Breathwork

Some human studies found a benefit. More research could change this.

All 4 studies with a clear result found a benefit. Strongest study: a randomized controlled trial.

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. A trial found it after a single session.

A 2024 randomized trial of 144 people found nausea in one patient in each intervention group, compared with six in the controls. A review also reported reduced motion-sickness symptoms in an analysis of 60 people.

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.

In a 2024 randomized trial of 90 people, the breathing intervention group had a between-group difference of 8.05 at 3 months of follow-up. A 2023 trial of 74 people also reported longer breath-holding duration with pranayama-assisted cognitive behavioral therapy.

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. Trials that found it ran as little as 6 weeks, typically 7.

A 2025 randomized trial of 40 people found improved upper-body flexibility compared with the control group. In another trial, chest expansion improved in 14 participants (93%) in the breathing-exercise group.

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. Trials that found it ran as little as 8 weeks, typically 10.

A 2026 randomized trial of 64 women found a 5.01-point greater increase in total sexual function scores with breathing exercises than with the control; 23.33% of the breathing-exercise group exceeded the specified score threshold, compared with none in the control group. A separate trial in men reported better intravaginal ejaculatory latency time when diaphragmatic breathing was added to treatment and pelvic floor muscle training.

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 trial found it after a single session.

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. Trials that found it ran as little as 7 weeks, typically 11.

A 2026 randomized trial of 62 people found fasting glucose was reduced by -6% compared with usual care. A review also reported a significant decrease in fasting blood sugar in one trial.

Studies disagree: some found a benefit, others found none. Do not count on it.

4 of 6 studies found a benefit; 2 found none. Strongest study: a pooled analysis of several trials.

Studies disagree: some found a benefit, others found none. Do not count on it.

15 of 20 studies found a benefit; 5 found none. Strongest study: a pooled analysis of several trials.

A meta-analysis pooling randomized and non-randomized human studies found a decrease of 4.4 points in Epworth Sleepiness Scale scores with inspiratory exercises. Sleep-quality findings are less consistent: one review reported improvement in one of two randomized trials, while a later review described its overall direction as neutral.

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 pooled analysis of several trials.

A pooled analysis of randomized trials involving 439 people found a small reduction in heart rate with pranayama. In a 2026 randomized trial, heart rate fell from 69.32 to 64.86 bpm in the pranayama group, while it rose in the control group.

Studies disagree: some found a benefit, others found none. Do not count on it.

3 of 4 studies found a benefit; 1 found none. Strongest study: a pooled analysis of several trials.

Pooled analyses whose designs were not recorded found improved neck alignment (131 people in the analysis). A separate review noted that only one study evaluated abdominal muscle activation and postural control.

Studies disagree: some found a benefit, others found none. Do not count on it.

44 of 52 studies found a benefit; 8 found none. Strongest study: a pooled analysis of several trials.

Studies disagree: some found a benefit, others found none. Do not count on it.

15 of 19 studies found a benefit; 4 found none. Strongest study: a pooled analysis of several trials.

Pooled analyses whose designs were not recorded found improvements in six-minute walk test distance, including a mean difference of 36.42 in 2025. Results differ across studies: a pooled analysis of randomized trials found no improvement in six-minute walk distance.

Studies disagree: some found a benefit, others found none. Do not count on it.

14 of 17 studies found a benefit; 3 found none. Strongest study: a pooled analysis of several trials.

Pooled analyses whose designs were not recorded found lower stress scores, including among pregnant women (355 people in the maternal stress analysis) and in an analysis of 785 people. Another pooled analysis found no effect on stress, so findings differ across studies.

Studies disagree: some found a benefit, others found none. Do not count on it.

13 of 16 studies found a benefit; 3 found none. Strongest study: a pooled analysis of several trials.

Pooled analyses whose designs were not recorded reported lower systolic blood pressure, including a mean reduction of −7.06 in an analysis of 1,097 people. Results are not consistent: a pooled analysis of randomized trials found the systolic blood pressure reduction was not statistically significant under its random-effects model.

Studies disagree: some found a benefit, others found none. Do not count on it.

18 of 21 studies found a benefit; 3 found none. Strongest study: a pooled analysis of several trials.

A 2025 pooled analysis whose study designs were not recorded found a small reduction in post-intervention symptom severity with prāṇāyāma compared with passive controls (315 people in its analysis). A separate pooled analysis found no effect on overall mental health outcomes, so findings are mixed.

Studies disagree: some found a benefit, others found none. Do not count on it.

17 of 22 studies found a benefit; 5 found none. Strongest study: a pooled analysis of several trials.

Pooled analyses whose designs were not recorded found lower depressive symptom scores (g = −0.40); a separate review reported lower post-test scores after yogic breathing in an analysis of 40 people. A pooled analysis of randomized trials did not find an effect on anxiety and depression, so results are not consistent.

Studies disagree: some found a benefit, others found none. Do not count on it.

27 of 35 studies found a benefit; 8 found none. Strongest study: a pooled analysis of several trials.

Pooled analyses whose designs were not recorded found lower self-reported anxiety (g = −0.32) and, among patients with lung cancer, an average 3.42-point decrease in postoperative anxiety scores. However, a pooled analysis of randomized trials did not find an effect on anxiety and depression, so results are not consistent across studies.

Studies disagree: some found a benefit, others found none. Do not count on it.

14 of 19 studies found a benefit; 5 found none. Strongest study: a randomized controlled trial.

A 2024 review reported improved high-frequency heart-rate variability in one trial and increased heart-rate variability alongside a lower breathing rate in a 60-person diaphragmatic-breathing study. Results differ across analyses: a pooled analysis of randomized trials found a heart-rate-variability effect in one statistical model, but not another.

Studies disagree: some found a benefit, others found none. Do not count on it.

3 of 4 studies found a benefit; 1 found none. Strongest study: a randomized controlled trial.

A review covering 31 people in its analysis reported that one breathing session reduced stop-signal reaction time, while go reaction time increased. A separate 2025 randomized trial of 86 people found lower auditory and visual reaction times and P300 latency in the test group than in the control group.

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.

A pooled analysis whose study designs were not recorded found lower malondialdehyde levels in the experimental group than in the control group.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

In a 2025 randomized trial of 404 healthy participants, the breathing protocol produced greater momentary self-reported energy increases than meditation after daily practice.

Too few or too small studies to know if it works.

7 of 9 studies found a benefit; 2 found none. Strongest study: a randomized controlled trial.

In a 2025 randomized trial of 45 people, quadriceps strength increased more with inspiratory muscle training than in the control group. Another trial reported increased pelvic floor muscle strength and endurance after 8 weeks in men with premature ejaculation; a pooled analysis of randomized trials found no additional benefit when breathing exercises were added to pelvic floor muscle training.

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.

Pooled analyses whose designs were not recorded included 476 people in the GERD symptom analysis and found lower GERD symptom scores with diaphragmatic breathing. Results varied substantially across studies.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a systematic review of studies.

Pooled analyses whose designs were not recorded included 61 people in the attention-task analysis; a single 18-minute session incorporating Kapalbhati increased total attempts and net scores on an attention-based cancellation task compared with a control group.

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.

Too few or too small studies to know if it works.

All 3 studies with a clear result found a benefit. Strongest study: a randomized controlled trial.

A 2024 randomized trial of 44 people found a decrease in IL‐6. A review reported that TNF-α decreased throughout a 3-month training period.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a systematic review of studies.

A review covering 516 people in its analysis reported that five studies found significant decreases in thyroid-stimulating hormone levels.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a systematic review of studies.

A review covering 516 people in its analysis reported a significant reduction in body mass index.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a systematic review of studies.

A review covering 27 people in its analysis reported increased natural killer cell counts at 12 and 24 weeks compared with baseline, and higher counts than the control group at 24 weeks.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

In a randomized trial involving 76 people with Breast Cancer, lack-of-appetite scores were lower in the intervention group than in the control group. This finding reflects reduced reported lack of appetite, not a measured increase in appetite.

Too few or too small studies to know if it works.

3 of 4 studies found a benefit; 1 found none. Strongest study: a randomized controlled trial.

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.

A 2025 randomized trial of 90 people found improved scores on the social dysfunction subscale after both online and brochure-based interventions. Participants in a separate qualitative study described the shared experience as helpful when mobility was limited.

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.

In a 2025 randomized trial of 85 people with GERD, bloating decreased over the study period, but the decrease was statistically significant only at Week 4 after correction for multiple comparisons. A review concluded that diaphragmatic breathing training may relieve symptoms in selected patients with GERD.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a non-randomized trial.

A quasi-experimental study in 46 healthy participants found self-acceptance rose from 24.05 to 27.85.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

In a 2023 randomized trial of 74 people, the pranayama group had a 5.41-point decrease in PTSD symptom scores.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

A randomized trial involving 76 people with Breast Cancer found fewer reported problems remembering on the 7th day after chemotherapy in the intervention group than in the control group; this difference was not reported for other measurement times.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

In a 2022 randomized trial of 120 people with lung cancer, self-efficacy scores were higher than in the control group at 60 days.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a small pilot study.

In a 2026 pilot involving 85 people with upper abdominal pain, pain scores fell over time in both groups; scores at each follow-up week were lower than at baseline.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a randomized controlled trial.

In a randomized trial involving 24 healthy people, the breathwork group reported a higher EBI score than the meditation group.

Too few or too small studies to know if it works.

The one study found a benefit. Strongest study: a narrative review.

A review describes an 8–9% improvement in cognition associated with memory consolidation, but human evidence has not been established for this finding.

Only tested in animals, cells or theory. Results like these often do not carry over to people.

The one study found a benefit. Strongest study: an interview study.

In a qualitative study of 20 people, participants described improved movement and function, feeling more mobile,

Only tested in animals, cells or theory. Results like these often do not carry over to people.

The one study found a benefit.

Slow, controlled breathing may increase pain tolerance by activating descending inhibitory pain pathways and reducing anxiety.

Only tested in animals, cells or theory. Results like these often do not carry over to people.

The one study found a benefit.

Breathwork is theorized to support detoxification pathways through enhanced lymphatic flow and CO2-driven vasodilation, but direct evidence of toxin clearance is limited to theoretical models.

Only people's own stories online. No study has tested it.

The one study found a benefit.

Expert knowledge describes slow, paced breathing with a prolonged exhale as increasing respiratory sinus arrhythmia and vagally mediated heart-rate variability in healthy adults, shifting autonomic balance toward parasympathetic activity. The effect is strongest during the practice and smaller with fast breathing; human evidence is not reported.

Only people's own stories online. No study has tested it.

The one study found a benefit.

Focused breathing practices are described as enhancing interoceptive awareness—the ability to perceive internal bodily sensations.

Only people's own stories online. No study has tested it.

The one study found a benefit.

Diaphragmatic breathing at 4-6 breaths per minute is described as acutely increasing vagal tone, indexed by heart rate variability.

Only people's own stories online. No study has tested it.

The one study found a benefit.

Breath-focused attention is described as a way to anchor the mind in the present moment and reduce repetitive negative thinking.

Studies looked for a benefit and did not find one.

The one study found no benefit. Strongest study: a pooled analysis of several trials.

A pooled analysis whose study designs were not recorded included 111 people in its nitric oxide analysis and found no significant difference between groups.

Studies looked for a benefit and did not find one.

The one study found no benefit. Strongest study: a systematic review of studies.

Human evidence supporting an improvement in urinary flow has not been established.

Studies looked for a benefit and did not find one.

The one study found no benefit. Strongest study: a randomized controlled trial.

A randomized trial involving 30 healthy people did not support an improvement: it found no changes in two-point discrimination, nine-hole peg test time, or position sense on either side.

Studies looked for a benefit and did not find one.

The one study found no benefit. Strongest study: a randomized controlled trial.

A randomized trial involving 96 people did not support a reduction in triglycerides: the difference versus control was -36.1 mg/dL.