
Open Dance
→
PracticesPilatesCertaintyMediumSourceRCTModerate effectconflicting findings54 other sources
A pooled analysis of randomized trials found a moderate improvement in balance among 930 people. Results conflict: a separate pooled analysis using the Berg Scale found no significant difference between Pilates and control groups.
How big is the effect
Moderate effectA clear difference on average.
Who was studied
Healthy people and patients
Addiction
Back Pain
Breast Cancer
Diabetes Type2
Down Syndrome
Gestational Diabetes
+11 more
Study result
The overall estimated effect size was estimated at 0.65, SE = 0.0817, Z = 7.985, p < 0.001, CI [0.4924 to 0.8127]
Open Pilates
→
PracticesQigongCertaintyMediumSourceRCTLarge effectconflicting findings23 other sources
A 2026 meta-analysis of randomized trials found a 3.98-point improvement on the Berg Balance Scale among 3,743 people in its analysis. Another 2026 meta-analysis found no statistically significant difference in proprioception, so the findings vary by the balance-related measure assessed.
How big is the effect
Large effectA big difference on average. Check how strong the evidence is.
Who was studied
Healthy people and patients
Atrial Fibrillation
Back Pain
Breast Cancer
Chronic Fatigue Syndrome
Copd
Depression
+11 more
Study result
TC [MD = 3.98, 95%CrI (2.91, 5.02)] improved balance on the Berg Balance Scale (BBS, higher = better).
Open Qigong
→
PracticesSocial DancingCertaintyMediumSourceRCTconflicting findings11 other sources
A meta-analysis of randomized trials found that creative dance improved Berg Balance Scale scores in older adults. Another review reported improvement compared with pharmacological usual care alone (MD = 3.24), but reviews disagree: some reported better balance or mobility, while others found mixed or non-significant effects.
Who was studied
Healthy people and patients
Alzheimers Disease
Arthritis
Dementia
Elderly
Heart Failure
Mild Cognitive Impairment
+3 more
Study result
Creative dance (CD) significantly outperformed the control group (CG) in health older adults (p < 0.05) on the Berg Balance Scale (BBS)
Open Social Dancing
→
PracticesBrisk WalkingCertaintyMediumSourceClinical7 other sources
A 2022 review pooling randomized trials found better Get Up and Go test results in the walking intervention group. In a 2016 trial, the difference between groups was confirmed only for mean center-of-pressure velocity in the anterior-posterior direction with eyes closed.
Who was studied
Healthy people and patients
Menopause
Osteoporosis
Study result
The results of the Get Up and Go test were significantly better in the intervention group in the study of regular long-term walking.
Open Brisk Walking
→
PracticesTENS / EMS TherapyCertaintyMediumSourceClinical8 other sources
A review covering 243 people with conditions including stroke, multiple sclerosis, and peripheral neuropathy reported reduced spasticity and improved balance after TENS. In one randomized trial, low-frequency, high-amplitude TENS reduced envelope area by 20.7% under the eyes-open condition.
Who was studied
People with a condition
Back Pain
Diabetes Type2
Multiple Sclerosis
Neuropathy Peripheral
Stroke
Study result
post-TENS reduction in spasticity and enhancement in balance ability
Open TENS / EMS Therapy
→
PracticesFoam RollingCertaintyMediumSourceClinical10 other sources
Trials reported improved dynamic balance and reach-test scores, including a 4.61 cm between-group difference in one trial of 40 people. Another trial of 40 found no significant effects on balance, so results differ across the studied groups, which included people with Back Pain and Osteoarthritis.
Who was studied
Healthy people and patients
Back Pain
Osteoarthritis
Study result
in the FR group, dynamic balance scores increased for both limbs at postintervention (p = 0.001)
Open Foam Rolling
→
PracticesRed Light Therapy (Photobiomodulation)CertaintyMediumSourceClinical5 other sources
In studies involving people with Parkinson’s disease, assessments reported improvements in mobility and balance tests, including the Timed Up and Go, Step Test and Single-Leg Stance; 19 people were included in this outcome analysis. A separate randomized trial of 72 reported positive effects on posturography outcomes.
Who was studied
Healthy people and patients
Alzheimers Disease
Cancer
Parkinsons Disease
Study result
Motor and functional outcomes were assessed in three studies from two independent clinical cohorts, involving Parkinson's disease (n = 19 patients) ... with tests reporting improvements in mobility (Timed Up and Go), balance (Step Test, Single-Leg Stance), fine motor skills (Nine-Hole Peg Test), and
Open Red Light Therapy (Photobiomodulation)
→
PracticesCyclingCertaintyMediumSourceClinical3 other sources
In a randomized trial of 44 people, both ECT and ECT-HV groups improved their TUG times by −1.71 ± 1.11 s and −1.9 ± 1.58 s, respectively, with mean changes of −16.7% and −17.3%; both groups improved more than the control group. A separate trial found TUG performance improved after ECC and standard rehabilitation, with larger improvements after ECC.
Who was studied
People with a condition
Study result
Significant improvements in TUG variation were observed in the ECT (−1.71 ± 1.11 s; mean change: −16.7%, p < .001) and ECT-HV groups (−1.9 ± 1.58 s; mean change: −17.3%, p < .01), with significantly greater reductions in the TUG in both groups compared with the control group (p < .01 and .001, respe
Open Cycling
→
PracticesMassageCertaintyMediumSourceClinical2 other sources
A pooled analysis whose designs were not recorded found improved dynamic balance, particularly on the right foot, among Taekwondo athletes receiving a 10MMP or 15MMP rather than NMP; 12 people were included in this outcome’s analysis. A separate trial found better YBT composite scores on the dominant limb.
Who was studied
Healthy people
Study result
significant improvements in dynamic balance, particularly in the right foot, among Taekwondo athletes who received a 10MMP or 15MMP compared to those in the NMP group
Open Massage
→

Low probability(14)

PracticesYogaCertaintyLowSourceRCTLarge effectconflicting findings16 other sources
A 2025 pooled analysis of randomized trials found improved static balance compared with no intervention among four trials (SMD = 2.36; 95% CI 1.13-3.58; P = 0.0002, I 2 = 93%). Results differed when yoga was compared with other exercise therapies: a 2026 pooled analysis found no statistically significant differences (p > 0.05).
How big is the effect
Large effectA big difference on average. Check how strong the evidence is.
Who was studied
Healthy people and patients
Alzheimers Disease
Cancer
Chronic Pain
Copd
Depression
Elderly
+10 more
Study result
Meta-analysis of four RCTs comparing yoga to NIC demonstrated significant improvements in static balance with yoga (SMD = 2.36; 95% CI 1.13-3.58; P = 0.0002, I 2 = 93%)
Open Yoga
→
PracticesAcupunctureCertaintyLowSourceRCTLarge effect2 other sources
In people with stroke, a pooled analysis of randomized trials found balance scores favored acupuncture on the Berg Balance Scale. The result varied substantially across studies (I² = 91.5%) and was rated very low certainty.
How big is the effect
Large effectA big difference on average. Check how strong the evidence is.
Who was studied
People with a condition
Stroke
Study result
Balance assessed by the Berg Balance Scale also favored acupuncture but showed substantial heterogeneity (SMD = 0.96, 95% CI = 0.35–1.56; I^2 = 91.5%; very low certainty).
Open Acupuncture
→
PracticesCalisthenicsCertaintyLowSourceClinical7 other sources
A 2022 randomized trial of 17 participants found increased posterolateral reach on the Y Balance Test when standing on the left leg; a 2025 quasi-experimental study found static balance increased by 56% in the flexible flatfeet group and 41.2% in the normal feet group. In contrast, a 2023 randomized trial of 69 found no significant changes in the reported balance-related measures.
Who was studied
Healthy people and patients
Frailty
Study result
significant increase in Y balance test posterolateral reach with left leg stance (p = 0.046)
Open Calisthenics
→
PracticesVinyasa YogaCertaintyLowSourceClinical7 other sources
An 80-person randomized trial in 2020 found an average Balance Index improvement of 5.6 mm/s. A separate randomized trial also found balance improved over time, including among participants studied in connection with Menopause.
Who was studied
Healthy people and patients
Menopause
Study result
an average improved Balance Index of 5.6 mm/s
Open Vinyasa Yoga
→
PracticesIce Bath / Cold PlungeCertaintyLowSourceClinical4 other sources
A review involving 225 healthy people found varied balance results: some studies suggested improvement, while others found no effect. Overall, the review described more positive than negative findings, but noted that results were contradictory.
Who was studied
Healthy people
Study result
The effects of CWI on balance in healthy individuals are varied, with some studies suggesting a positive impact and others suggesting no effect. Overall, its use yields more positive effects, although contradictions persist.
Open Ice Bath / Cold Plunge
→
PracticesLatin Dance FitnessCertaintyLowSourceClinical4 other sources
A 2020 trial in 98 healthy participants with Menopause reported one-stand leg test results of ET = 13.532 ± 2.65 and CnT = 11.656 ± 2.88. A 2025 trial found no group effect on other physical-performance variables.
Who was studied
Healthy people
Menopause
Study result
one-stand leg test: ET = 13.532 ± 2.65, p < 0.001; CnT = 11.656 ± 2.88, p < 0.001
Open Latin Dance Fitness
→
PracticesJump Rope TrainingCertaintyLowSourceClinical4 other sources
In an RCT of 16 participants, LDC scores rose from 49.50±5.71 to 67.37±5.87 in the exercise group and from 50.13±3.44 to 58.12±3.17 in the control group, with a significant interaction. A quasi-experimental study also found a higher composite score in the NSSG than in the SSG and CNTG.
Who was studied
Healthy people and patients
Leukemia
Study result
LDC (%) 49.50±5.71 to 67.37±5.87 (P<0.01) in ExG; ConG 50.13±3.44 to 58.12±3.17 (P<0.01); significant interaction (F=10.080, P=0.007)
Open Jump Rope Training
→
PracticesBrain Games / PuzzlesCertaintyLowSourceClinical3 other sources
A randomized trial of 36 healthy participants found differences in balance scores between game groups after training (SMD 0.59). Results were not uniform: another randomized trial of 31 found no significant training effects on physical functions or brain volume.
Who was studied
Healthy people
Study result
there was a significant difference between the groups scores in the posttest phase (F_2, 32=6.09, *P* = 0.006, η^2 = 0.27, Cohen's d = 0.59). The pairwise comparisons revealed significant differences between the Neurolight and Maghzineh groups (*P* = 0.021), as well as between the Neurolight and con
Open Brain Games / Puzzles
→
PracticesHot YogaCertaintyLowSourceClinical3 other sources
In a quasi-experimental study with 21 participants, balance time increased by an average of 228%, from 19.5 +/- 14 to 34.3 +/- 18 seconds; the control group did not change.
Study result
Percent change in balance time for individual yoga subjects averaged +228% (19.5 +/- 14 to 34.3 +/- 18 seconds, p < 0.05), with no change in controls.
Open Hot Yoga
→
PracticesForest Bathing (Shinrin-yoku)CertaintyLowSourceClinical2 other sources
An RCT involving 88 people found small improvements in both groups over time: 1.16% in the HG group and 2.63% in the FTG group. A pilot study in people with cancer also reported changes in SLS and Y-Balance scores.
Who was studied
Healthy people and patients
Cancer
Study result
A significant main effect for time (p < 0.01) indicates an improvement in both groups. However, the improvement is in the HG (T2-T1: 1.16%) and FTG (T2-T1: 2.63%) rather small.
Open Forest Bathing (Shinrin-yoku)
→
PracticesIndoor RowingCertaintyLowSourceClinical2 other sources
A randomized trial in 38 people with osteoarthritis found functional reach increased by 3.74 cm [0.68, 6.80].
Who was studied
People with a condition
Osteoarthritis
Study result
functional-reach (cm) (MD = 3.74 [0.68, 6.80], p = 0.018)
Open Indoor Rowing
→
PracticesGardening TherapyCertaintyLowSourceClinical2 other sources
A 2021 systematic review reported that falls decreased from 288 to 200 in the year after a garden opened; the review examined people with dementia.
Who was studied
People with a condition
Dementia
Study result
The number of falls decreased from 288 to 200 in the year after the garden was opened.
Open Gardening Therapy
→
PracticesTabata TrainingCertaintyLowSourceClinical1 other source
A 2021 randomized trial of 48 healthy participants found a time-by-group interaction on the Y-balance test composite score (F = 6.497, p = 0.014).
Who was studied
Healthy people
Study result
The Y-balance test of composite score showed an interaction between time and group (F = 6.497, p = 0.014)
Open Tabata Training
→
PracticesSwedish (Classic) MassageCertaintyLowSourceAnecdotal1 other source
This effect was mentioned in 3 of 241 Reddit reports discussing it.
→

Very low probability(4)

PracticesThai MassageCertaintyVery lowSourceClinical2 other sources
Balance has been examined in people with obesity or overweight, but there is no human evidence supporting an improvement.
Who was studied
People with a condition
Obesity
Overweight
Open Thai Massage
→
PracticesMindfulness MeditationCertaintyVery lowSourceClinical1 other source
Studies in people with Parkinson's disease did not support improved balance; a pooled analysis of randomized and non-randomized human studies described the evidence for benefit on balance and mobility as limited.
Who was studied
People with a condition
Parkinsons Disease
Open Mindfulness Meditation
→
PracticesMeditationCertaintyVery lowSourceClinical1 other source
Who was studied
People with a condition
Parkinsons Disease
Open Meditation
→
PracticesAnusara YogaCertaintyVery lowSourceTheoretical2 other sources
The studies did not support a balance benefit. A 2026 meta-analysis pooling randomized and non-randomized human studies involving 1,471 people in its balance analysis found that potential benefits for dynamic balance in unadjusted analyses became non-significant after trim-and-fill adjustment.
→

Practices compared

Practices with at least one trial of their own for this goal. Ranked by how sure the evidence is, then effect size.Top 10 of 31

Practice	How sure	Effect size	Own trials	What worked in trials
Cardio	CertaintyHighSourceMeta-analysis	Large effect	3 trials or pooled analyses	3–3.5×/week · 4–8 weeks
Qigong	CertaintyMediumSourceRCT	Large effect	14 trials or pooled analyses	45–60 min · 2–4×/week · 2–12 weeks
Hatha Yoga	CertaintyMediumSourceRCT	Large effect	8 trials or pooled analyses	60 min · 2–3×/week · 8–12 weeks
Tai Chi	CertaintyMediumSourceMeta-analysis	Moderate effect	45 trials or pooled analyses	40–60 min · 1–3×/week · 8–14 weeks
Pilates	CertaintyMediumSourceRCT	Moderate effect	43 trials or pooled analyses	45–60 min · 2–2.8×/week · 3–12 weeks
Strength Training	CertaintyMediumSourceMeta-analysis	Moderate effect	24 trials or pooled analyses	4–55 min · 2–3×/week · 4–11 weeks
Dance	CertaintyMediumSourceRCT	Small effect	15 trials or pooled analyses	30–60 min · 1–3×/week · 8–12 weeks
High-Intensity Interval Training (HIIT)	CertaintyMediumSourceMeta-analysis	Small effect	7 trials or pooled analyses	30 min · 3×/week · 3–13 weeks
Swimming	CertaintyMediumSourceMeta-analysis		14 trials or pooled analyses	45–50 min · 2–3×/week · 5–11 weeks
Foam Rolling	CertaintyMediumSourceClinical		8 trials or pooled analyses	2–12 min · 2–3×/week · 2–5 weeks

What worked in trials: From the smallest to the typical amount in the trials that found this benefit. Not the most anyone did, not a target and not a limit.

How to read the evidence

Evidence tier — overall strength of the evidence for the whole compound

StrongWell established — backed by strong, consistent human research.
ModerateSome genuine support, but mostly from small or lower-quality studies. Promising, not proven.
PreliminaryEarly-stage evidence — a promising hypothesis, not an established effect yet.
LimitedVery limited evidence — treat this as a hypothesis, not proof.
InsufficientNot enough good evidence to draw conclusions yet.

Certainty — how sure we can be that the effect is real

Very highVery sure — strong, consistent evidence
HighFairly sure — consistent evidence from good studies
MediumModerately sure — new studies could change this
LowNot sure — little or mixed evidence
Very lowVery unsure — almost no usable evidence

Source — where the data comes from

Meta-analysisPooled analysis of many trials — the strongest evidence
RCTRandomized controlled trial — a strong human study design
ClinicalA human clinical study
Reference knowledgeFrom general health-science or pharmacology references, not a study of this exact claim
Animal researchStudied in animals, not humans
Community reportsAggregated user reports — anecdotal signal, not evidence
TheoreticalMechanism-based reasoning, not yet tested

Study grade — quality of each underlying study (A best, D weakest)

AHigh-quality — strong, consistent human trials
BGood — solid evidence with some limitations
CModerate — limited, small, or mixed-quality studies
DWeak — preliminary, indirect, or animal-only

Confidence — how strongly the sources back a finding

A 0–100 score from study quality and agreement

Most supplement evidence is graded C–D. Read it as preliminary signal, not proof.

Full write-up

What is it

Balance Improvement describes better steadiness and control of body position during standing or movement. It is an outcome, not a treatment or diagnosis, and refers specifically to balance rather than a general improvement in physical performance. Activities studied in relation to this outcome include Tai Chi, Cycling, Swimming, and TENS / EMS Therapy.

How interventions help

Balance-related outcomes have been studied across several intervention types, including movement practices, rehabilitation-based exercise, electrical stimulation, and nutritional approaches. The movement interventions range from Tai Chi, yoga, cycling, and swimming to programs adapted for people with clinical conditions. Their outcomes are not interchangeable: studies have assessed fall risk, timed balance tasks, postural control, walking performance, and broader functional measures, so an improvement on one test does not necessarily represent the same change as an improvement on another.

The timing of an intervention also differs. Some findings followed a single acute exposure, such as Beetroot Extract at 140 mL over acute (2.5 h), while Caffeine was assessed during an acute test session and did not require daily use. Other approaches were evaluated over repeated use, including BCAA at 10.85 g/day for 16 weeks, Collagen at 2000 mg/day for 8 weeks, L-Leucine at 5000 mg/day for 1-4 weeks, Beta-Alanine at 6.4 g/day for 4 weeks, and Coenzyme Q10 at 100 mg/day for 8 weeks. These durations should not be treated as interchangeable, and the findings do not establish that benefits persist after stopping.

Results also vary by intervention and outcome. Several exercise and training approaches were associated with better balance or related functional performance, whereas Ice Bath / Cold Plunge did not significantly change proprioception after the intervention. Some supplement findings came from small studies or non-human evidence, and one Beta-Alanine comparison did not show a significant interaction. The populations studied also ranged from mixed groups and unspecified participants to people with clinical conditions, so applicability depends on how closely a person resembles the participants and the test used.

Choosing guidance

For a practice-based approach, Tai Chi, Cycling, Swimming, and yoga formats may suit readers looking for repeated movement training rather than a supplement. Tai Chi is especially relevant when the goal includes fall-risk reduction in older adults, although findings are not fully consistent. Yoga evidence is also mixed, so Yoga may suit readers willing to practice regularly while accepting greater uncertainty about results. Hatha Yoga and Vinyasa Yoga are more specific choices when the preferred format or ability-adapted practice matters.

Cycling and TENS / EMS Therapy were studied mainly in people with clinical conditions, so their findings may not transfer directly to otherwise healthy readers. Swimming was studied in a less clearly specified population. These options may be more appropriate when the activity or therapy matches the setting in which it was evaluated.

Timing and commitment differ. Beetroot Extract was examined as an acute option, while BCAA, Collagen, L-Leucine, Beta-Alanine, and Coenzyme Q10 were studied over periods of weeks. Caffeine was tested in an acute session, but its listed onset category is weeks, so it should not automatically be treated as a rapid balance aid. Longer-course options are a better conceptual fit for readers willing to maintain a daily or multiweek routine; acute options may appeal to those considering a time-limited trial.

The evidence is human for most listed options, but L-Leucine and Beta-Alanine are not classified that way, and Coenzyme Q10 is also not classified as human evidence. That distinction matters when choosing between an established movement approach and a supplement with more limited or less directly applicable support.